Citation
Dissimilar Cognitive Styles and their Relationships with Undergraduate Student Stress, Motivation, and Engagement

Material Information

Title:
Dissimilar Cognitive Styles and their Relationships with Undergraduate Student Stress, Motivation, and Engagement
Creator:
FRIEDEL, CURTIS ROBERT
Copyright Date:
2008

Subjects

Subjects / Keywords:
Cognitive style ( jstor )
College students ( jstor )
Educational incentives ( jstor )
Learner engagement ( jstor )
Learning motivation ( jstor )
Learning styles ( jstor )
Motivation ( jstor )
Student motivation ( jstor )
Students ( jstor )
Undergraduate students ( jstor )
Rolfs Hall ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Curtis Robert Friedel. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
3/1/2007
Resource Identifier:
659563271 ( OCLC )

Downloads

This item has the following downloads:

friedel_c ( .pdf )

friedel_c_Page_342.txt

friedel_c_Page_335.txt

friedel_c_Page_324.txt

friedel_c_Page_277.txt

friedel_c_Page_332.txt

friedel_c_Page_182.txt

friedel_c_Page_173.txt

friedel_c_Page_203.txt

friedel_c_Page_208.txt

friedel_c_Page_058.txt

friedel_c_Page_094.txt

friedel_c_Page_092.txt

friedel_c_Page_004.txt

friedel_c_Page_107.txt

friedel_c_Page_006.txt

friedel_c_Page_241.txt

friedel_c_Page_015.txt

friedel_c_Page_281.txt

friedel_c_Page_327.txt

friedel_c_Page_085.txt

friedel_c_Page_116.txt

friedel_c_Page_021.txt

friedel_c_Page_283.txt

friedel_c_Page_117.txt

friedel_c_Page_126.txt

friedel_c_Page_054.txt

friedel_c_Page_121.txt

friedel_c_Page_050.txt

friedel_c_Page_261.txt

friedel_c_Page_141.txt

friedel_c_Page_192.txt

friedel_c_Page_018.txt

friedel_c_Page_221.txt

friedel_c_Page_133.txt

friedel_c_Page_171.txt

friedel_c_Page_197.txt

friedel_c_Page_325.txt

friedel_c_Page_269.txt

friedel_c_Page_286.txt

friedel_c_Page_272.txt

friedel_c_Page_229.txt

friedel_c_Page_314.txt

friedel_c_Page_213.txt

friedel_c_Page_156.txt

friedel_c_Page_275.txt

friedel_c_Page_023.txt

friedel_c_Page_174.txt

friedel_c_Page_291.txt

friedel_c_Page_089.txt

friedel_c_Page_151.txt

friedel_c_Page_078.txt

friedel_c_Page_164.txt

friedel_c_Page_045.txt

friedel_c_Page_039.txt

friedel_c_Page_211.txt

friedel_c_Page_068.txt

friedel_c_Page_228.txt

friedel_c_Page_252.txt

friedel_c_Page_273.txt

friedel_c_Page_137.txt

friedel_c_Page_244.txt

friedel_c_Page_179.txt

friedel_c_Page_262.txt

friedel_c_Page_267.txt

friedel_c_Page_100.txt

friedel_c_Page_225.txt

friedel_c_Page_299.txt

friedel_c_Page_224.txt

friedel_c_Page_066.txt

friedel_c_Page_010.txt

friedel_c_Page_202.txt

friedel_c_Page_190.txt

friedel_c_Page_279.txt

friedel_c_Page_138.txt

friedel_c_Page_274.txt

friedel_c_Page_287.txt

friedel_c_Page_210.txt

friedel_c_Page_163.txt

friedel_c_Page_101.txt

friedel_c_Page_318.txt

friedel_c_Page_223.txt

friedel_c_Page_313.txt

friedel_c_Page_134.txt

friedel_c_Page_297.txt

friedel_c_Page_143.txt

friedel_c_Page_014.txt

friedel_c_Page_105.txt

friedel_c_Page_268.txt

friedel_c_Page_239.txt

friedel_c_Page_070.txt

friedel_c_Page_307.txt

friedel_c_Page_248.txt

friedel_c_Page_160.txt

friedel_c_Page_233.txt

friedel_c_Page_071.txt

friedel_c_Page_256.txt

friedel_c_Page_232.txt

friedel_c_Page_175.txt

friedel_c_Page_003.txt

friedel_c_Page_051.txt

friedel_c_Page_329.txt

friedel_c_Page_242.txt

friedel_c_Page_196.txt

friedel_c_Page_340.txt

friedel_c_Page_108.txt

friedel_c_Page_096.txt

friedel_c_Page_079.txt

friedel_c_Page_276.txt

friedel_c_Page_219.txt

friedel_c_Page_188.txt

friedel_c_Page_288.txt

friedel_c_Page_326.txt

friedel_c_Page_247.txt

friedel_c_Page_165.txt

friedel_c_Page_253.txt

friedel_c_Page_321.txt

friedel_c_Page_343.txt

friedel_c_Page_304.txt

friedel_c_Page_312.txt

friedel_c_Page_259.txt

friedel_c_Page_184.txt

friedel_c_Page_139.txt

friedel_c_Page_099.txt

friedel_c_Page_005.txt

friedel_c_Page_334.txt

friedel_c_Page_209.txt

friedel_c_Page_271.txt

friedel_c_Page_149.txt

friedel_c_Page_186.txt

friedel_c_Page_122.txt

friedel_c_Page_072.txt

friedel_c_Page_112.txt

friedel_c_Page_309.txt

friedel_c_Page_012.txt

friedel_c_Page_074.txt

friedel_c_Page_330.txt

friedel_c_Page_159.txt

friedel_c_Page_093.txt

friedel_c_Page_020.txt

friedel_c_Page_217.txt

friedel_c_Page_234.txt

friedel_c_Page_236.txt

friedel_c_Page_120.txt

friedel_c_Page_157.txt

friedel_c_Page_290.txt

friedel_c_Page_218.txt

friedel_c_Page_250.txt

friedel_c_Page_338.txt

friedel_c_Page_028.txt

friedel_c_Page_185.txt

friedel_c_Page_095.txt

friedel_c_Page_002.txt

friedel_c_Page_337.txt

friedel_c_Page_019.txt

friedel_c_Page_007.txt

friedel_c_Page_158.txt

friedel_c_Page_183.txt

friedel_c_Page_042.txt

friedel_c_Page_067.txt

friedel_c_Page_214.txt

friedel_c_Page_009.txt

friedel_c_Page_025.txt

friedel_c_Page_298.txt

friedel_c_Page_243.txt

friedel_c_Page_220.txt

friedel_c_Page_081.txt

friedel_c_Page_048.txt

friedel_c_Page_062.txt

friedel_c_Page_038.txt

friedel_c_Page_056.txt

friedel_c_Page_119.txt

friedel_c_Page_063.txt

friedel_c_Page_150.txt

friedel_c_Page_076.txt

friedel_c_Page_097.txt

friedel_c_Page_013.txt

friedel_c_Page_132.txt

friedel_c_Page_170.txt

friedel_c_Page_257.txt

friedel_c_Page_177.txt

friedel_c_Page_086.txt

friedel_c_Page_303.txt

friedel_c_Page_205.txt

friedel_c_Page_222.txt

friedel_c_Page_147.txt

friedel_c_Page_226.txt

friedel_c_Page_052.txt

friedel_c_Page_059.txt

friedel_c_Page_301.txt

friedel_c_Page_195.txt

friedel_c_Page_022.txt

friedel_c_Page_087.txt

friedel_c_Page_263.txt

friedel_c_Page_201.txt

friedel_c_Page_284.txt

friedel_c_Page_311.txt

friedel_c_Page_026.txt

friedel_c_Page_322.txt

friedel_c_Page_306.txt

friedel_c_Page_008.txt

friedel_c_Page_123.txt

friedel_c_Page_293.txt

friedel_c_Page_266.txt

friedel_c_Page_111.txt

friedel_c_Page_064.txt

friedel_c_Page_024.txt

friedel_c_Page_294.txt

friedel_c_Page_098.txt

friedel_c_Page_044.txt

friedel_c_Page_161.txt

friedel_c_Page_049.txt

friedel_c_Page_033.txt

friedel_c_Page_200.txt

friedel_c_Page_148.txt

friedel_c_Page_246.txt

friedel_c_Page_308.txt

friedel_c_Page_136.txt

friedel_c_Page_172.txt

friedel_c_Page_154.txt

friedel_c_Page_040.txt

friedel_c_Page_131.txt

friedel_c_Page_129.txt

friedel_c_Page_341.txt

friedel_c_Page_235.txt

friedel_c_Page_104.txt

friedel_c_Page_118.txt

friedel_c_Page_135.txt

friedel_c_Page_231.txt

friedel_c_Page_230.txt

friedel_c_Page_110.txt

friedel_c_Page_194.txt

friedel_c_Page_153.txt

friedel_c_Page_113.txt

friedel_c_Page_041.txt

friedel_c_Page_065.txt

friedel_c_Page_128.txt

friedel_c_Page_088.txt

friedel_c_Page_046.txt

friedel_c_Page_280.txt

friedel_c_Page_238.txt

friedel_c_Page_115.txt

friedel_c_Page_181.txt

friedel_c_Page_289.txt

friedel_c_Page_296.txt

friedel_c_pdf.txt

friedel_c_Page_027.txt

friedel_c_Page_302.txt

friedel_c_Page_130.txt

friedel_c_Page_167.txt

friedel_c_Page_260.txt

friedel_c_Page_331.txt

friedel_c_Page_060.txt

friedel_c_Page_146.txt

friedel_c_Page_315.txt

friedel_c_Page_082.txt

friedel_c_Page_199.txt

friedel_c_Page_305.txt

friedel_c_Page_109.txt

friedel_c_Page_237.txt

friedel_c_Page_216.txt

friedel_c_Page_155.txt

friedel_c_Page_204.txt

friedel_c_Page_178.txt

friedel_c_Page_166.txt

friedel_c_Page_212.txt

friedel_c_Page_270.txt

friedel_c_Page_240.txt

friedel_c_Page_193.txt

friedel_c_Page_017.txt

friedel_c_Page_323.txt

friedel_c_Page_047.txt

friedel_c_Page_152.txt

friedel_c_Page_090.txt

friedel_c_Page_180.txt

friedel_c_Page_168.txt

friedel_c_Page_176.txt

friedel_c_Page_206.txt

friedel_c_Page_187.txt

friedel_c_Page_080.txt

friedel_c_Page_124.txt

friedel_c_Page_333.txt

friedel_c_Page_265.txt

friedel_c_Page_207.txt

friedel_c_Page_169.txt

friedel_c_Page_069.txt

friedel_c_Page_061.txt

friedel_c_Page_254.txt

friedel_c_Page_037.txt

friedel_c_Page_036.txt

friedel_c_Page_264.txt

friedel_c_Page_075.txt

friedel_c_Page_285.txt

friedel_c_Page_114.txt

friedel_c_Page_073.txt

friedel_c_Page_227.txt

friedel_c_Page_189.txt

friedel_c_Page_245.txt

friedel_c_Page_053.txt

friedel_c_Page_292.txt

friedel_c_Page_029.txt

friedel_c_Page_043.txt

friedel_c_Page_255.txt

friedel_c_Page_316.txt

friedel_c_Page_295.txt

friedel_c_Page_057.txt

friedel_c_Page_336.txt

friedel_c_Page_125.txt

friedel_c_Page_258.txt

friedel_c_Page_198.txt

friedel_c_Page_328.txt

friedel_c_Page_320.txt

friedel_c_Page_278.txt

friedel_c_Page_251.txt

friedel_c_Page_030.txt

friedel_c_Page_127.txt

friedel_c_Page_282.txt

friedel_c_Page_031.txt

friedel_c_Page_215.txt

friedel_c_Page_344.txt

friedel_c_Page_249.txt

friedel_c_Page_035.txt

friedel_c_Page_310.txt

friedel_c_Page_142.txt

friedel_c_Page_317.txt

friedel_c_Page_145.txt

friedel_c_Page_106.txt

friedel_c_Page_102.txt

friedel_c_Page_016.txt

friedel_c_Page_103.txt

friedel_c_Page_339.txt

friedel_c_Page_032.txt

friedel_c_Page_011.txt

friedel_c_Page_140.txt

friedel_c_Page_055.txt

friedel_c_Page_191.txt

friedel_c_Page_001.txt

friedel_c_Page_034.txt

friedel_c_Page_300.txt

friedel_c_Page_083.txt

friedel_c_Page_077.txt

friedel_c_Page_091.txt

friedel_c_Page_319.txt

friedel_c_Page_084.txt

friedel_c_Page_144.txt

friedel_c_Page_162.txt


Full Text












DISSIMILAR COGNITIVE STYLES AND THEIR RELATIONSHIPS WITH
UNDERGRADUATE STUDENT STRESS, MOTIVATION, AND ENGAGEMENT















By

CURTIS ROBERT FRIEDEL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Curtis Robert Friedel

































This document is dedicated to my parents, Larry and Susan Friedel.
Thank you for everything.















ACKNOWLEDGMENTS

There are many people to thank who have pushed and pulled me to the creation of

this document. First, I would like to thank Dr. Rick Rudd for his advising,

encouragement, and friendship, all of which gave reason for me to start and complete my

Ph.D. My advisory committee members, Drs. Gary Fairchild, Howard Ladewig, and Ed

Osborne, all contributed in their own way by adding their expertise. This document is a

reflection of the scholarship and rigor imposed by these three gentlemen. Also, I would

like to thank Dr. Michael Kirton for his wisdom and counsel regarding cognitive style.

To say that my Ph.D. in agricultural education was earned in my three years at the

University of Florida would be an exaggeration. I owe many thanks to my undergraduate

advisor, Dr. Marvin Hoskey, a master teacher. Likewise, I appreciate the teachers and

administrators at Laurel-Concord Public Schools who served as models of excellent

teaching and learning early in my career.

I cannot say thank you enough to my parents for teaching me the value of hard

work and never letting me forget what is important in life. I also must thank my wife,

Beth, not only for patiently waiting for me to write this document, but for showing me

her approach towards life which helps me become better each and every day.

Last, but not least, a special thank you goes to my fellow graduate students with

whom I shared an office. The times we spent making trips for ice cream, discussing the

unspeakable over lunch, spending time at the Jones's on weekends, and building

camaraderie have all helped the time fly by. I have learned from all of them.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES .............. ................. ........... ................... ........ xi

LIST OF FIGURES ......... ....................... .......... ....... ............ xvi

ABSTRACT .............. ..................... .......... .............. xvii

CHAPTER

1 INTRODUCTION TO THE STUDY AND PROBLEM STATEMENT ....................1

B background of Study ............................................... ..... .................... .. 1
Introduction to Learning Styles .............................................................. ............... 2
Introduction to Kirton's Theory of Adaption-Innovation .........................................7
The Need for Student Engagem ent......... ............................................ ............. .. 9
Problem Statem ent ........ .... ....... ............ ........ .................... ......10
S ig n ifican ce ................................................................12
Purpose of the Study ....... ..... ................................ ... ............................. 13
A ssu m p tio n s ............................................................................................................... 14
L im station s .................................................................................................... 14
D efin itio n s ................................................................1 5
S u m m ary ............. ............. .................................................................18

2 REVIEW OF LITERATURE ................ ........ .... .........19

T h eoretical F ram ew ork ......................................................................................... 2 0
Learning and Learning Styles ............... ..... .....................29
Learning Style and Learning Process .............................................. ........33
Learning Style and Learning Level ........................................ ............... 35
Kirton's Cognitive Style...................................................................36
The Scale of Adaption and Innovation .....................................................37
Summary of Learning and Learning Styles.............................. ............ ...39
L earn in g an d S tre ss ............................................................................................... 4 0
Stress in the C classroom ....................................................... 4 1
Demographic Variables Related to Stress ........................................ ....43
M easurem ent of Stress ................................................. ........ 44
Summary of Learning and Stress ........................................... 45


v









L earning and M otivation .................................................. .............................. 46
Expectancy-V alue Theory ............................................................................ 47
The expectancy component........................ ........... ................... 48
The value com ponent ...........................................................................49
A effective m motivation ............................................. ............ ............... 51
Demographic Variables Related to Motivation......... ...... ........ ............... 52
Measurement of Motivation ................................................................ 52
Summary of Learning and Motivation ......................................................54
Learning and Engagem ent ............. .................................... ..................................... 55
The Influence of Faculty Members ........................................ ............... 56
Demographic variables related to engagement.................................................58
Measurement of Engagement ... .. ..........................................58
Summary of Learning and Engagement................................. ......................61
Su m m ary ...................................... .................................................. 62

3 M E T H O D S ........................................................................................................... 6 4

R e se arch D e sig n ................................................................................................... 6 5
P o p u la tio n ............................................................................................................. 6 7
Instrumentation .............................................. .. .......... ............... ..69
Kirton Adaption-Innovation Inventory.............................. .................. 69
Student-life Stress Inventory .............. .......................... ........................... 70
Motivated Strategies for Learning Questionnaire ............................................72
National Survey of Student Engagement .................................... ...............73
D ata C collection Procedures ............................................... ............................. 75
Data Analysis Procedures ........... ........ ................................ 79
R research O bjectiv e O ne ........................................................... .....................79
R research O objective Tw o........................................................... ............... 80
R research O objective Three......................................................... ............... 80
R research Objective Four ................. ........ .............. ............................80
R research O objective Five...................... .... ................ ................. ............... 81
S u m m ary ...................... .. .. ......... .. .. ................................................. 8 1

4 R E S U L T S .............................................................................8 3

Instrument Post-hoc Reliability ................................. ........................... 84
O bje ctiv e O n e ...................... .. ............. .. ................................................8 5
C la s s A ........................................................................................................... 8 7
C la s s B ........................................................................................................... 8 8
C la s s C ........................................................................................................... 9 0
C la s s D ........................................................................................................... 9 1
C la s s E ........................................................................................................... 9 3
C la s s F ........................................................................................................... 9 4
C la s s G ........................................................................................................... 9 6
C la s s H ........................................................................................................... 9 7
C la s s I ...................... .. ............. .. .....................................................9 9
All Students ..................................... ............ ... .........101









Summary of Findings for Objective One .............. ...........................................103
Objective Tw o .................................... .. .. ........... ........ ..... 105
C la ss A ................................................................. 1 0 5
C la s s B ..................................................................................................... 1 0 9
C la s s C ..................................................................................................... 1 1 2
Class D .................. .......... ............... 116
Class E ...................................... ........................... 119
C la ss F ...............................................................12 3
C la ss G ............................................................12 7
C lass H ................................................................................................ 13 0
C la ss I ................................................................................................ 13 4
All Students .......................................................................... ........ ...................137
Com prisons Betw een Classes .............. ................................................. 141
Summary of Findings for Objective Two .............. ............. ...............143
O bje ctiv e T h re e ................................................................................................... 14 7
C la ss A ............................................................14 8
Class B .......... ............................ ............... 155
C la ss C ............................................................1 6 1
Class D ................. .......... .................... 167
Class E ...................................... ........................... 173
C la ss F ...............................................................1 7 8
C la ss G ..............................................................1 8 4
C lass H ................................................................................................ 18 9
C la ss I ................................................................................................ 19 5
A ll S tu d en ts ...............................................................2 0 0
Group Comparisons.................. ............................206
Summary of Findings for Objective Three ................................................ ......... 208
Objective Four ..... ....................... .................. 214
C la ss A ............................................................ 2 1 6
C la ss B .............................................................. 2 1 8
C la ss C ............................................................ 2 1 8
C la ss D ............................................................2 2 0
C la ss E ..............................................................2 2 1
C la ss F ...............................................................2 2 3
C la ss G ............................................................2 2 4
C la ss H ............................................................2 2 4
Class I ................. .......... .................... 225
A ll Students ........................................227
Summary of Findings for Objective Four ...... ................. ...............229
O objective Five .............. ........ ...................................................................... 231
C la ss A ............................................................2 3 2
C la ss B ............................................................2 3 3
C la ss C ............................................................2 3 4
C la ss D ..............................................................2 3 5
C la ss E ..............................................................2 3 6
C la ss F ...............................................................2 3 7










C la ss G ...............................................................2 3 7
C la ss H ...............................................................2 3 8
Class I ................................................. .................... 239
A ll Stu dents ....................................................... 240
Summary of Findings for Objective Five ............ ......... ................... 242
S u m m a ry ................................ ............................................................................. 2 4 3

5 SUMMARY AND DISCUSSION ........................................ ... ........ 245

Summary of Problem Statement ................. .......................245
Purpose and O objectives ........................................... ............................................246
M eth o d o lo g y ................................ ....................................................................... 2 4 7
Conclusions and D discussion ............................................... ........ 249
Objective One ............... ......... ............... 250
C la ss A ...............................................................2 5 0
C lass B ................................................................. ......251
C lass C ................................................................. ......251
C la ss D ...............................................................2 5 2
C la s s E ................................................................................................... 2 5 2
C la ss F ........................................................................2 5 3
C la ss G ...............................................................2 5 3
C la ss H ...............................................................2 5 4
C la ss I ............................................................2 5 4
All Participants ................................................ 255
Discussion for objective one ..............................................255
O objective T w o .............. .................. ......... .......................................... .. .257
C la ss A ...............................................................2 5 7
C la s s B ................................................................................................... 2 5 8
C la s s C ................................................................................................... 2 5 8
C la ss D ...............................................................2 5 8
C la s s E ................................................................................................... 2 5 9
C la ss F ........................................................................2 5 9
C la ss G ...............................................................2 6 0
C la ss H ...............................................................2 6 0
C la ss I ............................................................2 6 1
All students........................................... 261
Comparison between classes ............... .. ............. ..... ................ 262
Discussion for objective two ..............................................263
Objective Three ......................................................................267
C la ss A ...............................................................2 6 8
C la s s B ................................................................................................... 2 6 9
C la s s C ................................................................................................... 2 6 9
C la ss D ...............................................................2 7 0
C la s s E ................................................................................................... 2 7 1
C la ss F ........................................................................2 7 1
C la ss G ...............................................................2 7 2
C la ss H ...............................................................2 7 2


v111









C la ss I ............................................................2 7 3
All Students ..... ......... .. ........ .......... ........274
G roup C om parisons........................................................... ............... 274
Discussion for objective three..... .................... .............276
Objective Four ................. ........ .................... 280
C la ss A ...............................................................2 8 0
C la s s B ................................................................................................... 2 8 1
C la s s C ................................................................................................... 2 8 1
C la ss D ...............................................................2 8 2
C la s s E ................................................................................................... 2 8 2
C la s s F ................................................................................................... 2 8 2
C la ss G ...............................................................2 8 3
C la ss H ...............................................................2 8 3
C la ss I ............................................................2 8 4
All Students ..... ......... .. .......... ..........284
Discussion for objective four ............................................................. 285
Objective Five ................. ......... ..................287
C la ss A ...............................................................2 8 8
C la s s B ................................................................................................... 2 8 8
C la s s C ................................................................................................... 2 8 9
C la ss D ...............................................................2 8 9
C la s s E ................................................................................................... 2 8 9
C la s s F ................................................................................................... 2 9 0
C la ss G ...............................................................2 9 0
C la ss H ...............................................................2 9 0
C la ss I ............................................................2 9 1
All Students ..... ............ ............ ..........291
Discussion for objective five ...... ...................... ........ 291
Concluding Discussion ............. .. ......... .................293
Recommendations.......................................................298
S u m m a ry .............. ..... ............ ................. .................................................. 3 0 0

APPENDIX

A INTERNAL REVIEW BOARD SUBMISSION FORM ........................................301

B FACULTY INFORMED CONSENT ............................................... 304

C STUDENT INFORMED CONSENT ..................................... ............................. 306

D EXAMPLE ITEMS OF KIRTON' S ADAPTION-INNOVATION
IN V E N T O R Y .................................................................................................. 3 0 8

E EXAMPLE ITEMS OF STUDENT-LIFE STRESS INVENTORY............309

F EXAMPLE ITEMS OF MOTIVATED STRATEGIES FOR LEARNING
Q U E S T IO N N A IR E .................................................................................................. 3 10










G EXAMPLE ITEMS OF NATIONAL SURVEY OF STUDENT
E N G A G E M E N T ............................................................................ .. .............. 11

L IST O F R EFE R E N C E S ........................................................................... ..............312

B IO G R A PH IC A L SK E T C H ........................................... ...........................................325



















































x
















LIST OF TABLES


Table page

2-1 Factor Traits of Adaptors and Innovators ..................................... .................36

3-1 Construct Reliability of Kirton Adaption-Innovation Inventory.............................70

3-2 Construct Reliability of Student-life Stress Inventory ..........................................71

3-3 Construct Reliability of Motivated Strategies for Learning Questionnaire ............73

3-4 Construct Reliability of National Survey of Student Engagement ........................75

4-1 Listing of Classes designated by Faculty Members' Total Cognitive Style Score
(N = 9 ) ...............................................................................8 6

4-2 Summary of Classes Regarding Reported Demographic Information of
Participating Faculty M em bers ........................................ ........................ 103

4-3 Summary of Classes Regarding Reported Demographic Information of
P participating Students.................................................. ............................... 104

4-4 Class A Student Mean Scores of Cognitive Style Constructs (n=58)....................106

4-5 Class A Student Mean Scores of Stress Constructs ............................................107

4-6 Class A Student Mean Scores of Motivation Constructs................................ 108

4-7 Class A Student Mean Scores of Engagement Constructs (n=70).........................109

4-8 Class B Student Mean Scores of Cognitive Style Constructs (n=46) .................. 109

4-9 Class B Student Mean Scores of Stress Constructs (n=68)..................................110

4-10 Class B Student Mean Scores of Motivation Constructs ................ ............... 111

4-11 Class B Student Mean Scores of Engagement Constructs (n=70) ......................... 112

4-12 Class C Student Mean Scores of Cognitive Style Constructs (n=56)....................113

4-13 Class C Student Mean Scores of Stress Constructs ......................... ............114









4-14 Class C Student Mean Scores of Motivation Constructs (n=71) ...........................115

4-15 Class C Student Mean Scores of Engagement Constructs (n=71) .........................115

4-16 Class D Student Mean Scores of Cognitive Style Constructs (n=73)....................116

4-17 Class D Student Mean Scores of Stress Constructs ............... ..........117

4-18 Class D Student Mean Scores of Motivation Constructs ............. ............... 118

4-19 Class D Student Mean Scores of Engagement Constructs (n= 108)....................19

4-20 Class E Student Mean Scores of Cognitive Style Constructs (n=32) ..................120

4-21 Class E Student Mean Scores of Stress Constructs................... ... .............121

4-22 Class E Student Mean Scores of Motivation Constructs (n=48)..........................122

4-23 Class E Student Mean Scores of Engagement Constructs (n=48) .......................123

4-24 Class F Student Mean Scores of Cognitive Style Constructs (n=71)................... 124

4-25 Class F Student Mean Scores of Stress Constructs (n= 12)..............................125

4-26 Class F Student Mean Scores of Motivation Constructs (n= 115).......................... 126

4-27 Class F Student Mean Scores of Engagement Constructs (n= 115)........................127

4-28 Class G Student Mean Scores of Cognitive Style Constructs (n=65)....................127

4-29 Class G Student Mean Scores of Stress Constructs (n=84) .................................128

4-30 Class G Student Mean Scores of Motivation Constructs (n=85) .........................129

4-31 Class G Student Mean Scores of Engagement Constructs (n=85).........................130

4-32 Class H Student Mean Scores of Cognitive Style Constructs (n=50)....................131

4-33 Class H Student Mean Scores of Stress Constructs (n=70) .................................132

4-34 Class H Student Mean Scores of Motivation Constructs (n=70) .........................132

4-35 Class H Student Mean Scores of Engagement Constructs................................... 133

4-36 Class I Student Mean Scores of Cognitive Style Constructs (n=60)....................134

4-37 Class I Student Mean Scores of Stress Constructs (n=77)...................................135

4-38 Class I Student Mean Scores of Motivation Constructs (n=77).............................136









4-39 Class I Student Mean Scores of Engagement Constructs .....................................137

4-40 All Student Mean Scores of Cognitive Style Constructs (n=511) .......................138

4-41 All Student M ean Scores of Stress Constructs...................................................... 139

4-42 All Student M ean Scores of M otivation Constructs..............................................140

4-43 All Student Mean Scores of Engagement Constructs .........................................140

4-44 Summary of Classes Regarding Student Cognitive Style, Stress, Motivation and
Engagem ent M ean Scores ......................................................... .............. 147

4-45 Class A Student Mean Scores of Cognitive Style Construct Gaps (n=58) ............149

4-46 Class A Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=56) ......................153

4-47 Class B Student Mean Scores of Cognitive Style Construct Gaps (n=46)............155

4-48 Class B Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=44) ......................160

4-49 Class C Student Mean Scores of Cognitive Style Construct Gaps (n=56).............162

4-50 Class C Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=56) ......................166

4-51 Class D Student Mean Scores of Cognitive Style Construct Gaps (n=73) ............168

4-52 Class D Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=67) ......................172

4-53 Class E Student Mean Scores of Cognitive Style Construct Gaps (n=32).............174

4-54 Class E Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=31)......................177

4-55 Class F Student Mean Scores of Cognitive Style Construct Gaps (n=71)............ 179

4-56 Class F Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=69) ......................182

4-57 Class G Student Mean Scores of Cognitive Style Construct Gaps (n=65) ............184

4-58 Class G Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=64) ........................188

4-59 Class H Student Mean Scores of Cognitive Style Construct Gaps (n=50) ...........190









4-60 Class H Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=50) ......................194

4-61 Class I Student Mean Scores of Cognitive Style Construct Gaps (n=60)............196

4-62 Class I Correlations, Means, and Standard Deviations among Constructs of
Cognitive Style Gap, Stress, Motivation and Engagement (n=59) ......................199

4-63 All Student Class Mean Scores of Cognitive Style Construct Gaps (n=511) ........201

4-64 All Student Class Correlations, Means, and Standard Deviations among
Constructs of Cognitive Style Gap, Stress, Motivation and Engagement
(n=496) ............................................................... ..... ..... ......... 205

4-65 Class A Backward Stepwise Multiple Regression Explaining Student Total
S tre ss (n = 5 7 ) ..................................................................... 2 16

4-66 Class A Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=57) ...................... .... ............ ................. .... ....... 217

4-67 Class B Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=45) ...................... .... ............ ................. ..........218

4-68 Class C Backward Stepwise Multiple Regression Explaining Student Total
S tre ss (n = 5 6 ) ..................................................................... 2 19

4-69 Class C Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=56) ...................... .... ............ ................. .... ....... 220

4-70 Class D Backward Stepwise Multiple Regression Explaining Student Total
S tre ss (n = 6 6 ) ..................................................................... 2 2 1

4-71 Class E Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=32) ...................... .................. ................. .... ....... 222

4-72 Class F Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=71) .................. ..................................... .. ........ .... 223

4-73 Class H Backward Stepwise Multiple Regression Explaining Student Total
S tre ss (n = 4 9 ) ..................................................................... 2 2 5

4-74 Class I Backward Stepwise Multiple Regression Explaining Student Total Stress
(n=60) ............................................................... ..... ...... ........ 226

4-75 Class I Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=60) ...................... .... ............ ................. .... ....... 227

4-76 All Students Backward Stepwise Multiple Regression Explaining Student Total
S tre ss (n = 4 9 6 ) .................................................................... 2 2 8









4-77 All Students Backward Stepwise Multiple Regression Explaining Student Total
M motivation (n=502) ........................................ ............... .... ....... 229

4-78 Class A Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=56) .................. ..................................... .. ........ .... 233

4-79 Class B Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=45) .................. ........................... ...................... 234

4-80 Class C Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=56) ............ ..... ...... ....... .......... .. .... .............. 235

4-81 Class D Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=66) .................. ........................... ...................... 236

4-82 Class E Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=3 1) .................. ............................. ........ .. ........ .... 237

4-83 Class G Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=63) .................. ..................................... .. ........ .... 238

4-84 Class H Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=49) .................. ........................... ...................... 239

4-85 Class I Backward Stepwise Multiple Regression Explaining Student Total
Engagem ent (n=59) .................. ..................................... .. ........ .... 240

4-86 All Student Backward Stepwise Multiple Regression Explaining Student Total
E ngagem ent (n= 493) ................................................................... .....................24 1
















LIST OF FIGURES


Figure page

2-1 Kirton's theoretical model of Adaption and Innovation applied to the
undergraduate classroom ........................................ ......... ..... ................. 24

2-2 Interaction of cognitive style gap, stress, motivation and engagement....................27















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DISSIMILAR COGNITVE STYLES AND THEIR RELATIONSHIPS WITH
UNDERGRADUATE STUDENT STRESS, MOTIVATION, AND ENGAGEMENT

By

Curtis R. Friedel

December, 2006

Chair: Rick Rudd
Major Department: Agricultural Education and Communication

Few studies have examined dissimilar cognitive style between faculty members

and students and its relationships with student stress, motivation and engagement. In the

search for variables explaining student engagement, researchers have examined the

cognitive level of instructional discourse, but few have examined the cognitive style of

the instructor as it corresponds to the cognitive style of the student. Identifying how

dissimilar learning styles relate to student stress and motivation may have implications

for increasing student engagement.

Kirton's Adaption-Innovation theory provided the theoretical framework for this

study. An ex post facto design was used to accomplish the objectives of this study. Nine

undergraduate courses in the College of Agriculture and Life Sciences at the University

of Florida were administered four standardized instruments: Kirton's Adaption-

Innovation Inventory, Student-life Stress Inventory, Motivated Strategies for Learning

Questionnaire, and National Survey of Student Engagement. Dissimilar cognitive style









was measured by subtracting the faculty members' cognitive style score from each

individual student's cognitive style score. The difference provided a cognitive style gap

which was used to determine relationships among student stress, motivation and

engagement in the classroom.

Findings indicated that in these courses, students with more than a 20-point

cognitive style gap did not have higher levels of stress than students with less than a 20-

point cognitive style gap. However, students with more than a 20-point cognitive style

gap did have significantly lower levels of motivation. Student motivation was further

decreased when students were taught by an adaptive faculty member. Backwards

stepwise multiple regression was used to explain student stress, student motivation and

student engagement based on cognitive style gap and selected demographic variables.

Results provided evidence that cognitive style gap did contribute to the explanation of

student stress, motivation and engagement in some classes. However, cognitive style gap

was not significant in explaining these dependent variables in all nine classes.

Further research should employ qualitative methods to examine how cognitive

style is exhibited in instructional discourse. Researchers may use these findings to further

examine the relationships between learning style and learning process in the effort of

improving student engagement.


xviii














CHAPTER 1
INTRODUCTION TO THE STUDY AND PROBLEM STATEMENT

Background of Study

This study examined relationships between dissimilar cognitive styles (Kirton,

2003), stress, motivation and classroom engagement at the University of Florida.

Educators have long known through pragmatic and empirical evidence that students learn

in a variety of styles. Many assessments of learning styles and personality have been

developed to determine how students interact with and process the learning environment.

All learning style theories share in common the contention that learning improves in

learning environments that match a student's learning style (Coffield, Moseley, Hall and

Ecclestone, 2004). Nevertheless, few studies have been conducted to assess the

interaction between the learning style of the student and the learning style of the

instructor; nor the positive and negative effects from this interaction. Cognitive style as

defined by Kirton may offer insight into how a style of learning influences other variables

in the college classroom. Using the Kirton's Adaption-Innovation Inventory (KAI), the

researcher determined the cognitive style of undergraduate students and university

instructors in the college classroom. The interval difference between an individual

student's cognitive style and the instructor's cognitive style yielded a cognitive style gap

score (Kirton). Individual cognitive style gap scores were compared to other student

variables including student stress, motivation, engagement, and selected demographics.









Introduction to Learning Styles

A current trend in undergraduate education is a transition from teacher-centered

classrooms to student-centered classrooms (Acharya, 2002). Although teachers are still

accountable for student learning, great interest has been given in support of self-regulated

learning (Winne, 1995) and student engagement (Kuh, 2003). Boekaerts (1999) stated

that student understanding of learning styles is critical to increasing successful learning.

But, what is a learning style? To define learning style, the researcher reviewed an

array of definitions, theories, measurements and beliefs which included those written by

Allinson and Hayes (1996), Apter (2001), Dunn (1984), Gregorc (1979), James and

Gardner (1995), Keefe (1979), Kirton (1976), Kolb (1984), Sternberg (1988) and Witkin

and Goodenough (1977). Throughout the literature the terms learning style and cognitive

style have been used interchangeably with few authors making distinctions between the

two concepts (Sadler-Smith, 2001). Additionally, authors have been congruent in

defining both learning style and cognitive style. For example, Claxton and Ralston (1978,

p. 7) define learning style as a "consistent way of responding to and using stimuli in the

context of learning." Messick (1984, p. 143) defines cognitive style as a "consistent

individual differences in ways of organizing and processing information and experience."

Both definitions were deemed appropriate for use in this study with the two terms used

synonymously throughout this document.

In a literature review of learning styles, Coffield, Moseley, Hall and Ecclestone

(2004) reported 71 different assessments of learning style. To categorize the various

learning style theories, a continuum of learning style families was developed based on the

degree of the learning style being fixed and innate. Anchoring one end of the continuum

were strongly fixed and innate learning styles possibly stemming from genetics. The









other end of the continuum is anchored by learning styles which are influenced by

situational factors such as the environment, motivation, stress and learning strategies.

From one through five, this continuum of learning style families and their respective

theorists included the following:

1. Learning styles and preferences are largely constitutionally based including the four
modalities: visual, auditory, kinesthetic, and tactile. Learning style theorists in this
family include Bartlett, Betts, Dunn and Dunn, Gordon, Gregorc, Marks, Paivio,
Richardson, Sheehan and Torrance.

2. Learning styles reflect deep-seated features of the cognitive structure, including
patterns of ability. Learning style theorists in this family include Broverman,
Cooper, Gardner et al., Guilford, Holzman and Klein Hudson, Hunt, Kagan, Kogan,
Messick, Pettigrew, Riding, and Witkin.

3. Learning styles are one component of a relatively stable personality type. Learning
style theorists in this family include Apter, Epstein and Meier, Harrison-Branson,
Jackson, Miller, and Myers-Briggs.

4. Learning styles are stable learning preferences with a degree of flexibility. Learning
style theorists in this family include Allinson and Hayes; Felder and Silverman;
Herrmann; Hermanussen, Wierstra, de Jong, and Thij ssen; Honey and Mumford;
Kaufmann; Kirton; Kolb; and McCarthy.

5. Learning styles are approaches, strategies, orientations and conceptions of learning.
Learning style theorists in this family include Biggs; Conti and Kolody; Entwistle;
Grasha-Riechmann; Hill; Marton and Saljo; McKenney and Keen; Pask; Pintrich,
Smith, Garcia and McCeachie; Schmeck; Sternberg; Vermunt; Weinstein,
Zimmerman and Palmer; and Whetton and Cameron.

(Coffield, Moseley, Hall & Ecclestone, 2004, p. 9)

This simple categorization of learning styles ignores the depth of theory

supporting learning, validity and reliability of the instrumentation, and classroom

implications; however, the work done by Coffield, Moseley, Hall and Ecclestone gives

the best reference point for the discussion of strengths, weaknesses, and similarities of


learning styles.









Even within categories of learning styles there are disagreements. For example in

category four, Kolb (1984) defines his measure of learning style as stable cognitive

programs formed by past experiences of education, choice of career, and the current task

at hand. Kirton (2003) asserts his measure of learning style is based on fixed and innate

traits used in problem solving, but people may cope outside of their style. Kolb's (1984)

model of experiential learning requires four learning styles to complete the experiential

learning process. Kirton (2003) claims that the process of learning is independent of

learning style (de Ciantis & Kirton, 1996). Additionally, Kolb (1984) claims that a

preferred learning style will allow a person to be more competent in completing a

particular task. Kirton (2003) argues that everyone has the capability to solve all

problems within their own style. That is, learning style is distinct and independent of

learning level or capacity. Level of learning can be defined as an individual's ability to

learn and can be expressed as intelligence (Kirton, 2003). The literature is unclear in

distinguishing an individual's style of learning and level of learning, which is an assertion

of Kirton's theory. This assertion is supported by the literature as Ericsson, Krampe and

Tesch-Romer (1993) found that in all domains of knowledge, the single greatest

contribution to expertise was deliberate practice. In addition, no evidence was found that

a particular learning style or individual difference played a role in acquiring expertise

within a domain of knowledge. However, it may be possible that experts have a preferred

style when producing an expert performance.

Learning styles have been included as a variable in many educational studies.

Still, research conducted in the area of learning styles has provided conflictive findings.

For example, Smith, Sekar and Townsend (2002) reviewed studies in which students









were given instructional discourse that matched their individual learning style. Results

were mixed with only nine of eighteen studies supporting the contention that students

learn best when taught in their preferred learning style. In another example, Wingenbach

(2000) used the Group Embedded Figures Test (GEFT) (Witkin, Oltman, Raskin & Karp,

1971) and found undergraduates with field-independent learning styles had a moderately

positive and significant relationship with student academic achievement. However,

DeTure (2004) used the GEFT and found no evidence to significantly predict student

academic achievement from learning style. Why do these findings conflict? If Kirton

(2003) is correct, all students are capable of achieving academic success within their own

cognitive style. If academic achievement is dependent on the expectations of the faculty

member, could dissimilar cognitive style between students and the faculty member be an

intervening variable?

Kirton's Adaption-Innovation (A-I) Theory (2003) identifies cognitive style as a

preference to the amount of structure that one undertakes when solving problems. Within

the continuum of learning style families (Coffield et al., 2004), Kirton's work resides in

the fourth family noted above, "flexibly stable learning preferences" (p.9). Some people

are more adaptive when solving problems, and other people are more innovative when

solving problems. Adaptiveness and innovativeness anchor on continuums of how people

approach problems, produce ideas, and adhere to structures. These three constructs

measure a fixed perception of an environment in which the individual interacts with and

responds in the process of learning while solving a problem; a fit to this study's definition

of learning style (Claxton and Ralston, 1978). More detail will be given to Kirton's

cognitive style's relationship to learning in chapter 2.









Instructional material designed to incorporate learning styles may increase

students' motivation to learn, (Larkin-Hein, 2000); however, more research is warranted

in this area. Few studies have been conducted to measure the impact of curriculum

delivered in a learning style different than the student. Nor have studies been conducted

to examine how instructional discourse of a dissimilar learning style relates to a student's

stress and motivation in the classroom; an assertion of Kirton's A-I theory.

The primary focus of research and practice of learning style theory is aligning

curriculum and instruction to meet the needs of diverse learning styles in a given

classroom (Curry, 1990). However, much of the literature fails to recognize the learning

style of the instructor and how this variable plays in the context of the classroom. The

effect of dissimilar learning styles of people working together to develop a new idea or

product and the stress resulting from this interaction is lacking in the literature (Smulders,

2004); yet the study of this phenomenon is the aim of A-I theory (Kirton, 2003). Since

students and teachers of different learning styles commonly come together in a classroom

to generate ideas with varying degrees of motivations, attitudes and stressors, it is clear

that many research questions are still unanswered.

Limitations have been placed on the application of learning styles in the

classroom. After an extensive review of the literature, Coffield, Moseley, Hall and

Ecclestone (2004) cautioned practitioners' application of learning styles in the classroom

as many types of style have little theoretical backing, low reliability and insufficient

validity. Furthermore, few studies have found learning styles to "explain more than 16%

of the variance in academic achievement" (p. 127). Other authors have agreed that many

measures of learning style are psychometrically weak and therefore offer little use to the









practitioner (Curry, 1990; Hargreaves, 2005; Price, 2004; Smith 2002,). It is evident that

future research must use psychometrically solid measures of learning style with strong

theoretical ties in order to provide recommendations to education practitioners.

Introduction to Kirton's Theory of Adaption-Innovation

In Kirton's (2003) A-I theory, a person's cognitive style can be determined as

more adaptive versus more innovative along a continuum. Cognitive style is ultimately a

measure of how individuals solve problems with indication of preferences for learning

(de Ciantis & Kirton, 1996). Kirton (2003) describes cognitive style in context as

individuals interacting and reacting to the environment in consistent and stable

approaches when solving problems, which is consistent with definitions of learning style

and cognitive style used in this study (Claxton & Ralston, 1978; Messick, 1984). A

person may find their preferred style of problem solving along the adaptive-innovative

continuum by completing the Kirton Adaption-Innovation Inventory (KAI).

Kirton's (2003) measure of cognitive style along a continuum allows for interval

scale distinction between dissimilar cognitive styles of two individuals. Additionally,

dissimilarity of cognitive style can be determined between two innovative individuals or

two adaptive individuals. The calculated difference between two individuals' cognitive

style scores yields a cognitive style gap (Kirton) which determines the dissimilarity of

two approaches to problem solving. Cognitive style gap was calculated in this study by

subtracting faculty members' cognitive style score from each student's cognitive style

score.

A cognitive style gap larger than 20 points may contribute to inefficiencies in

communication (Kirton, 2003). Stress is a result of cognitive gap (Kirton) which may

present important implications to learning. The literature suggests that stress in the form









of anxiety may affect students' self-esteem and academic achievement (Pintrich &

Schunk, 2002). Fortunately, individuals can be motivated to operate outside of a

preferred cognitive style and bridge the cognitive gap found with another individual

(Kirton, 2003). Kirton uses the term coping behavior to describe bridging a cognitive

style gap and further elaborates that motivation initiates coping behavior.

It is necessary to note that Kirton's measure of cognitive style is distinct from

motivation. However, authors of other learning style instruments integrate aspects of

motivation and cognitive style into one measure of how individuals approach learning

(Apter, 2001; Dunn, 1984; Entwistle, 1978; Herrmann, 1989; Vermunt, 1992). The note

was not made to argue for or against motivational factors as a determinant of a preferred

style of learning. Rather the pure and distinct measure of learning style separated from

motivation allows for a more precise measure of motivation (Kirton, 2003), which in turn

permits the study of motivation as it relates to dissimilar learning styles and to learning.

The literature provides evidence that student motivation is necessary for learning

(Schunk, 1991) and student engagement (McKeachie, Pintrich, Lin & Smith, 1986)

making Kirton's measure of cognitive style particularly valuable.

Kirton's A-I theory provides a theoretical framework to incorporate the

interaction of dissimilar learning styles in the production of new products and ideas.

Adaption-Innovation theory was chosen for this study, because of its ability to separate

cognitive style from cognitive level while giving attention to motivation, stress, and

learning. Distinction between cognitive style and cognitive level may provide insight to

the literature which asserts a confused meshing of the two in explaining student

achievement. A reference point is also established to examine the relationships of









variables found in the educational classroom such as student engagement, motivation and

stress. The KAI has been determined as psychometrically reliable and valid (Kirton,

1999); as opposed to many other learning style assessments (Coffield et al., 2004; Curry,

1990). Given the above discussion of A-I theory, the researcher used Kirton (2003) to

operationalize learning style. Chapter 2 further elaborates on Kirton's A-I theory and

measure of cognitive style.

The Need for Student Engagement

Involvement in Learning (Study Group on the Conditions of Excellence in

American Higher Education, 1984), a report proposing the reform of higher education,

stated a quality education was more dependent on student engagement and less on the

reputation and resources of the academic institution. Suddenly faculty members of

colleges and universities were asked to be accountable for undergraduate student

achievement. To assist faculty members with their new challenge, Chickering and

Gameson (1987) developed a set of engagement indicators known as the Seven Principles

of Good Practice in Undergraduate Education. They include:

1. Encourages contact between students and faculty,
2. Develops reciprocity and cooperation among students,
3. Encourages active learning,
4. Gives prompt feedback,
5. Emphasizes time on task,
6. Communicates high expectations, and
7. Respects diverse talents and ways of learning. (p. 5)

The Seven Principles of Good Practice in Undergraduate Education are widely

accepted by the American Association of Higher Education (AAHE) as research based

process standards that can be applied and measured in undergraduate classrooms.

In a longitudinal, multi-institutional study, Astin (1993) used 82 student outcome

measures, 140 student characteristics, 190 environmental characteristics and 57 types of









student involvement and found that academic involvement, faculty involvement and

student peer group involvement were the largest indicators of enhancing student learning

and personal development of undergraduate students. Koljatic and Kuh (2001) conducted

a study to determine if cooperation among students, active learning, and faculty contact

with students has improved since 1983 and found the degree of student engagement in

these three areas has not improved. Furthermore, 20% of freshman and seniors in

colleges and universities across the nation were identified as disengaged (Kuh, 2003).

Even though the literature has addressed the need for student engagement, it seems that

faculty members' instructional techniques still do not encourage undergraduate students'

classroom engagement. In colleges of agriculture, Whittington (1998) found almost 80%

of faculty members' instructional discourse to be low at cognitive level. Is student

engagement explained by cognitive level, cognitive style, or both? Does cognitive style

gap between faculty members and students matter?

Problem Statement

Trends in undergraduate education are moving from teacher-centered classrooms

to student-centered classrooms for the improvement of student engagement (Acharya,

2002). From classroom teaching experiences, most faculty members concur that students

learn differently. This has increased the application of learning styles in the classroom as

faculty members struggle to develop instructional techniques to engage student learning.

Advocating the awareness of differing learning styles is commendable. However, further

implications for education practitioners are limited as studies examining instructional

discourse provided to students in similar learning styles have conflicting results (Smith,

Sekar & Townsend, 2002). Many learning styles measures are psychometrically weak

(Curry, 1990) and offer small explanation to the variance of student achievement









(Coffield et al., 2004). Simply stated, future research must be able to accurately

determine students' learning styles and provide theoretical and empirical evidence that

learning styles contribute to student learning.

The literature comprising the investigation of learning styles is vast with little

agreement or collaboration to find agreement among the respective theorists (Coffield et

al.). Little attention has been given to how dissimilar learning styles between students and

faculty members contribute to any measure of academic success (Coffield et al.).

Furthermore, the literature has been consistently unclear concerning the separation of

learning style from learning capability (Kirton, 2003). Research conducted to explain

student achievement by measures of inherent learning styles without controlling for other

variables, such as motivation, stress and intelligence, tend to support beliefs of academic

superiority (Hargreaves, 2005). However, if students learn best in their preferred learning

style and instructional discourse is only provided in that learning style, academic

superiority may be an outcome. Investigating how dissimilar learning styles contribute to

academic success is necessary. Also, the separation of learning style from learning level

is warranted to allow for the distinct relationship and accurate measure of other

classroom variables that facilitate student engagement in the classroom and promote

academic achievement. That is, if dissimilar learning style is not found to be a large

contributor to student engagement, perhaps learning level is a superseding variable.

Faculties of colleges of agriculture and life sciences have been reliant on scholars

in agricultural education to examine teaching and learning models that promote student

engagement. Nevertheless, studies conducted in agricultural education have made few

attempts to identify relationships among student characteristics, thinking skills and









motivations that increase student engagement (Rudd, Baker & Hoover, 2000). Kirton's

A-I theory provides a framework to examine how motivation may overcome the strain of

students and faculty members with dissimilar cognitive styles. Adaption-Innovation

theory has been applied as a model to problem solving in business management,

consumer marketing and aspects of change; however, little research has been done to

apply A-I theory to the context of undergraduate education. As A-I theory is not context

specific (Kirton, 1994), the theory should also apply in the undergraduate classroom.

Faculty members are increasingly faced with more accountability for increasing

student achievement. As student engagement is closely related to academic achievement

(Kuh, 2001), faculty members need to utilize instructional strategies and curricula to

increase student engagement to foster student learning. Research is needed to determine

if cognitive style contributes to the explanation of student engagement. Does the

cognitive style gap between students' preferred cognitive style and faculties' preferred

cognitive style impair student engagement? Are students able to overcome cognitive style

gap as the faculty member engages the learner? What are the implications of Kirton's A-I

theory as applied within the context of the undergraduate classroom?

Significance

The goal of many university faculties is to help students learn the content of the

course in order to prepare them for graduate school, professional school or a successful

career. The goals of many university students mirror the faculty members' goals.

Increasing student engagement in the classroom will facilitate the successful achievement

of these same goals. Often there are impediments that restrain student engagement.

Dissimilar cognitive style may be one impediment.









This study can improve undergraduate instruction in colleges of agriculture and

life sciences by providing evidence of the educational impact of dissimilar cognitive

style. Given this finding, faculty members can better develop programs and curricula to

increase student engagement by reducing cognitive gap caused by dissimilar styles.

Departments of agricultural education can help faculty members in colleges of agriculture

and life sciences better prepare for teaching students with dissimilar cognitive styles.

Also, university faculty can gain valuable knowledge concerning how learning style, not

learning level, dictates how students view problems.

This study contributed to the knowledge base of learning styles by providing data

concerning the relationships between dissimilar cognitive styles, student stress,

motivation and engagement in the undergraduate classroom. Furthermore, collaborative

research among learning style authors may be initiated to determine if they also can

contribute to improving student engagement.

Purpose of the Study

The purpose of this study was to determine if significant relationships exist

between cognitive style gap, student stress, student motivation, student engagement and

selected demographic variables of undergraduate students at the University of Florida in

the College of Agriculture and Life Sciences. The specific objectives of this study were

to:

1. Describe selected faculty and students according to their selected demographic
variables;
2. Determine the cognitive style, student stress, student motivation, and student
engagement of undergraduate students;
3. Determine the cognitive style gap between faculty and students and to explore its
relationship with student stress, student motivation, student engagement and
specific demographic variables of undergraduate students;
4. Explain undergraduate student stress and student motivation based on cognitive
style gap and selected student demographic variables;









5. Explain undergraduate student engagement based on cognitive style gap, student
stress, student motivation and selected student demographic variables.

Assumptions

Basic assumptions of this study were as follows:

* Student engagement is desired by faculty members in colleges of agriculture and
life sciences.

* Faculty members teach in their preferred cognitive style.

* Problem solving learning exists in the undergraduate classroom.

Limitations

Conclusions and implications derived from this study were subject to the

following limitations. Undergraduate courses are self-selected groups with students

having individual differences prior to data collection. It was anticipated that group

differences existed between classes with regard to mean scores and demographic

variables. Therefore, data were limited to this population of faculty members teaching

undergraduate courses in the College of Agricultural and Life Sciences at the University

of Florida. Applications of the results of this study to other college classrooms were

limited to the degree of similarity they have with courses used in this study.

The researcher worked with the conditions allowed by the faculty member to

collect data from their course. All instruments were administered in the same manner, but

some students completed instruments in class and other students completed instruments

outside of class. As the faculty member made choices regarding students' completion of

the instrument, method of completion was confounded with the faculty member and

therefore provided no way to conclude there was a method of completion effect on

instrument scores. Even if courses were compared, self-selected courses have preexisting

differences with no way to statistically determine causality of these differences.









Faculty members volunteered students in their courses to participate in this study.

Although extra-credit points were often given as an incentive to increase response rate,

some students still chose not to respond. In this study, non-responding students included

students who were administered instruments and chose not to respond, and students who

never attended class the day of instrument administration. This study is subject to non-

response error as there was no data collected on non-responders.

Instruments used in the study were self-reported which may affect the accuracy of

the data. The halo effect is a common problem among student self-reports as students

tend to inflate variables such as grades and effort. However, Pike (1999) found that the

halo effect was relatively constant across different student groups and schools giving

evidence that comparisons can be made without disadvantaging a group. However, a halo

effect on the KAI presents an unusable cognitive style score thus increasing mortality rate

of the instrument. Removing unusable scores from the data may have influenced findings

in this study, especially if unusable cognitive style score is related to anxiety (Kirton,

1999).

Other confounding variables that may affect this study include coping behavior

(Kirton, 2003) of the faculty, competence of the faculty, motivation of the faculty,

availability of teaching resources, class sizes, and classroom environment.

Definitions

The following definitions are used in this study.

Adaptive. A preference of cognitive style which produces fewer ideas within a narrower
perspective while tending to utilize given rules. Adaptive cognitive style contrasts with
an innovative cognitive style (Kirton, 2003).

Behavior. An individual's actions and reactions while operating in an environment
(Bandura, 1986). Many factors contribute to an individual's behavior, but one factor is
preferred cognitive style blended with coping behavior (Kirton, 2003).










Cognitive affect. The source of problem selection with a degree of priority placed on the
search for a solution by means of motivation (Eccles, 1983; Kirton, 2003; Schunk, 2000).

Cognitive climate. The mode preferred cognitive style held by a group of individuals
working together to solve problems (Kirton, 2003).

Cognitive effect. A part of cognitive function incorporating both cognitive style, which
influences behavior, and cognitive level which limits behavior by means of cognitive
process (Kirton, 2003).

Cognitive function. The controlling influence on behavior through the interaction of
three components: cognitive effect, cognitive affect and cognitive resource (Kirton,
2003).

Cognitive level. The aptitude, scope or expertise an individual has in solving a problem.
It can be expressed as intelligence or level of creativity, but is not correlated with
cognitive style (Ericsson, Krampe & Tesch-Romer, 1993; Guilford, 1975; Kirton, 2003).

Cognitive process. An operation through a theoretical structure of thinking which one
advances step by step through the stages assumed to be the operations of the brain.
Cognitive process is distinguished from cognitive level and cognitive style, but utilizes
both in problem solving (Kirton, 2003; Kolb. 1984; Wallas, 1926).

Cognitive resource. The storage of learned information (memory) manifested through
constructs of knowledge, experiences, and skills (Kirton, 2003).

Cognitive style. The consistent individual differences found in ways of organizing and
processing information and experience (Messick, 1984). This study uses cognitive style
synonymously with learning style (Claxton & Ralston, 1978). Cognitive style was
operationalized by the Kirton Adaption-Innovation (KAI) Inventory with constructs of
sufficiency of originality, efficiency, and rule/group conformity with tendencies to be
more adaptive or more innovative within these constructs (Kirton, 2003).

Cognitive style gap. A degree of separation between cognitive styles of two individuals
or between an individual solving a problem and the cognitive style required to solve a
specific problem. (Kirton, 2003). Operationalized by subtracting a faculty member's
cognitive style score from a student's cognitive style score.

Cognitive technique. A learned skill an individual may utilize to increase cognitive level
or avert from a preferred cognitive style (Kirton, 2003).

Coping behavior. A learned skill that is initiated with motivation and used to perform in
a cognitive style that is different from one's preferred cognitive style. Coping behavior
can only last for a duration and intensity (Kirton, 2003). Coping behavior in this study is
not to be confused with coping behavior as a reaction to stress which is defined as a









conscious effort to deal with an event to which an individual has not yet developed
automatic patterns of response (Shields, 2001).

Engagement. The extent that students invest their time and psychological energy in
empirically derived good educational practices. (Astin, 1996, Kuh, Hayek, Carini,
Ouimet, Gonyea & Kennedy, 2001). Engagement was operationalized with the use of the
National Survey for Student Engagement (NSSE) (2005) with constructs of academic
challenge, active learning, and student-faculty interaction.

Environment. The social context individuals interact with during the problem solving
process (Kirton, 2003). Within the classroom these variables include: personalization,
involvement, student cohesiveness, satisfaction, competition, task orientation, and rule
clarity (Fraser, 1998).

Faculty Member. An individual responsible for presenting instructional discourse,
providing problem sets for student completion and assigning the final course grade.

Innovative. A preference of cognitive style to produce a larger number of ideas with a
wider variety of perspectives to make a product different while tending to alter rules
when solving problems. Innovative cognitive style contrasts with adaptive cognitive
style. (Kirton, 2003).

Learning style. A consistent way an individual responds to, and uses stimuli in the
context of learning. (Claxton and Ralston, 1978). This study uses cognitive style
synonymously with cognitive style (Messick, 1984). Learning style was operationalized
by the Kirton Adaption-Innovation (KAI) Inventory with constructs of sufficiency of
originality, efficiency, and rule/group conformity with tendencies to be more adaptive or
more innovative within these constructs (Kirton, 2003).

Motivation. A process that includes both initiating and maintaining a goal-directed
activity (Pintrich & Schunk, 2002). Comprises of three different components:
expectancy-a student's belief about their ability to perform a task, value-a student's goal
and belief concerning the importance and interest of the task and affective-a student's
emotional reactions to the task (Pintrich & De Groot, 1990). Motivation was
operationalized with use of the Motivated Strategies for Learning Questionnaire (MSLQ)
with constructs of intrinsic motivation, extrinsic motivation, task value, control of
learning belief, self-efficacy for learning and test anxiety.

Problem. A desire or felt need for a solution, but not immediately knowing the mental
operations to arrive at the solution (Soden, 1994).

Products and ideas. A solution to a problem fitted to the degree of structure preferred by
the problem-solver. A more adaptive person will prefer more structure in the generation
of ideas recognizing boundaries of the relevant paradigm and a more innovative person
will prefer less structure in the generation of ideas disregarding the boundaries of the
relevant paradigm (Kirton, 2003).










Stress. The perception of a stimulus in the form of an event or condition and requires an
individual to adjust from normal life. Stressors can be positive or negative depending on
the individual's reaction to the experience (Gadzella, 1994; Romano, 1992). Stress was
operationalized with use of the Student-life Stress Inventory (SSI) with constructs of
frustrations, conflicts, pressures, changes and self-imposed stress.


Summary

In summary, learning styles theory has become popular as pragmatic educators

can easily see its existence in the classroom. However, many learning style instruments

have proven to be unreliable and invalid. There is little empirical evidence that dissimilar

learning styles contribute to understanding stress, motivation, and engagement; these are

common variables found in the undergraduate classroom. Kirton's theory of A-I provides

a framework to distinguish cognitive style from cognitive level to help understand the

influences of the aforementioned variables. As faculty members become more

accountable for student engagement and look for instructional methods to promote

learning, evidence is needed to determine the importance of learning styles in student

engagement. Kirton's definition of cognitive style and psychometrically sound instrument

to measure cognitive style may be of particular use.














CHAPTER 2
REVIEW OF LITERATURE

The purpose of this study was to determine if significant relationships exist

between cognitive style gap, student stress, student motivation, student engagement and

selected demographic of undergraduate students in the College of Agriculture and Life

Sciences at the University of Florida. The specific objectives of this study were to: 1)

describe selected faculty and students according to their selected demographic variables;

2) determine the cognitive style, student stress, student motivation, and student

engagement of undergraduate students; 3) determine the cognitive style gap between

faculty and students and to explore its relationship with student stress, student motivation,

student engagement and specific demographic variables of undergraduate students; 4)

explain undergraduate student stress and student motivation based on cognitive style gap

and selected student demographic variables; and 5) explain undergraduate student

engagement based on cognitive style gap, student stress, student motivation and selected

student demographic variables.

This chapter summarizes pertinent literature concerned with learning styles,

stress, motivation and engagement. Special attention will be given to Kirton's cognitive

style (Kirton, 2003) as applied to the context of the undergraduate classroom, thus

providing the theoretical framework for the study. Also a discussion of empirical research

conducted to examine learning style, stress, motivation, and engagement, is presented to

determine relationships between these variables.









Within this chapter are the following major sections: theoretical framework,

learning and learning styles, learning and stress, learning and motivation, and learning

and engagement.

Theoretical Framework

Literature concerning learning styles includes a wide variety of published sources

consisting of empirical research found in refereed journals, pedagogical

recommendations found in teacher newsletters, and books advocating the use of a

specific learning style inventory (Coffield, Moseley, Hall & Ecclestone, 2004). Said

differently, the scope of learning styles research is based on instrument development,

adding to theory, classroom improvement, and commercial use. Kirton's A-I theory

(1976) is a theoretical response to the literature stating that all people can be creative in

problem solving in their own cognitive style, thus distinguishing between cognitive style

and cognitive level (Kirton, 1978). Kirton's A-I theory provided the theoretical

framework of this study.

The most detailed explanation of A-I theory is given by Kirton (2003) in

Adaption-Innovtion: In the context of diversity and change, with illustration of cognitive

function, behavior and environment in relation to idea formation (p. 37). Within the

theoretical model, cognitive style and cognitive level are components of cognitive effect.

Cognitive effect is defined as a part of cognitive function incorporating both cognitive

style and cognitive level, both of which are employed by the individual to complete the

cognitive process (Kirton). The strength of A-I theory is the separation and independent

measure of cognitive style from cognitive level. Cognitive style is defined by Kirton

(2003) as "a strategic, stable characteristic-the preferred way in which people respond to

and seek to bring about change" (p.43). Kirton's definition of cognitive style regarding









adaptiveness and innovativeness will be explained further in subsequent sections of this

chapter. Examples of cognitive level include measures of capacity or ability such as

intelligence. However in problem solving, both style and level are required to complete a

cognitive process. Cognitive processes can be described as the "steps of how I operate"

and "not measured, but validated to provide information of where we are" (p.38) in

solving a problem. Examples of cognitive process include Kolb's cycle of Experiential

Learning (de Ciantis & Kirton, 1996; Kolb, 1984) and the stages of problem solving

(Wallas, 1926). Each stage of a cognitive process can be completed at a high cognitive

level or low cognitive level and through the use of different cognitive styles. The

outcome of cognitive process is knowledge (learning) stored in cognitive resource

(Kirton, 2003). The cognitive process can also be accomplished with low motivation or

high motivation as determined by cognitive affect (p. 89). Cognitive affect is defined as

the motivation and value placed on the search for a solution to a problem (Kirton).

Ultimately, cognitive affect, cognitive resource and cognitive effect are three distinct and

uncorrelated components of cognitive function (Kirton) and employed by the individual

to solve problems.

This study examined cognitive style gap between the teacher and the student and

its relationship with stress, motivation and engagement. Other gaps exist between the

cognitive function of teacher and student including: cognitive gap of level-a difference of

intelligence, cognitive gap of affect-a difference of motivation or value and cognitive gap

of resource-a difference of knowledge or skill (Kirton, 2003). These cognitive gaps

influence the interaction between two individuals, but the separation of cognitive style









from cognitive level, cognitive affect and cognitive resource allows for a more precise

measure of dissimilar cognitive style and its relationship with classroom engagement.

In the effort of solving a problem, individual differences become apparent at the

realization of cognitive gaps. Nevertheless, if the motivation is great enough for teacher

and student, both will be engaged in the process of learning (Kirton, 2003). Ideas and

solutions may come from both student and teacher. Although students are responsible for

engaging in learning activities, the teacher facilitates learning and therefore accountable

for student engagement. Student engagement includes constructs of academic challenge,

active learning and student-faculty interaction (Kuh, 2001). Teacher engagement

facilitates student engagement and can be measured by the teacher's knowledge of the

learning process, intrinsic motivation for teaching, and self-efficacy toward teaching

(Knowles, 1999).

Kirton (2003) defines the environment as the social context in which individuals

interact during the problem solving process. An individual's interaction with the

environment manifests as people work together in the environment to solve problems and

react to the feedback provided by others (Kirton). In the undergraduate classroom,

students interact with fellow students and the faculty member through discussions,

activities, assignments and projects; all of which provide student feedback either by

fellow students or the faculty member. In a review of instruments measuring factors

influencing the undergraduate classroom environment, Fraser (1998) categorized seven

scales: personalization, involvement, student cohesiveness, satisfaction, competition, task

orientation, and rule clarity. However, Fraser further discussed that this list may not be









conclusive as little research has been conducted regarding the environment of

undergraduate classrooms.

Finally, student academic achievement results from the faculty member and

students engaged in the classroom environment for the purpose of learning. Academic

achievement can be measured regarding a specific assignment or a cumulative total

provided by the faculty member with values represented as either a percentile or a letter

grade. Figure 2-1 displays Kirton's A-I theory in the context of a teacher and student

engaging in idea formation.

For the individual, coping behavior is used when the rewards gained from solving

the problem are greater than the effort exerted to operate outside of an individual's

preferred style (Kirton, 2003). Said differently, cognitive style gap causes stress and

motivation is needed to overcome this stress and bridge the cognitive style gap between

two individuals. Researchers have investigated the relationship cognitive style gap has

with stress. McCarthy (1988) found that women managers' stress increased as the size of

the cognitive style gap increased. Furthermore, the threshold at which stress became

apparent was at a cognitive style gap of 20 points. Another study conducted by Kubes

and Spillerova (1992) had concurring results with a large cognitive style gap influencing

the construct aggravation within a working group. However, both of these studies were

conducted in the context of business. Little is known if A-I theory is appropriate for the

undergraduate classroom and if these relationships between dissimilar cognitive style and

student stress also apply.



































Environment
(Classroom)
Personalization
Involvement
Student Cohesiveness
Satisfaction
Competition
Task Orientation
Rule Clarity


:Student Cognitive Processes
STeacher Cognitive Processes
SStudent Motivation
:Teacher Motivation
SStudent Learning
Teacher Learning
= Cognitive Gap of Style
= Cognitive Gap of Level
= Cognitive Gap of Affect
= Cognitive Gap of Resource


Figure 2-1. Kirton's theoretical model of adaption and innovation applied to the undergraduate classroom (Kirton, 2003)


A


G4

V









Both intrinsic and extrinsic motivation can be driving forces for increasing

engagement in the problem solving process (Hennessey & Amabile, 1993). In the event

that a course is taught by a faculty member with a dissimilar cognitive style, a student

may be motivated to cope outside of a preferred cognitive style in effort to interact more

with the faculty member and improve student engagement (Kuh, 2001). Motivation is not

a result of cognitive gap, but is a driving force for using a cognitive style that is not the

individual's preferred style. Specifically, motivation is a factor of coping behavior

(Kirton, 2003); if motivation can be partially explained by cognitive style gap, then there

is evidence that students are coping. Motivation is a component of cognitive affect and

therefore independent of cognitive style, a component of cognitive effect.

Utilizing the framework provided by A-I theory, student engagement was

determined to be an outcome of problem solving and idea formation. Student engagement

was used instead of academic achievement in this study to provide a better measurement

of the interaction between faculty member and student.

Kirton (2003) asserts that adaptive and innovative individuals prefer different

amounts of structure while completing the problem solving process. Whereas more

adaptive individuals tend to solve problems favoring structure, more innovative

individuals tend to solve problems unbound by structure. Adaptive individuals tend to

have fewer ideas making the solution more efficient; innovative individuals prefer to

have a wide range of ideas with the knowledge that many of the ideas are improbable

(Kirton, 1994). The difference between two styles of problem solving is often evident by

the differences of perception and communication of the solution. That is, a gap between

two dissimilar cognitive styles tends to inhibit effective communication. This failure of









communication can often lead the two individuals to focus on the new problem of

ineffective communication and not the problem at hand which is completing course

projects and assignments (Kirton). If this occurs in the undergraduate classroom, students

may dwell on the inefficiencies of communication with the faculty member instructing

the course and may not be motivated enough to engage in learning. The constructs of

student engagement include level of academic challenge, active and collaborative

learning, and student interaction with the faculty member, (Kuh, Hayek, Carini, Ouimet,

Gonyea & Kennedy, 2001), all of which require communication between the faculty

member and the student to be achieved (Umbach & Wawrzynski, 2005). In summary, the

faculty member determines the amount of structure used in the course, but students may

prefer different amounts of structure and find themselves unmotivated to effectively

communicate with the faculty member, thus impacting the level of student engagement.

Kirton's (2003) A-I theory may contribute to the explanation of student

engagement by identifying the relationships between cognitive style gap, stress and

motivation. An increase in cognitive style gap between the student and the faculty

member causes stress through working together in solving problems or engaging in

learning. If the student is motivated, coping behavior is used to operate outside of a

preferred cognitive style and employ a cognitive style closer to that of the faculty

member (Kirton). As a result, student and faculty member are operating in the same

cognitive style which may increase student engagement. However, student motivation

supplied to coping behavior only last for a period of time and at a limited level (Kirton).

Consequently the student reverts back to a preferred cognitive style which may be

detrimental to student engagement. For the student to be successfully engaged the reward









for learning must be greater than the stress imposed by cognitive style gap from the

faculty member (Kirton). Figure 2-2 portrays the interaction of student stress and

motivation in explaining student engagement.

O_

0 C
o Student
r
0(> 0 Engagement
(Dmm
S10
Student Behavior


oa

09D



Teacher Behavior


30


50


70


I Qn


Figure 2-2. Interaction of cognitive style gap, stress, motivation and engagement (Kirton,
2003)


Other variables that influence student engagement include faculty member

behavior (Umbach & Wawrzynski, 2005), class size (Ehrenberg, Brewer, Gamoran &

Willms, 2001), use of teaching methods (Marzano, Pickering & Pollock, 2001), gender,

ethnicity, full-time student status, organizational involvement and discipline of study

(Pike, 2004).

There were other confounding variables but were considered outside the

objectives of the study. They included teaching effectiveness, teacher engagement,

classroom environment and student's previous knowledge. Furthermore, it was









impossible to determine the variables outside of the classroom that inadvertently effect

student engagement. These included but were not limited to sickness, family crisis and

lack of time devoted to study.

Kirton (1994) offers the following anecdote to teachers and students in reference

to A-I theory:

We all teach children to be more adaptive: before answering the question, please
read it; be more methodical; present your work more neatly; nice essay, pity it is
not on the topic I set; and so on. The best teachers taught most of us a lot of
technique, but changed us (our preferred cognitive style) not at all. In the process
they found some of us adaptorss) more receptive to working more closely within
structure than others (innovators). Lesser teachers, especially when faced with the
less motivated innovators, got even less far. Of course, despite the groans in the
literature, not all teachers are or ever were, adaptors. The highly innovative ones
delighted most of us, whether we always understood them or not; some of us (the
higher adaptors) were just confused. As the saying goes, "You just can't win 'em
all'." The best teachers win more than most.

(p. xxiii).

The undergraduate classroom offers a new context for A-I theory as well as the

introduction of new variables. Although these variables were not discounted by A-I

theory, up to this point, they had not been fully explored.

Given the previous description of A-I theory, the researcher found two studies of

interest. First considering the undergraduate classroom, Puccio, Talbot and Joniak, (1993)

found that students' stress increased with the perceived adaptiveness of the course.

Specifically, the rule/group conformity construct of the KAI was most important in

determining the relationship between cognitive style and stress. It seems logical that an

innovative student would find an adaptive course stressful, but why would an adaptive

student find an adaptive course stressful? The researcher found the study conducted by

Puccio, Talbot and Joniak one of the few investigations regarding Kirton's A-I theory

applied to the undergraduate classroom.









In the second study, Barnhart (2002) set out to determine if increased cognitive

gap led to aviation flight instruction inefficiencies. No significant differences in

effectiveness were found between total cognitive style gap between students and flight

instructors, however significant effects were found by analyzing KAI construct gaps.

Specifically, a significant weak negative correlation of the efficiency cognitive style gap

and time spent preparing for the first solo flight (r=-.23, p<.05) and time spent preparing

for certification (r=-.26, p<.05). Furthermore, there was a weak negative correlation of

rule/group conformity style gap and time spent preparing for certification (r=-.29, p<.05)

(Barnhart). Although these correlations were considered low (Miller, 1998), one would

expect a positive correlation in this study. That is, cognitive style gap should increase

with the amount of time spent preparing for the first solo flight and certification.

Both studies previously mentioned found low negative correlations between

cognitive style gap and measures of stress and effort. Although these findings contradict

with the theoretical model of this study, it is important to recognize them as two of the

few published studies found by the researcher that applied A-I theory to the context of the

learning environment.

Learning and Learning Styles

To understand learning styles, it is not only necessary to define learning, but also

to provide learning theory to give evidence of different learning types, processes and

styles. Gagne (1965) proposed eight types of learning: 1) signal learning; 2) stimulus-

response learning; 3) chaining; 4) verbal-associate learning; 5) multiple discrimination; 6)

concept learning; 7) principle learning; and 8) problem solving (p. 33). The types of

learning are hierarchical as the latter is a prerequisite of the former for a type of learning

to take place (p. 60). Additionally, learning can be any one of these eight types which









delineates the learning process of a situation and giving reason for a learner to exhibit a

preferred style and a chosen level (Gange).

Research conducted in the domain of learning styles was categorized by Curry

(1987) into four dimensions which formed the "onion" model. Personality trait types of

learning styles were categorized at the core of the onion, as they were found the most

stable within a learner. The second layer of learning styles represents information

processing, while the third layer includes the social and situational interactions of the

learner. The most unstable category of learning styles includes the preference of

instructional methods, found on the outside layer of the "onion." Curry placed Kirton's

measure of learning style at the core of the onion. An individual's cognitive style defined

by Kirton (2003) does not change as a person ages, and is correlated with many measures

of personality (Kirton, 1999). Nor does it change as the individual solves any particular

type problem. However, in problem solving, an individual may cope outside of one's

cognitive style as the situation requires.

Shuell (1986; as cited in Schunk, 2000) defines learning as "an enduring change

in behavior, or in the capacity to behave in a given fashion, which results from practice or

other forms of experience" (p. 2). It is plausible that each of the eight types of learning

provided by Gagne (1965) is fitting with this definition. That is, each type of learning

may result from an experience and produce a behavioral change. Evidence of this may be

found by examining a type of learning process. Using stimulus-response learning as an

example, a student may be stimulated to increase focus during independent study time.

The discriminating stimulus is the allowance to use background music during

independent study time. This stimulus triggers a student response to either use or not use









background music. If the student uses background music to help focus during

independent study time, another stimulus (praise) is utilized to reinforce the behavior

(Skinner, 1953). Using problem solving learning as an example, a student is stimulated

by a felt need to solve a problem of completing all homework assignments before the end

of the night. The student knows he may use independent study time, but it is hard for him

to focus. He generates solutions to solve his problem of lack of focus by playing music in

the background and executes a plan of action (Wallas, 1926). For both types of learning,

the increase of these experiences (practice) results in a changed behavior (increasing

focus).

Given the definition of learning used in this study, it seems paradoxical that one

would have a preferred learning style that is evident by a stable and innate behavior not

altered by instruction (Kirton, 2003). However, learning styles are a preferred approach

to learning, or a perception of the learning environment to which the learner has a

preference. Claxton and Ralston (1978, p. 7) define learning style as "a consistent way an

individual responds to, and uses stimuli in the context of learning," which is similar to

Messick's (1984) definition of cognitive style. Furthermore, a person may cope outside of

a preferred learning style given a different situation, adequate motivation, and learned

techniques (Jones, Reichard & Mokharti, 2003; Kirton, 2003; Kolb, 1984).

Ford and Chen (2001) found that matching a student's learning style with a

faculty member's learning style improved student learning. However, Ford and Chen also

recognized that type of learning and other variables interact with the matching of similar

learning styles. Jensen (2000) connected scientific studies of the brain to teaching and

learning and wrote, "The whole notion of learning styles becomes irrelevant when we









consider how much variety the brain thrives on" (p.138). However, Jensen further wrote

that learners do have preferences of many learning styles and operate in those styles at

different levels. It is useful to understand aspects of a preferred style in order that

learning differences are realized.

This belief is concurrent with the most complete and thorough literature review

concerning learning styles to date (Coffield, Moseley, Hall & Ecclestone, 2004). In this

literature review, 71 learning styles were held up to criteria of psychometric properties,

theoretical backing and practical usage in the classroom. The authors agreed that even

though most learning style measures fall short of these criteria, each have strengths and

weaknesses. Realizing the complex nature of the brain, it is conceivable that more than

71 different learning styles do exist and further research will bring greater understanding

as to how these styles associate with learning. It is interesting to note that of the 71

learning styles reviewed by Coffield, Moseley, Hall and Ecclestone, only a number of

authors relied on related theories and concepts of learning and all used different

terminology, yet the term learning style and cognitive style encompass all measures of

preferred behaviors in learning. Coffield, Moseley, Hall and Ecclestone argue for more

collaborative research between learning style authors as findings may bring greater

understanding to the process of learning and student engagement.

It may be possible that a specific learning style is congruent with one of the eight

types of learning noted above by Gange (1965). For example, Kolb's (1984) measure of

learning style and process is harmonious with the learning of a principle. To learn a

principle a learner engages in a concrete experience, followed by self-reflection and

abstract conceptualization and ending with experimentation of that principle.









Furthermore, Kirton's (1976) measure of learning style is harmonious with learning from

problem solving. To learn from problem solving, a learner manipulates a principle and

interprets the meaning from the manipulation. For effective learning from a problem,

Pask (1976) claims the learner must use both holist and serialist cognitive styles.

Although Pask and Kirton measure cognitive style differently, they are much the same

when considering description and categorization (Cassidy, 2004). Both of these measures

of cognitive style are representative of problem solving style.

Other types of learning defined by Gange (1965) may be compatible with specific

learning styles. For example, the Group Embedded Figures Test (GEFT) (Witkin,

Oltman, Raskin & Karp, 1971) measures whether an individual is either field independent

or field dependent by the individual's ability to identify embedded figures within

geometric shapes. The cognitive processes used to complete the GEFT gives some

evidence that field independence and field dependence is a style of concept learning

(Gange, 1965). Concept learning is defined as comparing and distinguishing properties of

similar items based on knowledge of those items (Gange). However, Witkin (1973) has

found that his measure of learning style does have implications for problem solving with

regards to the need for structure. Field dependent learners tend to prefer more structure in

learning while field independent learners prefer less structure. This corresponds to

Kirton's (2003) adaptive individuals who favor structure and innovative individuals who

prefer less structure in the effort of problem solving.

Learning Style and Learning Process

Most psychologists concur that learning occurs through a process or cycle.

Research of the brain has found five stages for effective learning (Jensen, 2000),

including preparation-background recall of information, acquisition-inputting









information, elaboration-experimentation and feedback, memory formation-encoding

through emotion and context, and functional integration-application of knowledge

(Jensen). It is easy to see these stages within other learning processes. Kolb (1984)

defines the process of learning a principle as a concrete experience with the principle to

be learned, followed by self-reflection and abstract conceptualization and ending with

active experimentation of that principle. Teaching and learning the problem solving

process includes the stages of problem identification, idea generation, idea evaluation and

implementation (Swartz & Perkins, 1990). Still, the process the student uses to gather

information dictates how the information is stored (Jensen, 2000; Piaget, 1970). That is, a

learner studying a principle will acquire knowledge of the principle, but not how to

manipulate the principle to solve a problem; knowledge to solve a problem must be

acquired from the process of problem solving learning.

Kirton (2003) argues that learning style is independent from the learning process,

a contradiction of Kolb's experiential learning theory and Kolb's measure of learning

style. Kirton empirically tested (de Ciantis & Kirton, 1996) his claim by conducting a

factor analysis on the Learning Style Questionnaire (Honey & Mumford, 1992), a

learning style measure congruent with principle learning. The Learning Style

Questionnaire was found as a more reliable version of Kolb's Learning Style Inventory.

Following factor analysis, comparisons were made with Kirton's Adaption-Innovation

Inventory to make the argument of independence between style and process. However, it

may not be sensible to compare a learning style of principle and a learning style of

problem solving as the two are of different learning processes. That is, a learning style of

problem solving may be independent of principle learning, but not independent of









problem solving learning. Furthermore, Kirton (2003) claims that adaptive learners have

preferences of inductive thinking methods leading to better problem representation, while

innovative learners tend to use deductive thinking methods to proliferate more ideas (p.

172). Again, both problem representation and idea generation are components of the

problem solving process. The process of learning and learning style may not be

completely independent as Kirton suggests. In addition to this, effective learning from

problem solving requires the learner to use cognitive techniques that correspond with

adaptiveness and innovativeness (Pask, 1976) while manipulating a principle to solve a

problem. If coping behavior is required for an individual to be more adaptive or

innovative, there is theoretical evidence that in problem solving learning stress is

inevitable and motivation is needed to complete the problem solving learning process.

Learning Style and Learning Level

Kirton (1994) further separates problem solving in terms of cognitive level and

cognitive style. Whereas cognitive style is fixed, innate and guides preferences of

structure for solving problems, cognitive level is the capability to solve problems at a

degree of complexity (Kirton). Examples of cognitive level include intelligence,

creativity, academic achievement, etc. Again, the separation of cognitive style from

cognitive level is the strength of A-I theory. The distinction and accurate measurement of

cognitive style allows for the precise measure of other variables such as student

intelligence, engagement, and motivation.

According to A-I theory, all people can creatively solve problems and have

original ideas (Goldsmith, 1994). However when solving problems, people who are more

adaptive tend to produce a smaller number of ideas with a narrower perspective that tend

to be more relevant, sound and useful while utilizing given rules. People who are more









innovative generate a larger number of original ideas of which many are different with

widely changing perspectives and tendencies to alter rules (Kirton, 1994). Both adaptive

and innovative people can adequately solve any problem, just within their own style.

Kirton's Cognitive Style

From the definitions of learning style (Claxton and Ralston, 1978) and cognitive

style (Messick, 1984) used in this study, Kirton's cognitive style was determined as a

style of learning. Additionally, Kirton's Adaption-Innovation Inventory (KAI) simply

measures style of problem solving and is not convoluted with factors of cognitive level,

cognitive affect or other types of learning (Gange, 1965). Three factors located on a

continuum of more adaptive to more innovative have been identified and determined as

preferential behaviors of an individual. Table 2-1 displays the factor traits of adaptors and

innovators.

Table 2-1. Factor Traits of Adaptors and Innovators
Construct Adaptor Innovator
Sufficiency Tendency to produce fewer ideas Tendency to produce more ideas
of originality aimed at being sound, useful and that are both relevant and
relevant to solving the problem. irrelevant to the problem with


Efficiency Preference for attention to detail
to further develop ideas while
narrowly focused on improving.


knowledge that some ideas are
inappropriate.

Preference for discontinuous
thought to modify the problem
with broad approaches to make
different.


Rules/Group Often applies group accepted Often challenges group accepted
conformity rules in idea formation to solve rules in idea formation to solve
problems. problems.

Note. Source: Kirton, 2003, p. 58-60.

The factor traits of A-I theory are "innate, fixed and reliably measured" (Kirton,

2003, p. 76). Each of the three factor traits lends to a perceived structure of the problem


t









at hand which determines the approach needed to solve the problem. The more adaptive

tend to work within the frame of structure. A more adaptive cognitive style rarely

challenges the boundaries of structure and solutions to the problem are focused with

detail to derive a better solution that improves efficiency. A more innovative cognitive

style breaks boundaries as if they never existed and solutions generated are often

unrelated in an effort to derive a different solution that transforms the structure itself

(Kirton). The perception of the structure from the framed problem determines the

response of the learner, providing for the use of a preferred cognitive style.

The Scale of Adaption and Innovation

A benefit of A-I theory and the KAI is the interval measure of difference between

two cognitive styles, providing a specific degree of separation between an individual's

cognitive style from another individual's cognitive style. This is opposed to a learning

style measured nominally and determined as categorically different; a model for many

learning style inventories. The KAI measures cognitive style by scoring adaptiveness and

innovativeness along a continuum (Kirton, 2003). Anchoring the more adaptive on the

left and the more innovative on the right of the continuum, an interval scale is created

amid a theoretical range of 32 to 160 (Kirton). However, the actual range is

approximately 45 to 145 with a mean of 95 (Kirton).

It is important to note that the terms adaptors and innovators are used only to

distinguish two differing cognitive styles, when in actuality, cognitive styles refer to

being more adaptive or more innovative in the context of a situation (Kirton). For

example, an individual may have a cognitive score of 125 distinguishing them as more

innovative, but this individual will be perceived as adaptive to an individual with a score

of 140.









Kirton (2003) suggests that a 10-point difference in score between two

individual's cognitive style, as measured by the KAI, develops a "noticeable difference"

when working together on a problem. The difference found between the two cognitive

styles is defined as cognitive style gap (Kirton). A 20-point cognitive style gap leads to

difficulties in mutual understanding and collaboration which may have a definitive effect

on communication when problem solving (Kirton). For an individual with this amount of

cognitive gap between one's preferred cognitive style and the style required to work with

a problem or another problem solver, coping behavior is needed and can exist in both

intensity and duration (Kirton). Coping behavior requires motivation (Walling, 1987) and

is psychologically expensive (Kirton, 2003).

When solving a problem, more adaptive individuals can use techniques such as

brainstorming which can be taught, however, they are not preferred and less likely to use

it in their natural environment. Innovative individuals, likewise can be taught to master

details and operate within rules, but would prefer the unstructured situation. Kirton

further suggests that practice of coping behavior does not make coping easier for that

person (Kirton, 2003, p. 255). This suggestion does limit the definition of learning used

in this study; however the suggestion is supported in the creativity literature. Few

researchers have been able to provide hard evidence that an individual can be taught to be

more innovative or more adaptive (Nickerson, 1999). Furthermore, there is evidence that

teaching an individual to use different problem solving techniques has little long term

effect (Nickerson). Cognitive style (Kirton, 2003) is fixed, innate, and determines an

individual's preference for structure within problem solving. There is no association with









cognitive style and nationality or age, however females tend to be five points more

adaptive then the general population mean (Kirton).

Summary of Learning and Learning Styles

Gange (1965) described eight different types of learning that are hierarchical in

nature. Of the eight types, learning from problem solving is the highest form of learning

with evidence of a problem solving process and problem solving style or cognitive style.

Kirton's measure of cognitive style is a preferred approach to solve problems and

independent from cognitive level (Goldsmith, 1994). However, both cognitive style and

cognitive level are utilized to complete the problem solving process. There is evidence

that other measures of learning style are specific to other types of learning defined by

Gange (1965) with evidence found in the learning process.

Coffield, Mosley, Hall and Ecclestone (2004) identified 71 different types of

cognitive styles with varying degrees of reliability and validity. Kirton's measure of

cognitive style was determined a reliable and valid measure of cognitive style (Kirton,

1999). Furthermore, the measure was found independent from measures of motivation

and capability (Kirton). Kirton's cognitive style determines if a person is more adaptive

or more innovative in one's preference to solve problems and is stable characteristics of

the individual. Along a continuum of adaption anchoring the left and innovation

anchoring the right an individual's cognitive style score is placed in a theoretical range of

32 to 160 (Kirton). Dissimilarity between the cognitive styles of two individuals is

determined by calculating the difference between two cognitive style scores. This

cognitive style gap indicates differences between preferred methods of solving problems

with stress as a consequence. Motivation may be utilized by the student to operate in a

different cognitive style than preferred and bridge cognitive gap (Kirton, 2003). A









cognitive gap larger than 20 points may inhibit communication (Kirton) between the

student and faculty member and may explain lack of student engagement.

Learning and Stress

College students may be influenced by stress positively or negatively (Sandler,

2000). That is, stress as a stimulus can provoke a negative or positive reaction to the

situation at hand. If a student has a positive reaction it is the fear of failure that motivates

this student to succeed, but a negative reaction will likely be detrimental to classroom

success. Given this variability of outcome from stress, little research has been conducted

with attention to stress associated specifically concerning the undergraduate classroom.

A word of caution is given regarding terminology and definition in discussing

stress. Coping behavior specific to the reaction to a stressor can be defined as a conscious

effort to deal with an event to which an individual has not yet developed automatic

patterns of response (Shields, 2001). In this document, reaction is used instead of the

word coping so that the reader does not confuse coping as a reaction to stress with coping

behavior defined by Kirton (2003).

A student's reaction to the stressor may determine if the stress is beneficial or

detrimental to the academic achievement of the student. In a study conducted to examine

this hypothesis, Aspinwall and Taylor (1992) surveyed 672 freshman undergraduates.

Their findings provide evidence that a positive mood, higher optimism and an active

reaction to stress orientation had positive effects on adjusting to new college experiences.

Furthermore, the reaction to stress construct avoidance had detrimental effects on the

adjustment to new college experiences. Hockey (1979) considered the role of stress in

aiding the selectivity of attention and response in performing a cognitive task. Hockey









elaborated that there may be an optimal level of stress for cognitive function although

determining this level is dependent on the components of the task.

Researchers have studied the detrimental effects of anxiety, a correlate of

academic stress (Misra and McKean, 2000). In countless studies, test anxiety has been

found to be detrimental to academic achievement and self-esteem (Pintrich & Schunk,

2002). Specifically, worry is a factor of anxiety that causes distractions resulting in time

lost being able to perform at the level the student is capable of achieving. Most of the

literature concerning anxiety focuses on this worry component and its influence on

students' attention and time on task in completing an assignment or test. Like stress,

anxiety is an affective component of Kirton's A-I theory.

Undergraduate students face many stress-causing experiences throughout the time

spent in completing a bachelor's degree. Among these experiences are moving away

from home, academics, finances and work and relationships (Pfeiffer, 2001). The

experiences themselves do not cause stress, but the student's perception and reaction

turns the experience into a stressor (Romano, 1992).

Studies have been conducted to determine the effects of specific interventions of

student stress, such as better time management. Again, none of the studies found by the

researcher dealt with techniques specifically used to improve academics or students'

relationships with faculty members.

Stress in the Classroom

The college environment is much different than the work environment, making

the study of college stress unique in the literature. Research of undergraduate student

stress has been conducted concerning the global stress of a college student as well as

academic stress restricted to the classroom. Academic stressors for undergraduate









students may include homework, unclear assignments, uncomfortable classrooms and

time pressure set by the faculty member (Ross, Niebling & Heckert, 1999). The most

common stressors of college students may be intrapersonal, including the change in

sleeping habits, change in eating habits and new responsibilities (p.317). However,

changes in the environment may also change types of stressors. That is, measuring stress

at different times during the year and at different times leading to the completion of a

degree may find different levels of stressors. Shields (2001) found that stress had

detrimental effects on grade point average during the fall semester, but not the winter

semester. Shields postulates that the effects of a stressor diminishes once students begin

to activate positive reactions to the stressor. Regarding academic stress, Ross, Niebling

and Heckert (1999) found that increased class workload and lower grades than

anticipated were the two most frequent reported stressors among undergraduate students.

Pfeiffer (2001) believes that these academic stressors can cause depression and low self-

confidence (p. 13).

In a study conducted to find the relationship between stressors, learning strategies

and test anxiety, Gadzella, Masten, and Stacks (1998) found that frustration was

negatively correlated (r=-.20, p<.05) with the learning strategy deep processing

(Schmeck, Ribich, Ramanaiah, 1977). That is, students do not become as frustrated when

they become more organized (Gadzella, Masten & Stacks, 1998). Surprisingly, this study

found no significant correlations with factors of stressors and test anxiety.

Using Gadzella's Student-Life Stress Inventory (SSI), Misra and McKean (2000)

found that practices of time management lowered college student stress levels. However,

the Misra and McKean study contradicted the previously mentioned study by Gadzella,









Masten and Stacks (1998) and found significant positive correlations (p<.05) among

constructs of stress and general anxiety among the population of full-time undergraduate

students. Correlations between stress and anxiety should be positive as they both have a

worry component (Misra and McKean, 2000).

Felsten and Wilcox (1992) conducted a study to find the impact of stress on 146

college male's academic performance. After analysis they found that stress was a

detrimental effect on male's grade point average.

Shields (2001) conducted a study to examine the effects of stress and reactions to

stress among persisting and non-persisting undergraduate students. A persister is defined

as a full-time student enrolled in fall and winter semesters, while a non-persisting student

is one that enrolls in the fall semester but not the subsequent winter semester. The group

of study consisted of 220 persisters and 110 non-persisters and findings found that

persisters had significantly higher levels of stress than nonpersistors with no gender

difference. Stress was positively associated with reactive efforts to stress, and age among

persistors, but not nonpersisters. This gives evidence that full-time status or traditional

students have developed reactive efforts to stressors, but non-traditional students do not

develop these same reactive efforts (Shields). Finally with the use of regression, stress,

age and number of credit hours taken in the semester each had significant explanation of

student reaction to stress (Shields). Recognizing that stress affects students in the

classroom, it is necessary to examine the demographic variables related to stress.

Demographic Variables Related to Stress

Gadzella and Guthrie (1993) found that women undergraduate students tend to

have significantly higher levels in experiencing stress of both pressure and change.

Considering age, self-imposed stress was found to be highest among younger students









(16 to 18 years-old) and lowest in older students (31 to 57 years old) (Gadzella &

Fullwood, 1992). Moreover, young women experienced more self-imposed stress and

pressure than any age and gender combination (Gadzella & Fullwood).

Measurement of Stress

In the task of measuring stress, two types of instruments were found in the

literature. The first type lists possible stressors relevant to the situation of the student and

provides the research with either a count of stressors, or an indexed stress score with

certain stressors weighted more than others. Although this measure of stress objectively

attributes stress to the specific stressor, it ignores the interaction of the student with the

environment (Cohen, Kamarck & Mermelsein, 1983). That is, students may attribute

stress to a stressor without considering available resources to react to the stress.

Furthermore, for this type of stress instrument to be accurate a list of all possible stressors

must be presented to the student.

The second type of stress instruments measure perceived stress with regards to the

stressor, situation and ability to react to the stressor. This type of measure broadly

accounts for all stressors and provides a behavioral construct. However, perceived stress

instruments are limited as they may inadvertently measure chronic stress, or other sources

of stress (Cohen et al.). Research concerning perception of stress has found evidence that

"students often misattribute stress to the appropriate stressor" (p. 387). Given this

limitation, the measure of perceived stress is congruent with the definition of stress used

in this study. Again, stress is not a corollary of a particular stressor, but a result of the

student's perception and reaction to an experience (Romano, 1992).

To meet the objectives of this study, stress was measured specific only to the

classroom environment to which the instrument was administered. Again, the classroom









environment for problem solving is described as the social context to which individuals

interact with during the problem solving process (Kirton, 2003). This gave reason to use

an instrument that measured stress as perceived by the student and experienced in the

environment.

The researcher found that Gadzella's Student Life Stress Inventory (SSI) a valid

and reliable measure (Gadzella & Baloglu, 2001) of an undergraduate's perceived

classroom stress that could be adopted for use in this study. The theoretical framework

for the SSI comes from Morris (1990) and treats stress as a student experience, not just

stressor variables affecting students (Gadzella & Masten, 2005). Five stressors were

identified as experiential to undergraduate students and used in the SSI including: 1)

frustration-delays and failures in reaching academic goals, social acceptance and

disappointing grades, 2) conflict-one's choices among desirable and undesirable

outcomes in goal setting, 3) pressure-assessing class competition, deadlines, and

overload of assignments, 4) changes-unpleasant experiences and disruptive schedule

changes and 5) self-imposed-desire to compete, be loved by all, procrastinate finishing

assignments and anxiety during tests (p. 4). Each stressor forms a construct with items

using a five point Likert scale with one signifying never and five signifying most of the

time. Each stressor construct can be used individually or cumulatively for determining

total stress of the student. Chapter 3 further discusses how the SSI was modified to meet

the needs of the objectives for this study.

Summary of Learning and Stress

Stress is defined as an individual's perception and reaction to an experience that

requires an adjustment from normal life (Romano, 1992). Undergraduate students have

many unique stressors including moving away from home, academics, finances, work and









new relationships (Pfeiffer, 2001). Common academic stressors include increased

workload in a course and lower than expected grades received in a course (Ross, Niebling

and Heckert, 1999). Although stress in the form of anxiety was found to be detrimental to

academic achievement, an optimistic attitude may allow students to positively react to

stress (Aspinwall & Taylor, 1992). Evidence was also found that young women

undergraduate students tend to experience more stress than other undergraduate students

(Gadzella & Fullwood, 1992).

The measurement of stress can be categorized by instruments measuring a count

of stressors or a perceived experience which accounts for the reaction to a situation. Both

types of measurement have strengths and limitations (Cohen et al.). This study utilized

the SSI (Gadzella & Baloglu, 2001), a perceived measure of stress which is limited in

ability to determine source of stress but provides a measured level of stress in the form of

a behavioral construct (Gadzella & Baloglu). The measurement of stress as an experience

rather than a specific stressor was found to be more compatible with Kirton's A-I theory.

Learning and Motivation

Concerning Kirton's (2003) theoretical framework applied to the undergraduate

classroom, motivation is identified as part of cognitive affect. Motivation and its

influence on learning is a well-researched phenomena with many more questions still

unanswered. Research has found that learning is dependent on motivation, but

interestingly the achievement of learning goals increases motivation to learn more

(Schunk, 1991). Pace (1980) found that knowledge gained from a course was positively

correlated with the amount of effort students provided for learning. Theories of

motivation in education that are considered prominent at the time of this writing can be

grouped by expectancy-value theories, attribution theories, social cognitive theories, goal









theories, and intrinsic-extrinsic theories, (Pintrich & Schunk, 2002). Although, these

theories each have their strengths, this study used an expectancy-value theory for the

motivation theoretical framework (Eccles, 1983). Expectancy-value theory was found to

be compatible with Kirton's (2003) definition of the environment. For example, a student

may perceive himself as capable (expectancy) of solving a problem and receive some

kind of reward (value) from execution of the solution. However, the faculty member also

provides feedback on the student's effort and solution to which the student will react. If

the student's perceived ability and reward are great enough, the student will be able to

solve a problem despite cognitive style gap (Kirton, 2003). Furthermore, there is

evidence that expectancy beliefs are related to student engagement and value beliefs are

related to opportunities of future student engagement (Pintrich & Schunk, 2002). The

following is a short description of the expectancy-value theory with empirical studies to

support its importance to learning.

Expectancy-Value Theory

Current expectancy-value models have been adapted from Atkinson's (1957)

expectancy-value theory of achievement motivation and now provide the clearest

explanation to students' motivation in the classroom (Pintrich & Schunk, 2002).

However, much is still to be learned about motivation as no motivation theory has been

universally accepted by educational psychologists. The premise of expectancy-value

theory is that the learner is motivated by two components: 1) their perceived ability and

effort expected to be needed in accomplishing a goal (expectancy) and 2) the amount of

value or reward the learner places on attaining the same goal (value). These two

components will be further broken down and discussed theoretically and empirically.









The expectancy component

In this study the expectancy component consists of control of learning beliefs and

self-efficacy. The belief is that a positive increase of an expectancy component will

promote learning. A student's belief of their control in learning, or locus of control, is a

main component of many expectancy-value theories. Students who believe they are in

control (internal) of the learning succeed over students who explain away their control

(external) of learning to other factors (Eccles & Wigfield, 2002). Research has found

some evidence that control of learning beliefs change from externally to internally while

attending college, with the greatest change found in students involved in honors programs

(Pascarella & Terenzini, 1991). Research conducted to understand control of learning

beliefs was prominent between the 1960's and 1970's. After this period of time authors

moved on to answer other questions concerning motivation. Still, control of learning

belief is an important part of many expectancy-value theories and the measurement of

motivation.

Self-efficacy (Bandura, 1977) is accepting one has the capability to complete a

task or goal. Bandura (1994) states the four sources of self-efficacy include mastery of

challenging experiences, vicarious observation of a peer, verbal praise and reducing

stress. As it relates to social-cognitive theory (Bandura, 1986), self-efficacy affects

behavior and environment, but is also an effect of behavior and environment.

Furthermore, there is evidence that self-efficacy may influence the student's choice of

which problem to solve (Schunk, 2000). Referring to expectancy-value motivation in the

context of student engagement, student self-efficacy may affect a student's choice to

solve a problem while expectations of the student and the teacher raise the value placed

on the solution.









Research has shown that higher levels of self-efficacy have a significant effect on

students' persistence in completing a challenging task, academic performance and self-

regulated learning (Zimmerman, 2000). Linnenbrink and Pintrich (2003) reviewed

several studies regarding self-efficacy and claim a student with high self-efficacy will be

more cognitively engaged through the use of higher order thinking skills and

metacognitive strategies. In other research, Abouserie's (1994) findings supported the

negative relationship between self-esteem and the variables of stress and anxiety. He

elaborated that the major cause of test anxiety is the possibility of being found

incompetent; a bash to a learner's self-efficacy.

The literature indicates there is a domain specific nature to students' self-efficacy.

For example, academic self-efficacy was found to account for 21% of the variance in

predicting students' course grades (Choi, 2005); however global self-efficacy was not

found to be a significant predictor of academic achievement (p. 202).

Many authors have examined self-efficacy, self-esteem, and locus of control and

implications these traits have on learning. Though each of these expectancy components

have their own attributes (Ajzen, 2002) they have many of the same core constructs and

may be more alike than different (Judge, Erez, Bono & Thoresen, 2002).

The value component

The value components of expectancy-value theory consists of intrinsic

motivation, extrinsic motivation and task value constructs, all of which assign reward on

attainment of a goal. The belief regarding the relationship between these value

components and learning is simply that an increase of value in completing a learning goal

leads to an increase in learning.









Intrinsic motivation is defined as accomplishing a task for no reward except for

internal feelings that result from completing a task (Deci, 1975). Extrinsic motivation is

accomplishing a task for the benefit gained from a separate outcome (Deci & Ryan,

1985). Although intrinsic and extrinsic motivation is separate and distinct there is some

evidence that the two operate on a continuum. For example, a young boy may be

extrinsically motivated to learn how to play the violin as his music teacher rewards

successful learning with grades, extra-credit points or time off from their regularly

scheduled practice. As time moves on, the boy starts to find enjoyment in playing the

violin and will play more just for the sake of playing. The boy moved from being

extrinsically motivated to intrinsically motivated. This was empirically tested by Ryan

and Connell (1989) and concluded that this continuum does exist by examining

relationships among different types of extrinsic motivation and change in specific

attitudes.

Generally in the literature extrinsic motivation has been considered detrimental to

learning and the level to which students solve problems. However, extrinsic motivation

can be beneficial to learning if used in the correct manner. For example in problem

solving, if a student is told how a solution will be evaluated and rewarded, extrinsic

motivation positively influences engagement in the problem solving process (Collins &

Amabile, 1999). Furthermore, an extrinsic reward offered to a student before beginning

the problem solving process may increase level of problem solving if there is no

contingency contract (Collins & Amabile). In another study, Hennessey and Zbikowski

(1993) attempted to educate students about the negative effects of extrinsic motivation.

The results found that after these students were extrinsically rewarded to solve a problem,









there was no negative influence from the extrinsic reward. More research is needed to

understand the positive influences of extrinsic motivation, especially if students are often

motivated extrinsically and gradually become more intrinsically motivated over time.

A review of the literature finds disagreement in the definition of task value

motivation and how it is utilized in different expectancy-value motivation models. Task

value motivation sometimes incorporates intrinsic and extrinsic interests and is

sometimes referred to as incentive motivation (Pintrich & Schunk, 2002). Other

researchers refer to task value as a level of importance placed on a task (Battle, 1965)

which is how task value is defined in this study. Considering the importance of task value

to learning, Feather (1988) found task value to be specific to a college student's academic

major. For example, a student with a high task value in math had higher perceptions of

their ability in math, as well as an increase of personal values concerning achievement in

math.

Affective motivation

Affective motivation is not a component of expectancy value theory, but is

becoming more common in recent models of motivation as researchers explore the role of

emotion in motivating students. It is included here because the measure of motivation

chosen by the researcher included an affective component identified as test anxiety.

Throughout the literature, there is evidence that test anxiety is detrimental to academic

achievement and learning (Pintrich, Smith, Garcia & McKeachie, 1991). The two main

components of test anxiety are worry and emotion. Some evidence suggests that emotions

cause worry making test anxiety a behavioral construct (Hembree, 1988). Treatment of

students' test anxiety is best provided by reducing worry and the emotionality associated

with the test. This has shown to increase academic performance (Hembree).









Demographic Variables Related to Motivation

Motivation is an unstable consequence of the situation perceived by the individual

and therefore considered contextual or situational (Paris & Turner, 1994). Motivation

continues to change as a learner assesses and reassesses their ability to accomplish the

task and the value of the reward upon accomplishing the task. Demographic variables

such as age, gender, ethnicity, etc. are always factored in the individual's assessment of

the situation and their capability of success in the situation. The environment of the

classroom does allow for all learners to be in the same situation, but the learner

continually constructs motivation (Paris & Turner) making motivation highly

individualized and difficult to generalize to specific demographics.

Measurement of Motivation

The wide variety of motivational theories has led to the development of many

instruments to measure motivation. To meet the objectives of this study and to remain

consistent with expectancy-value theory, the Motivated Strategies for Learning

Questionnaire (MSLQ) (Pintrich, Smith, Garcia & McKeachie, 1991) was used to

determine motivation of each student specific to the class to which the instrument was

administered.

The MSLQ was designed specifically to measure undergraduate student

motivation for self-regulated learning. The three components of the instrument include an

expectancy component with items related to student beliefs of ability in performing a

task, a value component with items concerning goals and reasons for completing a task,

and an affective component which includes items regarding emotional reactions to the

task (Pintrich & De Groot, 1990).









The three components of the MSLQ can be further broken down into six scales of

motivation that can be used independently or cumulatively (Pintrich, Smith, Garcia,

McKeachie, 1991). The value component comprises three scales: intrinsic goal

orientation-participating in a task for reasons of challenge, curiosity or mastery, extrinsic

motivation-participating in a task for good grades, rewards or competition and task

value-an evaluation of the importance of the task (Pintrich et al.). The expectancy

component of the MSLQ has two scales which include control of learning beliefs-a belief

that results hinge on self and effort will provide positive results, and self-efficacy for

learning and performance-an appraisal and judgment of ability to complete a task

(Pintrich et al.). The affective component has one scale identified as test anxiety which

includes student performance and the resulting emotion (Pintrich et al.). It is important to

note that the MSLQ includes an affective component in measuring student motivation,

but there is no affective component in expectancy-value theory. Affective motivation, the

motivation scales that identify emotions, has increasingly been recognized in the

literature during the last 20 years and found to be incorporated into more cognitive

motivation models. More research is needed to understand how emotions play a part in

student motivation to learn.

In a study conducted to determine how the components of the MSLQ relate to

cognitive engagement of students, Pintrich and De Groot (1990) found that self-efficacy

and intrinsic orientation were both positively correlated with cognitive engagement and

academic performance. Test anxiety was not found to be correlated with cognitive

engagement, but did have a significant negative correlation with self-efficacy. The

affective scale of test anxiety was found to be either detrimental or show no relationship









with effort given to the task at hand (Pintrich & De Groot). Given that test anxiety is

composed of worry and resulting emotion, one can easily see the similarities between

stress and test anxiety. Furthermore, positive reactions to test anxiety may have the same

effect that they have on student stressors. Chapter 3 gives more detail as to how the

MSLQ was used to meet the objectives of this study.

Summary of Learning and Motivation

Expectancy value theories of motivation provide insight to how students are

motivated to achieve academic success (Pintrich & Schunk, 2002). This theory of

motivation was established as compatible with Kirton's (2003) cognitive affect as it

relates to the environment in which individuals interact and react during the problem

solving process. The expectancy component of the motivation theory refers to a

perceived ability in accomplishing a learning goal while the value component refers to

the reward placed on accomplishing the same goal (Eccles, 1983). As a student

consistently assesses and reassesses the ability to accomplish this goal and the value

placed on achieving success, motivation becomes situational and individualized (Paris &

Turner, 1994).

Evidence supports the assertion that motivation is necessary for learning and

motivation increases after a learning goal is achieved (Schunk, 1991). Research has

indicated that student engagement was positively correlated with measures of self-

efficacy and intrinsic motivation (Pintrich & De Groot, 1990). Although extrinsic

motivation was considered in the literature as detrimental to learning and student

engagement, extrinsic motivation was found to positively influence engagement in the

problem solving process when used without a contingency contract (Collins & Amabile,









1999). More research is warranted to distinguish between the harmful and useful

practices of providing extrinsic motivation to facilitate student engagement.

This study measured student motivation with the MSLQ (Pintrich, Smith, Garcia

& McKeachie, 1991) which included the following motivation scales: control of learning,

self-efficacy, intrinsic motivation, extrinsic motivation, task value and test anxiety.

Although test anxiety was not part of expectancy-value theory, it was a scale found in the

MSLQ and was included in the study given that affective motivational theories have been

increasingly found in the literature.

Learning and Engagement

Astin (1984) developed a theory of student involvement to help guide future

research in the area of student development and learning in completing an undergraduate

degree. In his writing, Astin explained the tie between student engagement and

motivation as "sharing a common psychological construct" (p. 522). Realizing the

importance of motivation, he also discussed students' time as a resource. Astin's theory

states that student effort and time are two resources required for student engagement in

learning. That is, an increase in effort and time by the student to accomplish a learning

goal will increase student learning. Astin concludes that "students' time and effort are

limited and exhausted" (p. 523) on activities related to college and also life. Astin's

theory also identifies contributors to student engagement, which guided the development

of the National Survey for Student engagement (NSSE), the measure of student

engagement in this study.

The roles that faculty members play as advisors and teachers have incredible

influence on undergraduate students. Even in informal situations, faculty members can

become positive role models for students, influencing attitudes and values (Pascarella,









1980). The importance of the faculty member cannot be denied in contributing to student

engagement in the classroom and therefore facilitating academic success. The study of

student engagement involves identifying the practices used by faculty members to

increase learning and personal development and then surveying students to determine if

those practices are being used. The following is a review of the literature regarding how

the faculty member can influence learning.

The Influence of Faculty Members

Umbach and Wawrzynski (2005) conclude that "the single most important" factor

in student development and classroom engagement is the faculty member. Whittington

(1998) however, has found that faculty members in colleges of agriculture typically teach

with methods that only require lower level thinking skills of the students and is

detrimental to student classroom engagement. However, many colleges and universities

now provide resources to help faculty members improve their instructional discourse. A

replication of Whittington's study is needed to determine if faculty members still teach at

lower cognitive levels.

In a study conducted to examine the relationship between student-faculty

interaction and students' critical thinking skills, Smith's (1977) findings provide evidence

that there is a positive relationship. Specifically, increases in critical thinking ability were

positively correlated among variables of class participation with the faculty member as

well as participation with other students in the class (Smith).

Increasing student engagement is more than just keeping students busy in the

learning process. Schlechty (2002) claims that the amount of time a faculty member

dedicates to improving the quality of assignments is directly related to student

engagement. Many studies have been conducted examining different instructional









techniques, distance education methods of instructional delivery, and uses of technology

and their impact on student academic achievement and engagement while controlling for

learning style. Most of these studies conclude that students have different styles of

learning and that teachers should use a variety of teaching approaches to meet learning

style needs of all students. As most good teachers incorporate a variety of instructional

techniques already, the significance of learning style theory has found little practical

application in the classroom (Coffield, Moseley, Hall & Ecclestone, 2004).

The literature revealed similar conclusions comparing the problem solving

teaching approach to the content based approach. Dyer and Osborne (1996) examined the

use of the two teaching approaches in a quasi-experimental design studying secondary

students enrolled in agriculture. Using the Group Embedded Figures Test (Witkin,

Oltman, Raskin & Karp, 1971) to measure learning style, they found that controlling for

teaching approach, neither field independent or field dependent students had significantly

higher achievement scores. However, they also found that field independent learners had

higher problem solving scores on non-agriculturally based problems. Though their

findings were limited, they also concluded that different instructional strategies should be

used to facilitate the learning of all students (Witkin et al.).

Pascarella and Terenzini (1991) summarize teacher/teaching attributes that are

effective for student learning. They include: command of subject matter, enthusiasm of

subject, clear explanation of concepts, efficient structure of class time, use of examples

and analogies in presenting new concepts, avoiding vagueness, and having good rapport

with students (Pascarella and Terenzini).









Realizing the tie between motivation and student engagement, Paris and Turner

(1994) identified four teaching techniques to increase student motivation to learn, which

include: allowing students to choose a course of action, providing academic challenge,

giving students control of achieving the learning objective, and allowing students to

collaborate in the problem solving process.

Demographic variables related to engagement

Pascarella and Terenzini (1978) conducted a study to examine the academic

performance and social development of student-faculty informal interaction (N=528).

Concerning the relationship between gender, academic major, ethnicity and parents' level

of education, no significant correlations were found among measures of academic

performance, personal development and intellectual development. The study was

replicated in 1980 and conclusions were consistent (Terenzini and Pascarella, 1980).

However, there is evidence that student engagement may be related to student

demographics. For example, a female student may be more influenced by a woman

faculty member than by a male faculty member (Pascarella & Terenzini, 1991). Pike

(2004) measured engagement with the NSSE and found that females exhibit a higher

level of academic challenge and students of different disciplines have different levels of

engagement. Pike did not find students' age a correlate of student engagement when

examining students over that age of 24. However, as students progress in attaining their

degree, student engagement increases (Kuh, 2001). That is, as students' classification

moves from freshman to senior, students tend to be more engaged in learning.

Measurement of Engagement

The operational framework of this study places student engagement as a product

or end result of the interaction between student and faculty member behavior. Many









studies of learning styles use final grade or academic achievement as the dependent

variable in order to demonstrate the practical contribution that a specific learning style

has in the classroom. However, as undergraduate classrooms become more student

centered (Acharya, 2002; Lunde, Baker, Buelow & Hayes, 1995; Study Group on the

Conditions of Excellence in American Higher Education, 1984), there is greater need to

identify variables related to student engagement. Student engagement is often determined

with constructs of academic challenge, collaborative learning and student-faculty

interaction (Kuh, 2001). Each construct is dependent on the teaching practices of the

faculty member and requires communication between the faculty member and student

(Umbach & Wawrzynski, 2005). If a dissimilar learning style between the faculty

member and the student inhibits communication during the process of learning, evidence

should be found in the variance of student engagement scores. That is, a large cognitive

style gap may explain why students do not show evidence of engagement in the

classroom.

Also, if a faculty member uses subjective evaluations (rubrics) of student projects

and assignments-which are commonly used to evaluate the solution to a problem-the

final course grade is also subjective and biased. That is, a faculty member may prefer

solutions to assigned problems that are congruent with their learning style. The researcher

found no evidence in the literature concerning instructors' evaluation biases based on

learning style, but that does not suggest that this may exist.

Finally, the use of academic achievement as a dependant variable, whether in the

form of a percentile or letter grade assigned by the faculty member, may subject the study

to measurement error. There is evidence that faculty members inflate students grades for









the purpose of receiving tenure and promotion (Germain & Scandura, 2005).

Furthermore, student engagement is highly correlated with grade point average and other

measures of academic achievement (Kuh, 2001). Given the previous discussion, the

researcher found student engagement a more objective and accurate measure of learning

and interaction between the student and faculty member.

A remarkable increase in the study of student engagement in the undergraduate

classroom came after the release of Involvement in Learning (Study Group on the

Conditions of Excellence in American Higher Education, 1984), which called for the use

of teaching strategies that increase student learning in higher education. Since then,

researchers have sought to identify empirically based practices that impact learning and

develop effective measures to determine students' level of engagement. The National

Survey of Student Engagement (NSSE) was designed to assess student engagement

among colleges across the nation. Scales of the NSSE form "benchmarks" in order that

colleges can compare the teaching practices of their faculty members to the teaching

practices of other faculty members across the nation (Kuh, Hayek, Carini, Ouimet,

Gonyea & Kennedy, 2001). The instrument was found to be stable and psychometrically

solid with evidence of validity and reliability (Kuh, 2001). Additionally, the NSSE has

been positively correlated with undergraduate student grade point average, critical

thinking, quality of learning and GRE scores (Kuh).

The NSSE was developed through an extensive literature review of classroom

activities and teaching practices that lead to valuable student outcomes (Kuh, Hayek,

Carini, Ouimet, Gonyea & Kennedy, 2001). Three of the five benchmarks measured by

the NSSE were specific to learning and engagement in a classroom. They include









academic challenge, active learning and student-faculty interaction. Two other

benchmarks measured engagement relevant to the college, including enriching

educational experiences and supportive campus environment. As the latter two

benchmarks were relevant to the college and not the classroom, they were not utilized in

this study. Said differently, they were not measuring engagement practices directly

attributed to the student and the faculty member.

The NSSE benchmarks of interest are described as follows: level of academic

challenge-time spent preparing for class, amount of reading and writing, and use of

higher order thinking skills; active learning-participating in class and working

collaboratively with other students inside and outside the classroom and student-faculty

interaction-talking with the faculty member, discussing ideas outside of class with the

faculty member, getting feedback from the faculty member, and conducting research with

the faculty member (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001, p. 5). More

detail of the NSSE is provided in Chapter 3.

Summary of Learning and Engagement

Student's have a limited amount of time and effort to apply in accomplishing a

learning goal (Astin, 1984). However, faculty members may be the "single most

important" factor in engaging the student to learn. Although the student is responsible for

engaging in learning, the faculty member is accountable for facilitating student

engagement. The study of student engagement primarily entails identifying teaching

practices that increase student engagement and surveying students to determine if

teaching practices are being used.

Many researchers have examined the relationship learning styles have with

academic achievement. However, an emphasis on increasing student engagement (Lunde,









Baker, Buelow & Hayes, 1995) gave reason to examine if dissimilar learning styles

between faculty member and student can contribute to the literature regarding student

engagement. Furthermore, student engagement as measured by the NSSE may serve as a

more objective dependent variable than academic achievement. Student engagement is

closely tied to learning and academic achievement (Kuh, 2001).



Summary

Chapter 2 explains Kirton's Adaption-Innovation theory as applied to the

undergraduate classroom providing a theoretical framework for this study. The

framework was derived from Kirton's Adaption-Innovation theory which may provide

insight to how dissimilar cognitive styles explain student engagement. Other major

sections of this chapter included: Learning and Learning Styles, Learning and Stress,

Learning and Motivation, and Learning and Engagement.

This review of the literature found ties between problem solving learning and

Kirton's measure of cognitive style. Although Kirton's cognitive style has been applied

in many contexts including business and management of change, little research has been

conducted to examine if it is appropriate in explaining student engagement in the

undergraduate classroom.

There is a need for more research considering the variable stress as it relates to

learning and student engagement in the undergraduate classroom. Specifically, why do

more adaptive courses have higher levels of stress and how does this contribute to student

engagement? However, research has consistently found motivation to significantly

contribute to learning, yet few studies have examined both stress and motivation while

controlling for an instructor with a dissimilar learning style from that of students.






63


Student engagement in a course is largely dependent on teaching practices of the

faculty member. The amount of structure preferred by faculty members can be

determined by identifying their cognitive style. However, students may have different

cognitive styles than the faculty member and prefer either more or less structure in

learning from problem solving. As student engagement is closely tied to students'

academic success, it is critical to consider if dissimilar cognitive styles between students

and the faculty member can explain variance in student engagement.














CHAPTER 3
METHODS

Chapter 1 introduced the theory of learning styles, the appropriateness of Kirton's

Adaption-Innovation (A-I) theory for the classroom, and the need for increasing student

engagement in undergraduate classrooms. A point was also made that the literature

contains many types of learning styles with varying degrees of reliability and validity.

However, Kirton's (2003) measure of learning style is valid and reliable and distinct from

cognitive level. Finally, definitions of key terms, purpose of the study, and objectives of

the study were presented.

Chapter 2 presented the theoretical and conceptual frameworks of the study.

Included were the following topics with attention given to theory and empirical evidence:

a) learning and learning styles, b) learning and stress, c) learning and motivation, and d)

learning and student engagement. Theoretical evidence was provided that Kirton's (2003)

measure of learning style is compatible with problem solving learning (Gange, 1965) and

may provide insight to why some students lack classroom engagement.

This chapter provides information concerning the methodology of the study. More

specifically, details were given concerning the research design, target population,

instrumentation, data collection procedures, and data analysis procedures.

The purpose of this study was to determine if relationships exist between

cognitive style gap, student stress, student motivation, student engagement and selected

demographic variables of undergraduate students at the University of Florida in the

College of Agriculture and Life Sciences. The objectives of the study were to: 1)









Describe selected faculty and students according to their selected demographic variables;

2) Determine the cognitive style, student stress, student motivation, and student

engagement of undergraduate students; 3) Determine the cognitive style gap between

faculty and students and to explore its relationship with student stress, student motivation,

student engagement and specific demographic variables of undergraduate students; 4)

Explain undergraduate student stress and student motivation based on cognitive style gap

and selected student demographic variables and 5) Explain undergraduate student

engagement based on cognitive style gap, student stress, student motivation and selected

student demographic variables.

Research Design

The researcher used an ex post facto design one (Ary, Jacobs & Razavieh, 2002)

to accomplish the objectives of the study. This design allows for the control and measure

of the independent variable(s) to test hypotheses concerning variation in the dependent

variable(s). Cognitive style is inherent, stable, and measurable as an interval score along a

continuum allowing for the precise measure of dissimilarity between two individual's

cognitive styles. The literature gave little guidance concerning A-I theory's application to

the context of the undergraduate classroom concerning relationships with dependent

variables. Thus, this empirical design was deemed most appropriate.

For objective one, student demographic variables were measured with items in the

National Survey of Student Engagement 2005 (NSSE) (Kuh, Hayek, Carini, Ouimet,

Gonyea & Kennedy, 2001) and included: age, gender, major, college classification, full-

time status, and number of classes taken with similar content as the class. Lastly, students

were asked to give the number of problem sets that had taken more than an hour to

complete during a typical week. This question was not used in data analysis, but was used









to validate that problem solving does exist in the classes used in this study. Faculty

members were asked questions regarding age, gender, years of teaching experience, and

academic department. Faculty members also provided the researcher with syllabi and

additional information to describe the course. Finally, participating faculty members were

administered Kirton's Adaption-Innovation Inventory (KAI) (Kirton, 1999) to determine

their cognitive style. Descriptive statistics were used to report determined values.

For objective two, four instruments were administered to nine undergraduate

student classes to determine each student's cognitive style, stress, motivation and

engagement in their respective course. The KAI provided a measure of each student's

cognitive style. Level of student stress experienced in the classroom was determined

using the Student-life Stress Inventory (SSI) (Gadzella & Baloglu, 2001). The Motivated

Strategies for Learning Questionnaire (MSLQ) (Pintrich, Smith, Garcia & McKeachie,

1991) was administered to measure student motivation in the classroom. The NSSE was

used to determine student engagement. Descriptive statistics were used to report

determined values.

For objective three, cognitive style gap was calculated by subtracting faculty

members' cognitive style scores from each student's cognitive style scores as measured

by the KAI. The same procedure was used to determine gaps within the constructs of

cognitive style. These calculated values were used to find correlations with constructs of

stress, motivation, engagement and selected student demographic variables. Objective

three was accomplished using bivariate correlation.

For objective four, dependent variables including student stress, student

motivation, and student engagement were each separately explained by the independent









variables including cognitive style construct gaps and student demographic variables.

Objective four was accomplished using backward stepwise multiple linear regression

For objective five, independent variables including cognitive style construct gaps,

student motivation, student stress and selected student demographic variables were used

to explain the dependent variable student engagement. Objective five was accomplished

using backward stepwise multiple linear regression.

Population

This study utilized nine purposefully chosen class groups in the College of

Agriculture and Life Sciences at the University of Florida. Since direction and degree of

cognitive style gap was dependent on the cognitive style of the faculty member, courses

were selected based on a faculty member's cognitive style: a) three faculty members with

similar adaptive cognitive style scores, b) three faculty members with similar cognitive

style score approximating 95, and c) three faculty members with similar innovative

cognitive style scores.

Researchers have found little evidence that class size influences undergraduate

student achievement (Williams, Cook, Quinn & Jensen, 1985). However, Cuseo (2004)

found the optimum class size for student engagement at approximately 15 students with

no evidence provided in the literature to support a linear or asymptotic relationship. The

literature supports small class sizes for the retention of knowledge, student participation,

and critical thinking (McKeachie, Chism, Menges, Svinicke & Weinstein, 1994), but

there is no agreement on what number of students constitutes a small class. Most studies

consider 50 or more students as a large class (Cuseo, 2004). A sample size of 46 is

needed to attain a power of .70 (a=.05, R2=.20) with the use of four independent

variables (Ary, Jacobs & Razavieh, 2002). The researcher found from previous









experience of administering the KAI to students at the University of Florida that

mortality rate approximates 40%, which is lower than Kirton's (1999) estimated 20%

mortality rate for this type of population. The researcher determined a cognitive style

score as unusable using guidelines provided by Kirton (1999). Given the high mortality

rate of the KAI, the researcher sought out classes with nearly 65 students. However the

researcher found that many courses in the College of Agriculture and Life Sciences at the

University of Florida were either below 50 students or approximated 100 students.

Therefore the researcher chose classes that approximated 100 students realizing the

limitations of large class sizes having less student engagement (McKeachie et al., 1994).

Finally, instructional techniques employed by the faculty members often promote

student centered classrooms providing for student engagement (McKeachie et al.). For

this reason, a review of prospective course syllabi and assignments was conducted by the

researcher to purposefully select courses conducive for student engagement. Specifically,

the researcher examined these documents for evidence that courses required students to

attend class in order to be successful in the course. The researcher was not concerned

with a percentage of the final grade requiring attendance, but examined if assignments

and projects provided reason for students to attend class. Furthermore, since the KAI is a

measure of problem solving style, courses selected for this study required the use of

problem solving assignments. A problem was defined in this study as a desire or felt need

for a solution, but not immediately knowing the mental operations to arrive at the

solution (Soden, 1994).









Instrumentation

Four instruments were utilized to collect the data and complete the objectives of

the study. The following is a description of the instrument and its constructs, as well as

assessments of validity, and reliability.

Kirton Adaption-Innovation Inventory

The Kirton Adaption-Innovation Inventory (KAI) was developed by Kirton

(1976) as a measure of cognitive style unrelated to cognitive ability and cognitive affect.

Appendix D provides example items concerning the KAI and contact information to

acquire permission to use the instrument. The KAI is a self-reported instrument

consisting of 33 items, with one item not relevant in determining cognitive style. In

distinguishing more adaptive individuals from more innovative individuals, items

delineate a theoretical total range of 32 (highly adaptive) to 160 (highly innovative)

(Kirton, 1999). However, the actual total range is approximately 45 to 145 with a general

population mean of 95 (Kirton). For possible ranges of cognitive style construct scores,

Kirton reported the following: sufficiency of originality (17 to 63), efficiency (7 to 33)

and rule/group conformity (14 to 56).

Individuals taking the KAI were asked "How easy or difficult do you find it to

present yourself, consistently, over a long period as:" followed by a list of personal

characteristics. Individuals then mark an "X" on a continuum of "Very Hard" and "Very

Easy". Construct scores were calculated by summing respective items and total scores

were calculated by summing construct scores.

The KAI measured the cognitive style of the students and faculty members

through the constructs of sufficiency of originality, efficiency, and rule/group conformity.

Cognitive style gap was then calculated using the difference of total and construct









cognitive style scores of the faculty member from the student. Reliability of the KAI was

determined with Cronbach's alpha coefficient (Kirton, 1999). Validity of the KAI was

established through correlations with personality assessments (criterion validity), lack of

correlations with assessments of cognitive level (criterion validity), and factor analysis

(construct validity) (Kirton). The researcher is a certified practitioner of the KAI. See

Table 3-1 for construct and reliability information concerning the KAI.

Table 3-1. Construct Reliability of Kirton Adaption-Innovation Inventory
Construct No. of items Mean SD a
Sufficience of originality 13 41 9 .83
Efficiency 7 19 6 .76
Rule/Group conformity 12 35 9 .83
Total KAI 32 95 18 .88
Note. Source: (Kirton, 1999).

Student-life Stress Inventory

The second instrument administered to students was the Student-life Stress

Inventory (SSI) developed by Gadzella (1994). Appendix E provides example items

concerning the SSI and contact information to acquire permission to use the instrument.

The SSI is a self-reported instrument designed for college students and comprised of two

parts: perceived stress (5 scales, 23 items) and reactions to stress (4 scales, 28 items).

Items were in the form of statements of which participants responded to five-point Likert

scales with numbers signifying responses of "never", "seldom", "occasionally", "often",

and "most of the time". Construct scores of stressors were calculated by summing the

items per each scale and include frustrations, conflicts, pressures, changes, and self-

imposed. The total stress score had a possible range from 23 to 115 and possible ranges

of constructs were frustration (7 to 35), conflicts (3 to 15), pressures (4 to 20), changes (3









to 15) and self-imposed (6 to 30). Total stress scores were calculated by summing

constructs scores.

The researcher determined constructs related to reactions to stress including

physiological, emotional, behavioral, and cognitive appraisal, unimportant for this study

as they measured physical and psychological responses to stress and not level of stress.

Therefore, these were removed from the instrument prior to data collection. Furthermore,

item number six was removed from the frustration construct as it was not related to stress

between the faculty member and the student. Stems of the items were altered to include

the words "in this course" to facilitate the measurement of stress specific to the course of

which the instrument was administered.

The SSI was used with permission. The SSI was assessed for reliability with

Cronbach's alpha coefficients and validity through confirmatory factor analysis

(construct validity) (Gadzella & Baloglu, 2001). See Table 3-2 for descriptive

information regarding constructs and reliability of the SSI.

Table 3-2. Construct Reliability of Student-life Stress Inventory
Construct No. of items Mean SD a
tTotal Stressors 23 70.09 11.53 .92
Frustrations 7 17.65 3.87 .70
tConflicts 3 8.43 1.91 .68
tPressures 4 14.08 2.83 .80
tChanges 3 8.03 2.39 .86
tSelf-imposed 6 21.82 3.78 .63
Total Reactions to Stressors 28 67.19 13.32 .75
Physiological 14 31.07 8.32 .86
Emotional 4 12.17 3.51 .82
Behavioral 8 11.47 4.32 .71
Cognitive appraisal 2 6.01 1.98 .82
Total SSI 51 137.28 20.98 .92
Note. Source: (Gadzella & Baloglu, 2001). t Signifies construct was used for this study.









Motivated Strategies for Learning Questionnaire

The third instrument administered to students was the Motivated Strategies for

Learning Questionnaire (MSLQ) developed by Pintrich, Smith, Garcia, and McKeachie

(1991). Appendix F provides example items concerning the MSLQ and contact

information to acquire permission to use the instrument. The MSLQ is a self-reported

instrument consisting of two components: motivation orientation (6 scales, 31 items) and

use of learning strategies (9 scales, 50 items). Students taking the questionnaire are asked

to rate themselves on seven point Likert scale items anchored by "not at all true of me"

and "very true of me". Constructs of the MSLQ motivation orientation included scales of

value components, expectancy components, and affective components. The total possible

range of the MSLQ motivational orientation was 6 to 42 with all construct scores ranging

from 1 to 7. Scores were calculated by summing items related to the construct and

dividing by the number of items.

Learning strategy scales of the MSLQ include cognitive and meta-cognitive

strategies, and resource management strategies. The researcher determined that the

learning strategy scales were unnecessary for this study and were removed from the

instrument prior to data collection. The learning strategy scales were a measure of

cognitive engagement (Pintrich et al.) which is highly correlated with student

engagement. Furthermore, the researcher found the items biased towards being more

adaptive in respect to A-I theory.

The MSLQ was specifically designed to be administered to college students

enrolled in a college course and to be flexible to the needs of the researcher. The 15

scales of the MSLQ can be used either cumulatively or individually. It was developed

over nine years with the developers assessing validity through factor analysis (construct









validity) and correlations with other instruments and final grade scores (criterion validity)

(Pintrich et al., 1991). Internal reliability was assessed by the developers with

Chronbach's alpha coefficients (Pintrich et al.). The MSLQ was used with permission.

See Table 3-3 for information describing the constructs and reliability of the MSLQ.

Table 3-3. Construct Reliability of Motivated Strategies for Learning Questionnaire
Construct No. of items Mean SD a
tTotal Motivation Scale 31 30.46 *
Value
Intrinsic goal motivation 4 5.05 1.41 .74
tExtrinsic goal motivation 4 5.03 1.23 .62
tTask value 6 5.54 1.25 .90
Expectancy
tControl of learning beliefs 4 5.74 0.98 .68
tSelf-efficacy for learning 8 5.47 1.14 .93
Affective
tTest anxiety 5 3.63 1.45 .80

Total Learning Strategies Scale 50 39.13 *
Cognitive and meta-cognitive strat.
Rehearsal 4 4.53 1.35 .69
Elaboration 6 4.91 1.08 .76
Organization 4 4.14 1.33 .64
Critical thinking 5 4.16 1.28 .80
Meta-cognitive self regulation 12 4.54 0.90 .79
Resource management strategies
Time and study environment 8 4.87 1.05 .76
Effort regulation 4 5.25 1.10 .69
Peer learning 3 2.89 1.53 .76
Help seeking 4 3.84 1.23 .52
Total MSLQ 81 69.59 *
Note. Source: (Pintrich, Smith, Garcia & McKeachie, 1991). t Signifies scale was used
for this study. Signifies value was not reported.

National Survey of Student Engagement

The fourth instrument administered to the students was the National Survey of

Student Engagement 2005 (NSSE) developed by the National Center for Higher

Education Management Systems (NCHEMS) (Kuh, 2001). Appendix G provides

example items concerning the NSSE and contact information to acquire permission to use









the instrument. The NSSE is a self-reported instrument consisting of 85 items measuring

the constructs of college activities, educational and personal growth, opinions about your

school, educational program characteristics and other measures of academic performance.

An additional 15 items were designated to determine student demographic information.

However, only 42 of the 85 construct items form "benchmarks of effective educational

practice" (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001) from which academic

institutions calculate yearly progress towards increasing student engagement.

Benchmarks of the NSSE include academic challenge, active and collaborative learning,

student faculty interaction, enriching educational experiences, and supportive campus

environment.

Only three benchmarks were used as a measure of student engagement in this

study, including academic challenge, active and collaborative learning, and student

faculty interaction. The benchmarks of enriching educational experiences and supportive

campus environment were not applicable to this study as many items were relevant to the

campus and not the classroom. Stems of items were altered to include the words "in this

course" in order that students answer items specific only to the class the instrument was

administered. All items were four-point scale items except for one specific to the number

of hours spent on homework; it was a seven-point scale and changed to a four-point scale

to better determine the number of hours spent preparing specifically for the course used

in this study. Anchors of the scale typically resembled "very often", "often",

"sometimes", and "never", while other items were anchored with numbers representing

frequencies of completing a specific task.









Possible benchmark ranges included academic challenge (11 to 44), active and

collaborative learning (7 to 28) and student faculty interaction (6 to 24). Benchmark

scores were calculated by summing items related to each construct. In this study these

three benchmark scales were added to form a total possible engagement range of 24 to

96. However, since two NSSE benchmark scales were removed in this study, this total

score range does not correspond with the total range reported by the NSSE.

Validity of the NSSE was assessed by the NSSE design team through well-

defined items (face validity), research of the literature (content validity), factor analysis

(construct validity), and review of the collected data (concurrent validity) (Kuh, 2001).

Reliability was assessed utilizing Spearman's rho correlations between yearly tests (Kuh,

Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). The NSSE was used with permission.

See Table 3-4 for a description of constructs and reliability of the NSSE.

Table 3-4. Construct Reliability of National Survey of Student Engagement
Construct No. of items Mean SD Spearman's rho
tAcademic challenge 11 33.18 .77
tActive and collaborative learning 7 17.92 .76
tStudent-faculty interaction 6 12.64 .83
Enriching educational experiences 12 15.57 .89
Supportive campus environment 6 23.06 .85
Total NSSE 42 102.37 .83
Note. Source: (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Values reported
regarding the senior population of college students. t Signifies scale was used for this
study. Signifies value was not reported.

Data Collection Procedures

The Institutional Review Board (IRB) at the University of Florida conducted a

formal review of the research proposal for ethical soundness. The research proposal was

approved and assigned the IRB protocol number 2006-U-0078.









Before the administration of instruments to collect data, all participants received a

letter of informed consent which was also read by the researcher. The letter of informed

consent described the study, identified the researcher, addressed potential risks associated

in the participation of the study and approximate amount of time expected to complete

the instruments. Participants were informed of the voluntary nature of the study with the

opportunity to quit the study at any time with no consequences. Individuals who

volunteered to participate in the study signed the letter of informed consent as acceptance

of the terms and returned the letter of informed consent with the completed instruments

utilized in the study. Instruments were packaged together with corresponding codes to

administer all instruments at the same time and ensure each instrument corresponded with

the same student. Additionally, directions were given orally and in written form to

students to complete the SSI, MSLQ and NSSE in regards to the specific class the

students received the instrument and not to other classes enrolled in that semester. The

students were also informed that these directions did not apply to the KAI as it measures

their style of solving problems regardless of class.

Sixty-four faculty members in the College of Agriculture and Life Sciences at the

University of Florida were contacted during February 2006 to participate in this study.

Faculty members were identified by contacting individuals serving as undergraduate

coordinator of each department in the college. Undergraduate coordinators provided the

researcher with contact information of faculty members teaching courses that used

problem solving assignments and had enough students enrolled in the course to meet the

requirements of this study. Although many times the identified faculty members believed

their course did not match the requirements of the study, they often provided contact









information of another faculty member within the same department whose course better

matched the requirements of this study. In the end, 21 faculty members were willing to

participate in the study, but only 15 faculty members were administered the KAI as their

courses were deemed suitable by the researcher. From this group, nine faculty members

were chosen based on their cognitive style score. The selection of faculty members

provided a range along the adaptive-innovative continuum as three faculty members had

similar adaptive cognitive style scores, three faculty members had similar innovative

cognitive style scores and three faculty members had similar cognitive style scores

located in the middle of the adaptive-innovative continuum. Kirton (2003) states that total

cognitive style scores less than 10 points are considered the same, between 10 and 19

points are considered similar and scores of 20 points or more are considered different.

Each of the nine faculty members' courses was coded with a letter to maintain

confidentiality in data presentation. In order of adaptiveness to innovativeness faculty

members were coded beginning with the letter "A" and ending with the letter "I". The

grouping of classes by faculty member cognitive style resulted in an adaptive teacher

group (Classes A, B and C), a middle score teacher group (Classes D, E and F) and an

innovative teacher group (Classes G, H and I).

Student data were collected from each faculty members' class after midterm

examinations during the spring 2006 semester, specifically March 27 to April 12, 2006.

This time of administration was selected to ensure students had opportunity to interact

with faculty members in the classroom as the interaction of dissimilar cognitive styles

causes stress (Kirton, 2003). Students were administered the instruments during regularly

scheduled class times, however, it was the faculty member's choice if the instruments









were to be completed during class time. If not completed during class time, students

completed the instruments outside of class and brought them back during the next

scheduled class time. Specifically, students enrolled in Classes C, D, G and H were

allowed to complete the instruments during regularly scheduled class time while students

enrolled in Classes A, B, E, F and I took the questionnaires out of the classroom and

brought them back completed the next regularly scheduled class meeting time.

Faculty members were given the choice to allow students to complete instruments

during class time or outside of class time out of respect for the faculty members'

instructional time with students. The estimated time for students to complete all four

instruments was 40 minutes. Some faculty members were reluctant to give up this amount

of time and would not have participated if instruments were required to be completed

during class time. Authors of the instruments used in this study gave no indication that

students must complete the instruments in the classroom.

Also, it was the faculty member's choice to offer an incentive of extra-credit to

encourage student participation. This was done primarily to encourage students to come

to class on the day instruments were administered and to prompt return of the instruments

if students completed instruments outside of class. Only students enrolled in Class E were

not given extra-credit for participating in this study, which may have influenced response

rate.

A limitation of this study was the lack of control for method of administering

instruments. Method of instrument administration is confounded with the faculty member

giving no way to defend if either completion of instruments or use of extra-credit affected

students' scores concerning cognitive style, stress, motivation or engagement.









Furthermore, differences of scores between classes were expected and supported by

authors of the instruments. For example, Kirton (2003, p. 75) provides evidence that self-

selected courses may be more adaptive or innovative depending on the nature of the

course. If students in each course were disproportional regarding the number of young

women, there may be higher levels of stress (Gadzella and Guthrie, 1993). Authors of the

MSLQ (Pintrich, Smith, Garcia, McKeachie, 1991) claim that "responses to the questions

might vary as a function of different courses" (p. 5). Finally, student engagement is an

outcome of teaching practices and may vary with individual differences found among

faculty members (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Given the

above reasons, the data were not analyzed to examine the effect of different types of

instrument administration.

Data Analysis Procedures

Data were analyzed according to the following objectives and presented in

chapter 4 with respect to study objectives and classes.

Research Objective One

Research objective one was to describe selected faculty and students according to

their selected demographic variables. Student participants were asked demographic

questions found on the NSSE including age, gender, college classification, full-time

status, number of similar courses taken and major. Student participants were also asked

how many problem sets were assigned during a typical week that took more than an hour

to complete. Faculty participants were asked to respond to demographic items including

age, gender, and years of teaching experience. To achieve objective one, data were

entered into SPSS statistical software for Windows and analyzed using descriptive

statistics (frequencies and measures of central tendency).









Research Objective Two

Research objective two was to determine student cognitive style, student stress,

student motivation, and student engagement. Student participants were asked to respond

to the four self-reported instruments: the KAI, the SSI, the MSLQ and the NSSE. All

instruments were hand-scored by the researcher. To meet objective two, data were

entered into SPSS statistical software for Windows. The researcher used descriptive

statistics (frequencies and measures of central tendency) to analyze the data. For further

analysis, tests of significance were used to compare groups.

Research Objective Three

Research objective three was to determine the cognitive style gap between faculty

and students and explore relationships with undergraduate student stress, motivation,

engagement and selected student demographic variables. Cognitive style gap was found

by subtracting each faculty member's total cognitive style score from each of their

student's total cognitive style score. Specific constructs of cognitive style gap scores

were calculated in the same manner. Therefore, cognitive style gap described students in

terms of being more adaptive or more innovative in relation to their faculty member.

Utilizing SPSS statistical software, the researcher used Pearson's correlation coefficient

to determine the strength and direction of the relationship between cognitive style gap

with student stress, student motivation, student engagement and student demographic

variables identified in objective one.

Research Objective Four

Research objective four was to explain both undergraduate student stress and

student motivation based on cognitive style gap and student demographic variables

identified in objective one. Using SPSS statistical software, backward stepwise multiple









linear regression was used twice for each class to achieve objective four. Data were

analyzed first to determine relationships between constructs of cognitive style gap

(independent variable), student demographics (independent variable) and student stress

(dependent variable); and second, to determine relationships of the same independent

variables with student motivation (dependent variable). For further analysis, tests of

significance were used to compare groups.

Research Objective Five

Research objective five was to explain undergraduate student engagement based

on cognitive style gap, student stress, student motivation and selected student

demographic variables identified in objective one. Employing SPSS statistical software,

backward stepwise multiple linear regression was used to explain the dependent variable

student engagement based on independent variables including cognitive style gap, student

stress, student motivation and student demographic variables identified in objective one.

For further analysis, tests of significance were used to compare groups.

Summary

This chapter provided details of the methodology of this study with regards to

research design, population, instrumentation, data collection, and data analysis. The study

utilized an ex post facto Design 1 methodology to study the relationships of Kirton's

cognitive gap with stress, motivation, engagement and selected student demographic

variables. The population consisted of undergraduate students enrolled in the College of

Agricultural and Life Sciences at the University of Florida. To ensure variance of

students' cognitive style gap, both in size and direction, faculty members were selected

based on their adaptiveness and innovativeness of cognitive style. From this group of









faculty members, nine courses were purposefully selected to meet the objectives of this

study.

Four instruments were utilized to collect data from the participants. They were the

following: Kirton Adaption-Innovation Inventory (KAI) (Kirton, 1976), National Survey

of Student Engagement 2005 (NSSE) with demographic questions (Kuh, 2001),

Motivated Strategies for Learning Questionnaire (MSLQ) (Pintrich, Smith, Garcia, and

McKeachie, 1991), and Student-life Stress Inventory (SSI) (Gadzella, 1994). All

instruments were assessed for validity and reliability prior to conducting the study. Post-

hoc reliabilities for this study were presented in chapter 4.

In chapter 4, data were analyzed per class and organized by objectives. This study

primarily used descriptive statistics, correlation statistics, and backward stepwise

multiple linear regression.




Full Text

PAGE 1

DISSIMILAR COGNITIVE STYLES AND THEIR RELATIONSHIPS WITH UNDERGRADUATE STUDENT STRESS, MOTIVATION, AND ENGAGEMENT By CURTIS ROBERT FRIEDEL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Curtis Robert Friedel

PAGE 3

This document is dedicated to my parents, Larry and Susan Friedel. Thank you for everything.

PAGE 4

iv ACKNOWLEDGMENTS There are many people to thank who have pushed and pulled me to the creation of this document. First, I would like to thank Dr. Rick Rudd for his advising, encouragement, and friendship, all of which ga ve reason for me to start and complete my Ph.D. My advisory committee members, Drs. Gary Fairchild, Howard Ladewig, and Ed Osborne, all contributed in th eir own way by adding their expertise. This document is a reflection of the scholarship and rigor imposed by these three gentlemen. Also, I would like to thank Dr. Michael Kirton for his wis dom and counsel regarding cognitive style. To say that my Ph.D. in ag ricultural education was earned in my three years at the University of Florida would be an exagge ration. I owe many thanks to my undergraduate advisor, Dr. Marvin Hoskey, a master teac her. Likewise, I appreciate the teachers and administrators at Laurel-Concord Public Sc hools who served as models of excellent teaching and learning early in my career. I cannot say thank you enough to my parent s for teaching me the value of hard work and never letting me forget what is im portant in life. I also must thank my wife, Beth, not only for patiently waiting for me to write this document, but for showing me her approach towards life which helps me become better each and every day. Last, but not least, a spec ial thank you goes to my fellow graduate students with whom I shared an office. The times we spent making trips for ice cream, discussing the unspeakable over lunch, spending time at the JonesÂ’s on weekends, and building camaraderie have all helped the time fly by. I have learned from all of them.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................xi LIST OF FIGURES.........................................................................................................xvi ABSTRACT....................................................................................................................xvi i CHAPTER 1 INTRODUCTION TO THE STUD Y AND PROBLEM STATEMENT....................1 Background of Study....................................................................................................1 Introduction to Learning Styles....................................................................................2 Introduction to KirtonÂ’s Theo ry of Adaption-Innovation.............................................7 The Need for Student Engagement...............................................................................9 Problem Statement......................................................................................................10 Significance................................................................................................................12 Purpose of the Study...................................................................................................13 Assumptions...............................................................................................................14 Limitations..................................................................................................................14 Definitions..................................................................................................................15 Summary.....................................................................................................................18 2 REVIEW OF LITERATURE.....................................................................................19 Theoretical Framework...............................................................................................20 Learning and Learning Styles.....................................................................................29 Learning Style and Learning Process..................................................................33 Learning Style and Learning Level.....................................................................35 KirtonÂ’s Cognitive Style......................................................................................36 The Scale of Adaption and Innovation................................................................37 Summary of Learning and Learning Styles.........................................................39 Learning and Stress.....................................................................................................40 Stress in the Classroom.......................................................................................41 Demographic Variables Related to Stress...........................................................43 Measurement of Stress........................................................................................44 Summary of Learning and Stress........................................................................45

PAGE 6

vi Learning and Motivation............................................................................................46 Expectancy-Value Theory...................................................................................47 The expectancy component..........................................................................48 The value component...................................................................................49 Affective motivation....................................................................................51 Demographic Variables Related to Motivation...................................................52 Measurement of Motivation................................................................................52 Summary of Learning and Motivation................................................................54 Learning and Engagement..........................................................................................55 The Influence of Faculty Members.....................................................................56 Demographic variables related to engagement....................................................58 Measurement of Engagement..............................................................................58 Summary of Learning and Engagement..............................................................61 Summary.....................................................................................................................62 3 METHODS.................................................................................................................64 Research Design.........................................................................................................65 Population...................................................................................................................67 Instrumentation...........................................................................................................69 Kirton Adaption-Innovation Inventory................................................................69 Student-life Stress Inventory...............................................................................70 Motivated Strategies for Learning Questionnaire...............................................72 National Survey of Student Engagement............................................................73 Data Collection Procedures........................................................................................75 Data Analysis Procedures...........................................................................................79 Research Objective One......................................................................................79 Research Objective Two......................................................................................80 Research Objective Three....................................................................................80 Research Objective Four.....................................................................................80 Research Objective Five......................................................................................81 Summary.....................................................................................................................81 4 RESULTS...................................................................................................................83 Instrument Post-hoc Reliability..................................................................................84 Objective One.............................................................................................................85 Class A.................................................................................................................87 Class B.................................................................................................................88 Class C.................................................................................................................90 Class D.................................................................................................................91 Class E.................................................................................................................93 Class F.................................................................................................................94 Class G.................................................................................................................96 Class H.................................................................................................................97 Class I..................................................................................................................99 All Students.......................................................................................................101

PAGE 7

vii Summary of Findings for Objective One..........................................................103 Objective Two..........................................................................................................105 Class A...............................................................................................................105 Class B...............................................................................................................109 Class C...............................................................................................................112 Class D...............................................................................................................116 Class E...............................................................................................................119 Class F...............................................................................................................123 Class G...............................................................................................................127 Class H...............................................................................................................130 Class I................................................................................................................134 All Students.......................................................................................................137 Comparisons Between Classes..........................................................................141 Summary of Findings for Objective Two..........................................................143 Objective Three........................................................................................................147 Class A...............................................................................................................148 Class B...............................................................................................................155 Class C...............................................................................................................161 Class D...............................................................................................................167 Class E...............................................................................................................173 Class F...............................................................................................................178 Class G...............................................................................................................184 Class H...............................................................................................................189 Class I................................................................................................................195 All Students.......................................................................................................200 Group Comparisons...........................................................................................206 Summary of Findings for Objective Three........................................................208 Objective Four..........................................................................................................214 Class A...............................................................................................................216 Class B...............................................................................................................218 Class C...............................................................................................................218 Class D...............................................................................................................220 Class E...............................................................................................................221 Class F...............................................................................................................223 Class G...............................................................................................................224 Class H...............................................................................................................224 Class I................................................................................................................225 All Students.......................................................................................................227 Summary of Findings for Objective Four.........................................................229 Objective Five...........................................................................................................231 Class A...............................................................................................................232 Class B...............................................................................................................233 Class C...............................................................................................................234 Class D...............................................................................................................235 Class E...............................................................................................................236 Class F...............................................................................................................237

PAGE 8

viii Class G...............................................................................................................237 Class H...............................................................................................................238 Class I................................................................................................................239 All Students.......................................................................................................240 Summary of Findings for Objective Five..........................................................242 Summary...................................................................................................................243 5 SUMMARY AND DISCUSSION...........................................................................245 Summary of Problem Statement...............................................................................245 Purpose and Objectives.............................................................................................246 Methodology.............................................................................................................247 Conclusions and Discussion.....................................................................................249 Objective One....................................................................................................250 Class A.......................................................................................................250 Class B........................................................................................................251 Class C........................................................................................................251 Class D.......................................................................................................252 Class E........................................................................................................252 Class F........................................................................................................253 Class G.......................................................................................................253 Class H.......................................................................................................254 Class I.........................................................................................................254 All Participants...........................................................................................255 Discussion for objective one......................................................................255 Objective Two...................................................................................................257 Class A.......................................................................................................257 Class B........................................................................................................258 Class C........................................................................................................258 Class D.......................................................................................................258 Class E........................................................................................................259 Class F........................................................................................................259 Class G.......................................................................................................260 Class H.......................................................................................................260 Class I.........................................................................................................261 All students.................................................................................................261 Comparison between classes......................................................................262 Discussion for objective two......................................................................263 Objective Three.................................................................................................267 Class A.......................................................................................................268 Class B........................................................................................................269 Class C........................................................................................................269 Class D.......................................................................................................270 Class E........................................................................................................271 Class F........................................................................................................271 Class G.......................................................................................................272 Class H.......................................................................................................272

PAGE 9

ix Class I.........................................................................................................273 All Students................................................................................................274 Group Comparisons....................................................................................274 Discussion for objective three....................................................................276 Objective Four...................................................................................................280 Class A.......................................................................................................280 Class B........................................................................................................281 Class C........................................................................................................281 Class D.......................................................................................................282 Class E........................................................................................................282 Class F........................................................................................................282 Class G.......................................................................................................283 Class H.......................................................................................................283 Class I.........................................................................................................284 All Students................................................................................................284 Discussion for objective four.....................................................................285 Objective Five...................................................................................................287 Class A.......................................................................................................288 Class B........................................................................................................288 Class C........................................................................................................289 Class D.......................................................................................................289 Class E........................................................................................................289 Class F........................................................................................................290 Class G.......................................................................................................290 Class H.......................................................................................................290 Class I.........................................................................................................291 All Students................................................................................................291 Discussion for objective five......................................................................291 Concluding Discussion.............................................................................................293 Recommendations.....................................................................................................298 Summary...................................................................................................................300 APPENDIX A INTERNAL REVIEW BOARD SUBMISSION FORM.........................................301 B FACULTY INFORMED CONSENT.......................................................................304 C STUDENT INFORMED CONSENT.......................................................................306 D EXAMPLE ITEMS OF KIRT ONÂ’S ADAPTION-INNOVATION INVENTORY...........................................................................................................308 E EXAMPLE ITEMS OF STUDENT -LIFE STRESS INVENTORY........................309 F EXAMPLE ITEMS OF MOTIVATED STRATEGIES FOR LEARNING QUESTIONNAIRE..................................................................................................310

PAGE 10

x G EXAMPLE ITEMS OF NATIONAL SURVEY OF STUDENT ENGAGEMENT ......................................................................................................311 LIST OF REFERENCES.................................................................................................312 BIOGRAPHICAL SKETCH...........................................................................................325

PAGE 11

xi LIST OF TABLES Table page 2-1 Factor Traits of Adaptors and Innovators................................................................36 3-1 Construct Reliability of Ki rton Adaption-Innovation Inventory..............................70 3-2 Construct Reliability of Student-life Stress Inventory.............................................71 3-3 Construct Reliability of Motivated Strategies for Learning Questionnaire.............73 3-4 Construct Reliability of Nationa l Survey of Student Engagement..........................75 4-1 Listing of Classes designated by Faculty MembersÂ’ Total Cognitive Style Score (N=9)........................................................................................................................86 4-2 Summary of Classes Regarding Re ported Demographic Information of Participating Faculty Members..............................................................................103 4-3 Summary of Classes Regarding Re ported Demographic Information of Participating Students.............................................................................................104 4-4 Class A Student Mean Scores of Cognitive Style Constructs (n=58)....................106 4-5 Class A Student Mean Scores of Stress Constructs...............................................107 4-6 Class A Student Mean Scor es of Motivation Constructs.......................................108 4-7 Class A Student Mean Scores of Engagement Constructs (n=70).........................109 4-8 Class B Student Mean Scores of Cognitive Style Constructs (n=46)....................109 4-9 Class B Student Mean Scores of Stress Constructs (n=68)....................................110 4-10 Class B Student Mean Scor es of Motivation Constructs.......................................111 4-11 Class B Student Mean Scores of Engagement Constructs (n=70).........................112 4-12 Class C Student Mean Scores of Cognitive Style Constructs (n=56)....................113 4-13 Class C Student Mean Scores of Stress Constructs................................................114

PAGE 12

xii 4-14 Class C Student Mean Scores of Motivation Constructs (n=71)...........................115 4-15 Class C Student Mean Scores of Engagement Constructs (n=71).........................115 4-16 Class D Student Mean Scores of Cognitive Style Constructs (n=73)....................116 4-17 Class D Student Mean Scores of Stress Constructs...............................................117 4-18 Class D Student Mean Scor es of Motivation Constructs.......................................118 4-19 Class D Student Mean Scores of Engagement Constructs (n=108).......................119 4-20 Class E Student Mean Scores of Cognitive Style Constructs (n=32)....................120 4-21 Class E Student Mean Scores of Stress Constructs................................................121 4-22 Class E Student Mean Scores of Motivation Constructs (n=48)............................122 4-23 Class E Student Mean Scores of Engagement Constructs (n=48).........................123 4-24 Class F Student Mean Scores of Cognitive Style Constructs (n=71).....................124 4-25 Class F Student Mean Scores of Stress Constructs (n=112)..................................125 4-26 Class F Student Mean Scores of Motivation Constructs (n=115)..........................126 4-27 Class F Student Mean Scores of Engagement Constructs (n=115)........................127 4-28 Class G Student Mean Scores of Cognitive Style Constructs (n=65)....................127 4-29 Class G Student Mean Scores of Stress Constructs (n=84)...................................128 4-30 Class G Student Mean Scores of Motivation Constructs (n=85)...........................129 4-31 Class G Student Mean Scores of Engagement Constructs (n=85).........................130 4-32 Class H Student Mean Scores of Cognitive Style Constructs (n=50)....................131 4-33 Class H Student Mean Scores of Stress Constructs (n=70)...................................132 4-34 Class H Student Mean Scores of Motivation Constructs (n=70)...........................132 4-35 Class H Student Mean Scores of Engagement Constructs.....................................133 4-36 Class I Student Mean Scores of Cognitive Style Constructs (n=60)......................134 4-37 Class I Student Mean Scores of Stress Constructs (n=77).....................................135 4-38 Class I Student Mean Scores of Motivation Constructs (n=77).............................136

PAGE 13

xiii 4-39 Class I Student Mean Scores of Engagement Constructs......................................137 4-40 All Student Mean Scores of C ognitive Style Constructs (n=511).........................138 4-41 All Student Mean Scor es of Stress Constructs.......................................................139 4-42 All Student Mean Scores of Motivation Constructs...............................................140 4-43 All Student Mean Scores of Engagement Constructs............................................140 4-44 Summary of Classes Regarding Studen t Cognitive Style, Stress, Motivation and Engagement Mean Scores......................................................................................147 4-45 Class A Student Mean Scores of Cognitive Style Construct Gaps (n=58)............149 4-46 Class A Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=56)........................153 4-47 Class B Student Mean Scores of Cognitive Style Construct Gaps (n=46).............155 4-48 Class B Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=44)........................160 4-49 Class C Student Mean Scores of Cognitive Style Construct Gaps (n=56).............162 4-50 Class C Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=56)........................166 4-51 Class D Student Mean Scores of Cognitive Style Construct Gaps (n=73)............168 4-52 Class D Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=67)........................172 4-53 Class E Student Mean Scores of Cognitive Style Construct Gaps (n=32).............174 4-54 Class E Correlations, Means, and St andard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=31)........................177 4-55 Class F Student Mean Scores of Cognitive Style Construct Gaps (n=71).............179 4-56 Class F Correlations, Means, and St andard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=69)........................182 4-57 Class G Student Mean Scores of Cognitive Style Construct Gaps (n=65)............184 4-58 Class G Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=64)........................188 4-59 Class H Student Mean Scores of Cognitive Style Construct Gaps (n=50)............190

PAGE 14

xiv 4-60 Class H Correlations, Means, and Sta ndard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=50)........................194 4-61 Class I Student Mean Scores of Cognitive Style Construct Gaps (n=60)..............196 4-62 Class I Correlations, Means, and St andard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=59)........................199 4-63 All Student Class Mean Scores of Cognitive Style Construct Gaps (n=511)........201 4-64 All Student Class Correlations, Me ans, and Standard Deviations among Constructs of Cognitive Style Gap, St ress, Motivation and Engagement (n=496)...................................................................................................................205 4-65 Class A Backward Stepwise Multiple Regression Explaining Student Total Stress (n=57)..........................................................................................................216 4-66 Class A Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=57)..................................................................................................217 4-67 Class B Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=45)..................................................................................................218 4-68 Class C Backward Stepwise Multiple Regression Explaining Student Total Stress (n=56)..........................................................................................................219 4-69 Class C Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=56)..................................................................................................220 4-70 Class D Backward Stepwise Multiple Regression Explaining Student Total Stress (n=66)..........................................................................................................221 4-71 Class E Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=32)..................................................................................................222 4-72 Class F Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=71)..................................................................................................223 4-73 Class H Backward Stepwise Multiple Regression Explaining Student Total Stress (n=49)..........................................................................................................225 4-74 Class I Backward Stepwise Multiple Regression Explaining Student Total Stress (n=60).....................................................................................................................226 4-75 Class I Backward Stepwise Multip le Regression Explai ning Student Total Motivation (n=60)..................................................................................................227 4-76 All Students Backward Stepwise Mul tiple Regression Expl aining Student Total Stress (n=496)........................................................................................................228

PAGE 15

xv 4-77 All Students Backward Stepwise Mul tiple Regression Expl aining Student Total Motivation (n=502)................................................................................................229 4-78 Class A Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=56)................................................................................................233 4-79 Class B Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=45)................................................................................................234 4-80 Class C Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=56)................................................................................................235 4-81 Class D Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=66)................................................................................................236 4-82 Class E Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=31)................................................................................................237 4-83 Class G Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=63)................................................................................................238 4-84 Class H Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=49)................................................................................................239 4-85 Class I Backward Stepwise Multip le Regression Explai ning Student Total Engagement (n=59)................................................................................................240 4-86 All Student Backward Stepwise Mul tiple Regression Expl aining Student Total Engagement (n=493)..............................................................................................241

PAGE 16

xvi LIST OF FIGURES Figure page 2-1 KirtonÂ’s theoretical model of Ad aption and Innovation applied to the undergraduate classroom..........................................................................................24 2-2 Interaction of cognitive style gap, stress, motivation and engagement....................27

PAGE 17

xvii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DISSIMILAR COGNITVE STYLES AN D THEIR RELATIONSHIPS WITH UNDERGRADUATE STUDENT STRESS, MOTIVATION, AND ENGAGEMENT By Curtis R. Friedel December, 2006 Chair: Rick Rudd Major Department: Agricultur al Education and Communication Few studies have examined dissimilar cognitive style between faculty members and students and its relationshi ps with student stress, motivation and engagement. In the search for variables explaining student e ngagement, researchers have examined the cognitive level of instructional discourse, but few have examined the cognitive style of the instructor as it corresponds to the cogni tive style of the student. Identifying how dissimilar learning styles relate to student stress and motivation ma y have implications for increasing student engagement. KirtonÂ’s Adaption-Innovation theory provide d the theoretical framework for this study. An ex post facto design was used to accomplish the objectives of this study. Nine undergraduate courses in the College of Agricu lture and Life Sciences at the University of Florida were administered four sta ndardized instruments: KirtonÂ’s AdaptionInnovation Inventory, Student-life Stress Invent ory, Motivated Strategies for Learning Questionnaire, and National Survey of Stude nt Engagement. Dissimilar cognitive style

PAGE 18

xviii was measured by subtracting the faculty membersÂ’ cognitive style score from each individual studentÂ’s cognitive style score. The difference provided a cognitive style gap which was used to determine relationshi ps among student stress, motivation and engagement in the classroom. Findings indicated that in these course s, students with more than a 20-point cognitive style gap did not have higher levels of stress than students with less than a 20point cognitive style gap. However, student s with more than a 20-point cognitive style gap did have significantly lower levels of motivation. Student motivation was further decreased when students were taught by an adaptive faculty member. Backwards stepwise multiple regression was used to e xplain student stress, student motivation and student engagement based on cognitive style gap and selected demographic variables. Results provided evidence that cognitive style gap did contribute to the explanation of student stress, motivation and engagement in some classes. However, cognitive style gap was not significant in explaining these depe ndent variables in all nine classes. Further research should employ qualitativ e methods to examine how cognitive style is exhibited in instruc tional discourse. Researchers may use these findings to further examine the relationships between learning st yle and learning proce ss in the effort of improving student engagement.

PAGE 19

1 CHAPTER 1 INTRODUCTION TO THE STUD Y AND PROBLEM STATEMENT Background of Study This study examined relationships between dissimilar cognitive styles (Kirton, 2003), stress, motivation and cl assroom engagement at the University of Florida. Educators have long known through pragmatic a nd empirical evidence that students learn in a variety of styles. Many assessments of learning styles and personality have been developed to determine how students interact with and process the learning environment. All learning style theories share in comm on the contention that learning improves in learning environments that match a studentÂ’s learning style (Coffi eld, Moseley, Hall and Ecclestone, 2004). Nevertheless, few studies have been conducted to assess the interaction between the lear ning style of the student a nd the learning style of the instructor; nor the positive and negative effects from this interaction. Cognitive style as defined by Kirton may offer insight into how a style of learning influences other variables in the college classroom. Using the Kirton Â’s Adaption-Innovation Inventory (KAI), the researcher determined the cognitive style of undergraduate students and university instructors in the college classroom. The interval difference be tween an individual studentÂ’s cognitive style and the instructorÂ’s cognitive style yielded a cognitive style gap score (Kirton). Individual c ognitive style gap scores were compared to other student variables including student stress, motivati on, engagement, and selected demographics.

PAGE 20

2 Introduction to Learning Styles A current trend in undergraduate educati on is a transition from teacher-centered classrooms to student-centered classrooms (Acharya, 2002). Although teachers are still accountable for student learning, great interest has been given in support of self-regulated learning (Winne, 1995) and st udent engagement (Kuh, 2003). Boekaerts (1999) stated that student understanding of lear ning styles is critical to in creasing successful learning. But, what is a learning style? To define learning style, the researcher reviewed an array of definitions, theories measurements and beliefs which included those written by Allinson and Hayes (1996), Apter (2001), Dunn (1984), Gregorc (1979), James and Gardner (1995), Keefe (1979), Kirton (1976), Kolb (1984), Sternberg (1988) and Witkin and Goodenough (1977). Throughout the literature the terms learning style and cognitive style have been used interchangeably with few authors making distinctions between the two concepts (Sadler-Smith, 2001). Additiona lly, authors have been congruent in defining both learning style and cognitive style. For example, Claxton and Ralston (1978, p. 7) define learning style as a “consistent way of responding to and using stimuli in the context of learning.” Messick (1984, p. 143) defines cognitive style as a “consistent individual differences in ways of organizing and processing information and experience.” Both definitions were deemed appropriate for use in this study with the two terms used synonymously throughout this document. In a literature review of learning styles, Coffield, Moseley, Hall and Ecclestone (2004) reported 71 different assessments of learning style. To cat egorize the various learning style theories, a continuum of lear ning style families was developed based on the degree of the learning style being fixed and innate. Anchoring one end of the continuum were strongly fixed and innate learning st yles possibly stemming from genetics. The

PAGE 21

3 other end of the continuum is anchored by learning styles which are influenced by situational factors such as th e environment, motivation, stress and learning strategies. From one through five, this continuum of l earning style families and their respective theorists included the following: 1. Learning styles and preferences are largely constitutionally based including the four modalities: visual, auditory, kinesthetic, and tactile. Learning style theorists in this family include Bartlett, Betts, Dunn a nd Dunn, Gordon, Gregorc, Marks, Paivio, Richardson, Sheehan and Torrance. 2. Learning styles reflect deep-seated featur es of the cognitive structure, including patterns of ability. Learning style theori sts in this family include Broverman, Cooper, Gardner et al., Guilford, Holzma n and Klein Hudson, Hunt, Kagan, Kogan, Messick, Pettigrew, Riding, and Witkin. 3. Learning styles are one component of a re latively stable person ality type. Learning style theorists in this family include Ap ter, Epstein and Meier, Harrison-Branson, Jackson, Miller, and Myers-Briggs. 4. Learning styles are stable learning preferen ces with a degree of flexibility. Learning style theorists in this family include Allinson and Hayes; Felder and Silverman; Herrmann; Hermanussen, Wierstra, de J ong, and Thijssen; Honey and Mumford; Kaufmann; Kirton; Kolb; and McCarthy. 5. Learning styles are approaches, strategies, orientations and con ceptions of learning. Learning style theorists in this family in clude Biggs; Conti and Kolody; Entwistle; Grasha-Riechmann; Hill; Marton and Slj; McKenney and Keen; Pask; Pintrich, Smith, Garcia and McCeachie; Schmeck; Sternberg; Vermunt; Weinstein, Zimmerman and Palmer; and Whetton and Cameron. (Coffield, Moseley, Hall & Ecclestone, 2004, p. 9) This simple categorization of learning styles ignores the depth of theory supporting learning, validity and reliability of the instrumentation, and classroom implications; however, the work done by Co ffield, Moseley, Hall and Ecclestone gives the best reference point for the discussion of strengths, weaknesses, and similarities of learning styles.

PAGE 22

4 Even within categories of learning styles there are disagreements. For example in category four, Kolb (1984) defines his measure of learning style as stable cognitive programs formed by past experiences of educa tion, choice of career, and the current task at hand. Kirton (2003) asserts his measure of learning style is base d on fixed and innate traits used in problem solving, but people may cope outside of their style. KolbÂ’s (1984) model of experiential learning re quires four learning styles to complete the experiential learning process. Kirton (2003) claims that the process of learning is independent of learning style (de Ciantis & Kirton, 1996). Additionally, Kolb (1984) claims that a preferred learning style will allow a person to be more competent in completing a particular task. Kirton (2003) argues that everyone has the capabi lity to solve all problems within their own style. That is, le arning style is distinct and independent of learning level or capacity. Leve l of learning can be defined as an individualÂ’s ability to learn and can be expressed as intelligence (Kirton, 2003). Th e literature is unclear in distinguishing an individualÂ’s st yle of learning and level of l earning, which is an assertion of KirtonÂ’s theory. This assertion is supporte d by the literature as Ericsson, Krampe and Tesch-Romer (1993) found that in all do mains of knowledge, the single greatest contribution to expertise was deliberate pract ice. In addition, no evidence was found that a particular learning style or individual difference played a role in acquiring expertise within a domain of knowledge. However, it ma y be possible that experts have a preferred style when producing an expert performance. Learning styles have been included as a variable in many educational studies. Still, research conducted in the area of learni ng styles has provided conflictive findings. For example, Smith, Sekar and Townsend ( 2002) reviewed studies in which students

PAGE 23

5 were given instructional discour se that matched their indivi dual learning style. Results were mixed with only nine of eighteen studi es supporting the contention that students learn best when taught in their preferred lear ning style. In another example, Wingenbach (2000) used the Group Embedded Figures Test (GEFT) (Witkin, Oltman, Raskin & Karp, 1971) and found undergraduates with field-indepe ndent learning styles had a moderately positive and significant relationship with student academic achievement. However, DeTure (2004) used the GEFT and found no evidence to significantly predict student academic achievement from learning style. Why do these findings conflict? If Kirton (2003) is correct, all students are capable of achieving academic succ ess within their own cognitive style. If academic achievement is dependent on the expectations of the faculty member, could dissimilar cognitive style betwee n students and the faculty member be an intervening variable? Kirton’s Adaption-Innovation (A-I) Theory ( 2003) identifies cognitive style as a preference to the amount of structure that one undertakes when solving problems. Within the continuum of learning style families (Co ffield et al., 2004), Kirton’s work resides in the fourth family noted above, “flexibly stab le learning preferences” (p.9). Some people are more adaptive when solving problems, and other people are more innovative when solving problems. Adaptiveness and innovativen ess anchor on continuums of how people approach problems, produce ideas, and adhere to structures. Thes e three constructs measure a fixed perception of an environment in which the individual interacts with and responds in the process of learning while solv ing a problem; a fit to this study’s definition of learning style (Claxton and Ralston, 1978) More detail will be given to Kirton’s cognitive style’s relationshi p to learning in chapter 2.

PAGE 24

6 Instructional material designed to in corporate learning styles may increase students’ motivation to learn, (Larkin-Hein, 2000); however, mo re research is warranted in this area. Few studies ha ve been conducted to measur e the impact of curriculum delivered in a learning style different than the student. Nor have studies been conducted to examine how instructional discourse of a di ssimilar learning style relates to a student’s stress and motivation in the classroom; an assertion of Kirton’s A-I theory. The primary focus of research and practi ce of learning style theory is aligning curriculum and instruction to meet the need s of diverse learning styles in a given classroom (Curry, 1990). However, much of th e literature fails to recognize the learning style of the instructor and how this variable plays in the context of the classroom. The effect of dissimilar learning styles of people working together to develop a new idea or product and the stress resulting from this interact ion is lacking in the literature (Smulders, 2004); yet the study of this phenomenon is th e aim of A-I theory (Kirton, 2003). Since students and teachers of differe nt learning styles commonly co me together in a classroom to generate ideas with varyi ng degrees of motivati ons, attitudes and stressors, it is clear that many research questions are still unanswered. Limitations have been placed on the a pplication of learning styles in the classroom. After an extensive review of the literature, Coffield, Moseley, Hall and Ecclestone (2004) cautioned practitioners’ appl ication of learning styles in the classroom as many types of style have little theoreti cal backing, low reliability and insufficient validity. Furthermore, few studies have found learning styles to “explain more than 16% of the variance in academic achievement” ( p. 127). Other authors have agreed that many measures of learning style are psychometrically weak and therefore offer little use to the

PAGE 25

7 practitioner (Curry, 1990; Harg reaves, 2005; Price, 2004; Smith 2002,). It is evident that future research must use psychometrically so lid measures of learning style with strong theoretical ties in order to provide recommendations to education practitioners. Introduction to KirtonÂ’s Theory of Adaption-Innovation In KirtonÂ’s (2003) A-I theo ry, a personÂ’s cognitive styl e can be determined as more adaptive versus more innovative along a continuum. Cognitive style is ultimately a measure of how individuals solve problems w ith indication of preferences for learning (de Ciantis & Kirton, 1996). Ki rton (2003) describes cogni tive style in context as individuals interacting and reacting to the environment in consistent and stable approaches when solving problems, which is co nsistent with definitions of learning style and cognitive style used in this study (Claxton & Ralston, 1978; Messick, 1984). A person may find their preferred style of pr oblem solving along the adaptive-innovative continuum by completing the Kirton Ad aption-Innovation Inventory (KAI). KirtonÂ’s (2003) measure of cognitive styl e along a continuum allows for interval scale distinction between dissi milar cognitive styles of two individuals. Additionally, dissimilarity of cognitive style can be determ ined between two innovative individuals or two adaptive individuals. The calculated diffe rence between two indi vidualsÂ’ cognitive style scores yields a cognitive style gap (Kir ton) which determines the dissimilarity of two approaches to problem solving. Cognitive style gap was calculated in this study by subtracting faculty membersÂ’ cognitive style score from each studentÂ’s cognitive style score. A cognitive style gap larger than 20 points may contribute to inefficiencies in communication (Kirton, 2003). Stress is a resu lt of cognitive gap (Kirton) which may present important implications to learning. Th e literature suggests that stress in the form

PAGE 26

8 of anxiety may affect studentsÂ’ self-esteem and academic achievement (Pintrich & Schunk, 2002). Fortunately, individuals can be motivated to operate outside of a preferred cognitive style and bridge the c ognitive gap found with another individual (Kirton, 2003). Kirton uses the term coping be havior to describe bridging a cognitive style gap and further ela borates that motivation in itiates coping behavior. It is necessary to note that KirtonÂ’s meas ure of cognitive style is distinct from motivation. However, authors of other learni ng style instruments integrate aspects of motivation and cognitive style into one meas ure of how individuals approach learning (Apter, 2001; Dunn, 1984; Entwistle, 1978; Herrmann, 1989; Vermunt, 1992). The note was not made to argue for or against motivatio nal factors as a determ inant of a preferred style of learning. Rather the pure and distinct measure of learning style separated from motivation allows for a more precise measure of motivation (Kirton, 2003), which in turn permits the study of motivation as it relates to dissimilar learning styl es and to learning. The literature provides evidence that stude nt motivation is necessary for learning (Schunk, 1991) and student engagement (M cKeachie, Pintrich, Lin & Smith, 1986) making KirtonÂ’s measure of cognitive style particularly valuable. KirtonÂ’s A-I theory provides a theore tical framework to incorporate the interaction of dissimilar learning styles in the production of new products and ideas. Adaption-Innovation theory was chosen for this study, because of its ability to separate cognitive style from cognitive level while gi ving attention to motivation, stress, and learning. Distinction between cognitive style and cognitive level ma y provide insight to the literature which asserts a confused meshing of the two in explaining student achievement. A reference point is also esta blished to examine th e relationships of

PAGE 27

9 variables found in the educational classroom such as student engagement, motivation and stress. The KAI has been determined as ps ychometrically reliabl e and valid (Kirton, 1999); as opposed to many other learning styl e assessments (Coffield et al., 2004; Curry, 1990). Given the above discussion of A-I theo ry, the researcher used Kirton (2003) to operationalize learning style. Chapter 2 furt her elaborates on KirtonÂ’s A-I theory and measure of cognitive style. The Need for Student Engagement Involvement in Learning (Study Group on the Conditions of Excellence in American Higher Education, 1984), a report prop osing the reform of higher education, stated a quality education wa s more dependent on student engagement and less on the reputation and resources of the academic institution. Suddenly faculty members of colleges and universities were asked to be accountable for undergraduate student achievement. To assist faculty members with their new challenge, Chickering and Gameson (1987) developed a set of e ngagement indicators known as the Seven Principles of Good Practice in Undergraduate Education They include: 1. Encourages contact between students and faculty, 2. Develops reciprocity and cooperation among students, 3. Encourages active learning, 4. Gives prompt feedback, 5. Emphasizes time on task, 6. Communicates high ex pectations, and 7. Respects diverse talents and ways of learning. (p. 5) The Seven Principles of Good Practice in Undergraduate Education are widely accepted by the American Association of Hi gher Education (AAHE) as research based process standards that can be applied a nd measured in undergraduate classrooms. In a longitudinal, multi-institutional study, Astin (1993) used 82 student outcome measures, 140 student characteristics, 190 envi ronmental characterist ics and 57 types of

PAGE 28

10 student involvement and found that academ ic involvement, faculty involvement and student peer group involvement were the larges t indicators of enha ncing student learning and personal development of undergraduate st udents. Koljatic and Kuh (2001) conducted a study to determine if cooperation among stude nts, active learning, and faculty contact with students has improved since 1983 and f ound the degree of student engagement in these three areas has not improved. Furtherm ore, 20% of freshman and seniors in colleges and universities across the nation were identified as disengaged (Kuh, 2003). Even though the literature ha s addressed the need for stude nt engagement, it seems that faculty membersÂ’ instructiona l techniques still do not enc ourage undergraduate studentsÂ’ classroom engagement. In colleges of agri culture, Whittington (1998) found almost 80% of faculty membersÂ’ instructional discourse to be low at cognitiv e level. Is student engagement explained by cognitive level, cogn itive style, or both? Does cognitive style gap between faculty members and students matter? Problem Statement Trends in undergraduate education are moving from teacher-centered classrooms to student-centered classrooms for the impr ovement of student engagement (Acharya, 2002). From classroom teaching experiences, mo st faculty members concur that students learn differently. This has increa sed the application of learning styles in the classroom as faculty members struggle to develop instruct ional techniques to engage student learning. Advocating the awarene ss of differing learning styles is commendable. However, further implications for education pr actitioners are limited as studi es examining instructional discourse provided to students in similar lear ning styles have conflicting results (Smith, Sekar & Townsend, 2002). Many learning styles measures are psychometrically weak (Curry, 1990) and offer small explanation to the variance of student achievement

PAGE 29

11 (Coffield et al., 2004). Simply stated, future research must be able to accurately determine studentsÂ’ learning styles and provi de theoretical and em pirical evidence that learning styles contribut e to student learning. The literature comprising the investigation of learning styles is vast with little agreement or collaboration to find agreemen t among the respective th eorists (Coffield et al.). Little attention has been given to how dissimilar learni ng styles between students and faculty members contribute to any measur e of academic success (Coffield et al.). Furthermore, the literature has been consistently unclear concerning the separation of learning style from learning capability (Kir ton, 2003). Research conducted to explain student achievement by measures of inherent learning styles without controlling for other variables, such as motivation, stress and inte lligence, tend to support beliefs of academic superiority (Hargreaves, 2005). Howe ver, if students learn best in their preferred learning style and instructional discourse is only provided in that learning style, academic superiority may be an outcome. Investigating how dissimilar learning styles contribute to academic success is necessary. Also, the separati on of learning style from learning level is warranted to allow for the distinct relationship and accurate measure of other classroom variables that facilitate studen t engagement in the classroom and promote academic achievement. That is, if dissimila r learning style is not found to be a large contributor to student engagement, perhaps learning le vel is a superseding variable. Faculties of colleges of agriculture and lif e sciences have been reliant on scholars in agricultural education to examine teachi ng and learning models that promote student engagement. Nevertheless, studies conducted in agricultural education have made few attempts to identify relationships among st udent characteristics, thinking skills and

PAGE 30

12 motivations that increase student engage ment (Rudd, Baker & Hoover, 2000). KirtonÂ’s A-I theory provides a framework to examine how motivation may overcome the strain of students and faculty members with dissi milar cognitive styles. Adaption-Innovation theory has been applied as a model to problem solving in business management, consumer marketing and aspects of change; however, little research has been done to apply A-I theory to the contex t of undergraduate e ducation. As A-I theory is not context specific (Kirton, 1994), the theory should also apply in the underg raduate classroom. Faculty members are increasingly faced w ith more accountability for increasing student achievement. As student engagement is closely related to academic achievement (Kuh, 2001), faculty members need to utilize in structional strategies and curricula to increase student engagement to foster student learning. Research is needed to determine if cognitive style contributes to the expl anation of student engagement. Does the cognitive style gap between studentsÂ’ prefer red cognitive style and facultiesÂ’ preferred cognitive style impair student engagement? Are students able to overcome cognitive style gap as the faculty member engages the learner? What are the implications of KirtonÂ’s A-I theory as applied within the cont ext of the undergraduate classroom? Significance The goal of many university faculties is to help students learn the content of the course in order to prepare them for graduate school, professional school or a successful career. The goals of many university stude nts mirror the faculty membersÂ’ goals. Increasing student engagement in the classroom will facilitate the su ccessful achievement of these same goals. Often there are impedim ents that restrain student engagement. Dissimilar cognitive style may be one impediment.

PAGE 31

13 This study can improve undergraduate instru ction in colleges of agriculture and life sciences by providing evidence of the e ducational impact of dissimilar cognitive style. Given this finding, faculty members can better develop programs and curricula to increase student engagement by reducing cognitive gap caused by dissimilar styles. Departments of agricultural education can help faculty members in colleges of agriculture and life sciences better prepare for teachi ng students with dissimilar cognitive styles. Also, university faculty can gain valuable knowledge concerning how learning style, not learning level, dictates ho w students view problems. This study contributed to the knowledge base of learning styles by providing data concerning the relationships between dissi milar cognitive styles, student stress, motivation and engagement in the undergradu ate classroom. Furthermore, collaborative research among learning style authors may be initiated to determine if they also can contribute to improving student engagement. Purpose of the Study The purpose of this study was to determ ine if significant relationships exist between cognitive style gap, student stress, student motivation, student engagement and selected demographic variables of undergraduate students at th e University of Florida in the College of Agriculture and Life Sciences The specific objectives of this study were to: 1. Describe selected faculty and student s according to their selected demographic variables; 2. Determine the cognitive style, student stress, student motivation, and student engagement of undergraduate students; 3. Determine the cognitive style gap between faculty and students and to explore its relationship with student stress, stude nt motivation, student engagement and specific demographic variables of undergraduate students; 4. Explain undergraduate student stress and student motivation based on cognitive style gap and selected stude nt demographic variables;

PAGE 32

14 5. Explain undergraduate student engagement based on cognitive style gap, student stress, student motivation and select ed student demographic variables. Assumptions Basic assumptions of this study were as follows: Student engagement is desired by faculty members in colleges of agriculture and life sciences. Faculty members teach in thei r preferred cognitive style. Problem solving learning exists in the undergraduate classroom. Limitations Conclusions and implications derived from this study were subject to the following limitations. Undergraduate courses are self-selected groups with students having individual differences prior to data collection. It was an ticipated that group differences existed between classes with regard to mean scores and demographic variables. Therefore, data were limited to this population of f aculty members teaching undergraduate courses in the College of Agricu ltural and Life Sciences at the University of Florida. Applications of the results of this study to other college classrooms were limited to the degree of similarity they have with courses us ed in this study. The researcher worked with the conditi ons allowed by the faculty member to collect data from their course. All instrument s were administered in the same manner, but some students completed instruments in cla ss and other students completed instruments outside of class. As the faculty member ma de choices regarding studentsÂ’ completion of the instrument, method of completion was confounded with the faculty member and therefore provided no way to conclude th ere was a method of completion effect on instrument scores. Even if courses were co mpared, self-selected courses have preexisting differences with no way to statistically determine causality of these differences.

PAGE 33

15 Faculty members volunteered students in th eir courses to partic ipate in this study. Although extra-credit points were often given as an incentive to increas e response rate, some students still chose not to respond. In this study, non-responding students included students who were administered instrument s and chose not to respond, and students who never attended class the day of instrument administration. This study is subject to nonresponse error as there was no da ta collected on non-responders. Instruments used in the study were self-re ported which may affect the accuracy of the data. The halo effect is a common pr oblem among student self-reports as students tend to inflate variables such as grades a nd effort. However, Pike (1999) found that the halo effect was relatively constant across different student groups and schools giving evidence that comparisons can be made wit hout disadvantaging a group. However, a halo effect on the KAI presents an unusable cognitiv e style score thus incr easing mortality rate of the instrument. Removing unusable scores fr om the data may have influenced findings in this study, especially if unusable cognitive style score is related to anxiety (Kirton, 1999). Other confounding variables that may aff ect this study include coping behavior (Kirton, 2003) of the faculty, competence of the faculty, motivation of the faculty, availability of teaching resources, cla ss sizes, and classroom environment. Definitions The following definitions are used in this study. Adaptive. A preference of cognitive style which pr oduces fewer ideas within a narrower perspective while tending to utilize given ru les. Adaptive cognitive style contrasts with an innovative cognitive style (Kirton, 2003). Behavior. An individualÂ’s actions and reactions while operating in an environment (Bandura, 1986). Many factors co ntribute to an individualÂ’s be havior, but one factor is preferred cognitive style blended wi th coping behavior (Kirton, 2003).

PAGE 34

16 Cognitive affect. The source of problem sel ection with a degree of priority placed on the search for a solution by means of mo tivation (Eccles, 1983; Kirton, 2003; Schunk, 2000). Cognitive climate. The mode preferred c ognitive style held by a group of individuals working together to solve problems (Kirton, 2003). Cognitive effect. A part of cognitive function incorpor ating both cognitive style, which influences behavior, and cognitive level which limits behavior by means of cognitive process (Kirton, 2003). Cognitive function. The controlling influence on beha vior through the interaction of three components: cognitive e ffect, cognitive affect and c ognitive resource (Kirton, 2003). Cognitive level. The aptitude, scope or expertise an individual has in solving a problem. It can be expressed as intelligence or leve l of creativity, but is not correlated with cognitive style (Ericsson, Krampe & Te sch-Romer, 1993; Guilford, 1975; Kirton, 2003). Cognitive process. An operation through a theoretical structure of thinking which one advances step by step through the stages a ssumed to be the operations of the brain. Cognitive process is distinguished from cogn itive level and cognitive style, but utilizes both in problem solving (Kirton, 2003; Kolb. 1984; Wallas, 1926). Cognitive resource. The storage of learned informati on (memory) manifested through constructs of knowledge, experien ces, and skills (Kirton, 2003). Cognitive style. The consistent individual differences found in ways of organizing and processing information and experience (Messi ck, 1984). This study uses cognitive style synonymously with learning style (Claxt on & Ralston, 1978). Cognitive style was operationalized by the Kirton Adaption-Innovati on (KAI) Inventory wi th constructs of sufficiency of originality, efficiency, and ru le/group conformity w ith tendencies to be more adaptive or more innovative with in these construc ts (Kirton, 2003). Cognitive style gap. A degree of separation between cogni tive styles of two individuals or between an individual solving a problem and the cognitive style required to solve a specific problem. (Kirton, 2003). Operationa lized by subtracting a faculty memberÂ’s cognitive style score from a st udentÂ’s cognitive style score. Cognitive technique. A learned skill an individual may utilize to increase cognitive level or avert from a preferred c ognitive style (Kirton, 2003). Coping behavior. A learned skill that is initiated with motivation and used to perform in a cognitive style that is different from oneÂ’s preferre d cognitive style. Coping behavior can only last for a duration and intensity (Kir ton, 2003). Coping behavior in this study is not to be confused with coping behavior as a reaction to stress which is defined as a

PAGE 35

17 conscious effort to deal with an event to which an individual has not yet developed automatic patterns of response (Shields, 2001). Engagement. The extent that students invest th eir time and psychological energy in empirically derived good educational pr actices. (Astin, 1996, Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Engagement was operationalized with the use of the National Survey for Student Engagement (N SSE) (2005) with constructs of academic challenge, active learning, a nd student-faculty interaction. Environment. The social context indivi duals interact with during the problem solving process (Kirton, 2003). Within the classroom these variables incl ude: personalization, involvement, student cohesiveness, satisfact ion, competition, task orientation, and rule clarity (Fraser, 1998). Faculty Member. An individual responsible for pr esenting instructional discourse, providing problem sets for student completi on and assigning the final course grade. Innovative. A preference of cognitive style to pr oduce a larger number of ideas with a wider variety of perspectives to make a product different while tending to alter rules when solving problems. Innovative cognitive style contrasts with adaptive cognitive style. (Kirton, 2003). Learning style. A consistent way an individual re sponds to, and uses stimuli in the context of learning. (Claxton and Ralst on, 1978). This study uses cognitive style synonymously with cognitive style (Messick, 1984). Learning style was operationalized by the Kirton Adaption-Innovation (KAI) Invent ory with constructs of sufficiency of originality, efficiency, and rule /group conformity with tendenc ies to be more adaptive or more innovative within these constructs (Kirton, 2003). Motivation. A process that includes both initiati ng and maintaining a goal-directed activity (Pintrich & Schu nk, 2002). Comprises of three different components: expectancy–a student’s belief a bout their ability to perform a task, value–a student’s goal and belief concerning the importance and intere st of the task and affective–a student’s emotional reactions to the task (Pin trich & De Groot, 1990). Motivation was operationalized with use of the Motivated St rategies for Learning Questionnaire (MSLQ) with constructs of intrinsic motivation, ex trinsic motivation, task value, control of learning belief, self-efficacy fo r learning and test anxiety. Problem. A desire or felt need for a solution, but not immediately knowing the mental operations to arrive at the solution (Soden, 1994). Products and ideas. A solution to a problem fitted to the degree of structure preferred by the problem-solver. A more adaptive person will prefer more struct ure in the generation of ideas recognizing boundaries of the relevant paradigm and a more innovative person will prefer less structure in the generation of ideas disregarding the boundaries of the relevant paradigm (Kirton, 2003).

PAGE 36

18 Stress. The perception of a stimulus in the form of an event or condition and requires an individual to adjust from normal life. Stresso rs can be positive or negative depending on the individualÂ’s reaction to the experien ce (Gadzella, 1994; Roma no, 1992). Stress was operationalized with use of the Student-life Stress Inventory (SSI) with constructs of frustrations, conflicts, pressures, changes and self-imposed stress. Summary In summary, learning styles theory has become popular as pragmatic educators can easily see its existence in the classroom However, many learning style instruments have proven to be unreliable a nd invalid. There is little empiri cal evidence that dissimilar learning styles contribute to understanding st ress, motivation, and e ngagement; these are common variables found in the undergraduate cl assroom. KirtonÂ’s theo ry of A-I provides a framework to distinguish cognitive style from cognitive level to help understand the influences of the aforementioned variables. As faculty members become more accountable for student engagement and l ook for instructional methods to promote learning, evidence is needed to determine th e importance of learning styles in student engagement. KirtonÂ’s definition of cognitive st yle and psychometrically sound instrument to measure cognitive style may be of particular use.

PAGE 37

19 CHAPTER 2 REVIEW OF LITERATURE The purpose of this study was to determ ine if significant relationships exist between cognitive style gap, student stress, student motivation, student engagement and selected demographic of undergraduate student s in the College of Agriculture and Life Sciences at the University of Florida. The specific objectives of th is study were to: 1) describe selected faculty and students accordi ng to their selected demographic variables; 2) determine the cognitive st yle, student stress, stude nt motivation, and student engagement of undergraduate students; 3) determine the cognitiv e style gap between faculty and students and to expl ore its relationship with stude nt stress, student motivation, student engagement and specific demographi c variables of undergraduate students; 4) explain undergraduate student stress and st udent motivation based on cognitive style gap and selected student demographic variable s; and 5) explain un dergraduate student engagement based on cognitive style gap, student stress, student motivation and selected student demographic variables. This chapter summarizes pertinent liter ature concerned with learning styles, stress, motivation and engagement. Special at tention will be given to KirtonÂ’s cognitive style (Kirton, 2003) as applie d to the context of the unde rgraduate classroom, thus providing the theoretical framework for the study Also a discussion of empirical research conducted to examine learning style, stress, motivation, and engagement, is presented to determine relationships between these variables.

PAGE 38

20 Within this chapter are the following major sections: theoretical framework, learning and learning styles, l earning and stress, learning and motivation, and learning and engagement. Theoretical Framework Literature concerning learning styles incl udes a wide variety of published sources consisting of empirical research f ound in refereed j ournals, pedagogical recommendations found in teacher newsle tters, and books advocating the use of a specific learning style inve ntory (Coffield, Moseley, Ha ll & Ecclestone, 2004). Said differently, the scope of learning styles rese arch is based on instrument development, adding to theory, classroom improvement, and commercial use. Kirton’s A-I theory (1976) is a theoretical response to the literature stating that all people can be creative in problem solving in their own cognitive style, thus distinguishing be tween cognitive style and cognitive level (Kirton, 1978). Kirton’s A-I theory provide d the theoretical framework of this study. The most detailed explanation of A-I theory is given by Kirton (2003) in Adaption-Innovtion: In the c ontext of diversity and change with illustration of cognitive function, behavior and envir onment in relation to idea formation (p. 37). Within the theoretical model, cognitive style and cogniti ve level are components of cognitive effect. Cognitive effect is defined as a part of cognitive function incorporating both cognitive style and cognitive level, both of which are employed by the individual to complete the cognitive process (Kirton). The strength of A-I theory is the sepa ration and independent measure of cognitive style from cognitive level. Cognitive style is defined by Kirton (2003) as “a strategic, stable characteristic–t he preferred way in which people respond to and seek to bring about change” (p.43). Kirt on’s definition of cogni tive style regarding

PAGE 39

21 adaptiveness and innovativeness wi ll be explained further in s ubsequent sections of this chapter. Examples of cognitive level include measures of capacity or ability such as intelligence. However in problem solving, both style and level are required to complete a cognitive process. Cognitive processes can be described as the “ste ps of how I operate” and “not measured, but validat ed to provide information of where we are” (p.38) in solving a problem. Examples of cognitive process include Kolb’s cycle of Experiential Learning (de Ciantis & Kirt on, 1996; Kolb, 1984) and the stages of problem solving (Wallas, 1926). Each stage of a cognitive pr ocess can be completed at a high cognitive level or low cognitive level and through th e use of different cognitive styles. The outcome of cognitive process is knowledge (learning) stored in cognitive resource (Kirton, 2003). The cognitive process can also be accomplished with low motivation or high motivation as determined by cognitive affect (p. 89). Cognitive affect is defined as the motivation and value placed on the search for a solution to a problem (Kirton). Ultimately, cognitive affect, cognitive resource a nd cognitive effect are three distinct and uncorrelated components of cognitive function (Kirton) and employed by the individual to solve problems. This study examined cognitive style gap between the teacher and the student and its relationship with stress, motivation and engagement. Other gaps exist between the cognitive function of teacher and student incl uding: cognitive gap of level–a difference of intelligence, cognitive gap of affect–a differe nce of motivation or va lue and cognitive gap of resource–a difference of knowledge or skill (Kirton, 2003). These cognitive gaps influence the interaction between two individuals, but the separation of cognitive style

PAGE 40

22 from cognitive level, cognitive affect and c ognitive resource allows for a more precise measure of dissimilar cognitive style and its relationship with classroom engagement. In the effort of solving a problem, indivi dual differences become apparent at the realization of cognitive gaps. Nevertheless, if the motivation is great enough for teacher and student, both will be engaged in the pr ocess of learning (Kirton, 2003). Ideas and solutions may come from both student and teacher. Although students are responsible for engaging in learning activities the teacher facilitates lear ning and therefore accountable for student engagement. Student engagement includes constructs of academic challenge, active learning and student-faculty inte raction (Kuh, 2001). Teacher engagement facilitates student engagement and can be measured by the teacher Â’s knowledge of the learning process, intrinsic motivation for teaching, and self-efficacy toward teaching (Knowles, 1999). Kirton (2003) defines the environment as the social context in which individuals interact during the problem solving process. An individualÂ’s interaction with the environment manifests as people work together in the environment to solve problems and react to the feedback provided by others (Kirton). In the undergraduate classroom, students interact with fellow students a nd the faculty member through discussions, activities, assignments and projects; all of which provide student feedback either by fellow students or the faculty member. In a review of instruments measuring factors influencing the undergraduate classroom environment, Fras er (1998) categorized seven scales: personalization, involvement, student cohesiveness, satisfaction, competition, task orientation, and rule clarity. Ho wever, Fraser further discusse d that this list may not be

PAGE 41

23 conclusive as little research has been conducted regarding the environment of undergraduate classrooms. Finally, student academic achievement results from the faculty member and students engaged in the classroom environmen t for the purpose of learning. Academic achievement can be measured regarding a specific assignment or a cumulative total provided by the faculty member with values re presented as either a percentile or a letter grade. Figure 2-1 displays Ki rtonÂ’s A-I theory in the cont ext of a teacher and student engaging in idea formation. For the individual, coping behavior is used when the rewards gained from solving the problem are greater than the effort exer ted to operate outside of an individualÂ’s preferred style (Kirton, 2003). Said differe ntly, cognitive style gap causes stress and motivation is needed to overcome this stre ss and bridge the cognitive style gap between two individuals. Researchers have investigated the relatio nship cognitive style gap has with stress. McCarthy (1988) found that women managersÂ’ stress increased as the size of the cognitive style gap increased. Furthermore, the threshold at which stress became apparent was at a cognitive style gap of 20 points. Another study conducted by Kubes and Spillerova (1992) had concurring results with a large cognitive style gap influencing the construct aggravation within a working group. However, both of these studies were conducted in the context of busin ess. Little is known if A-I theory is appropriate for the undergraduate classroom and if these relations hips between dissimilar cognitive style and student stress also apply.

PAGE 42

24 Figure 2-1. KirtonÂ’s theoretical model of adaption and innovation ap plied to the undergraduate classroom (Kirton, 2003)

PAGE 43

25 Both intrinsic and extrinsic motivation can be driving forces for increasing engagement in the problem solving proce ss (Hennessey & Amabile, 1993). In the event that a course is taught by a faculty member with a dissimila r cognitive style, a student may be motivated to cope outside of a preferred cognitive style in effort to interact more with the faculty member and improve stude nt engagement (Kuh, 2001). Motivation is not a result of cognitive gap, but is a driving fo rce for using a cognitive style that is not the individualÂ’s preferred style. Specifically, motivation is a factor of coping behavior (Kirton, 2003); if motivation can be partiall y explained by cognitive style gap, then there is evidence that students ar e coping. Motivation is a compone nt of cognitive affect and therefore independent of cognitive styl e, a component of cognitive effect. Utilizing the framework provided by A -I theory, student engagement was determined to be an outcome of problem so lving and idea formation. Student engagement was used instead of academic achievement in this study to provide a better measurement of the interaction between f aculty member and student. Kirton (2003) asserts that adaptive and innovative indi viduals prefer different amounts of structure while completing the problem solving process. Whereas more adaptive individuals tend to solve probl ems favoring structure, more innovative individuals tend to solve problems unbound by st ructure. Adaptive individuals tend to have fewer ideas making the solution more e fficient; innovative individuals prefer to have a wide range of ideas with the know ledge that many of the ideas are improbable (Kirton, 1994). The difference between two styl es of problem solving is often evident by the differences of perception and communicati on of the solution. That is, a gap between two dissimilar cognitive styles tends to inhi bit effective communica tion. This failure of

PAGE 44

26 communication can often lead the two indi viduals to focus on the new problem of ineffective communication and not the probl em at hand which is completing course projects and assignments (Kirton). If this occurs in the undergraduate classroom, students may dwell on the inefficiencies of communica tion with the faculty member instructing the course and may not be motivated enough to engage in learning. The constructs of student engagement include level of academ ic challenge, active and collaborative learning, and student interac tion with the faculty member, (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001), all of which requ ire communication between the faculty member and the student to be achieved (U mbach & Wawrzynski, 2005). In summary, the faculty member determines the amount of stru cture used in the course, but students may prefer different amounts of st ructure and find themselves unmotivated to effectively communicate with the faculty member, thus im pacting the level of student engagement. KirtonÂ’s (2003) A-I theory may contri bute to the explanation of student engagement by identifying the relationships between cognitive style gap, stress and motivation. An increase in cognitive style gap between the student and the faculty member causes stress through working together in solving problems or engaging in learning. If the student is motivated, coping be havior is used to operate outside of a preferred cognitive style and employ a cogniti ve style closer to that of the faculty member (Kirton). As a result, student and faculty member are operating in the same cognitive style which may increase student engagement. However, student motivation supplied to coping behavior only last for a period of time and at a limited level (Kirton). Consequently the student reverts back to a preferred cognitive style which may be detrimental to student engagement. For the stude nt to be successfully engaged the reward

PAGE 45

27 for learning must be greater than the stre ss imposed by cognitive style gap from the faculty member (Kirton). Figure 2-2 portray s the interaction of student stress and motivation in explaining student engagement. Student Engagement Student BehaviorStudent Cognitive Style Gap Initiating Stress Teacher Behavior 10 30 50 70 90 Student Coping Behavior Initiated By Motivation Figure 2-2. Interaction of cognitive style ga p, stress, motivation and engagement (Kirton, 2003) Other variables that influence studen t engagement include faculty member behavior (Umbach & Wawrzynski, 2005), cl ass size (Ehrenberg, Brewer, Gamoran & Willms, 2001), use of teaching methods (M arzano, Pickering & Po llock, 2001), gender, ethnicity, full-time student status, organiza tional involvement and discipline of study (Pike, 2004). There were other confounding variable s but were considered outside the objectives of the study. They included teach ing effectiveness, teacher engagement, classroom environment and studentÂ’s pr evious knowledge. Furthermore, it was

PAGE 46

28 impossible to determine the variables outside of the classroom that inadvertently effect student engagement. These incl uded but were not limited to sickness, family crisis and lack of time devoted to study. Kirton (1994) offers the following anecdote to teachers and students in reference to A-I theory: We all teach children to be more adaptiv e: before answering the question, please read it; be more methodical; present your wo rk more neatly; nice essay, pity it is not on the topic I set; and so on. The best teachers taught most of us a lot of technique, but changed us (our preferred c ognitive style) not at all. In the process they found some of us (adaptors) more re ceptive to working more closely within structure than others (innovators). Lesser teachers, especially when faced with the less motivated innovators, got even less far. Of course, despite the groans in the literature, not all teachers are or ever were, adaptors The highly innovative ones delighted most of us, whether we always understood them or not; some of us (the higher adaptors) were just confused. As the saying goes, “You just can’t win ‘em all’.” The best teachers win more than most. (p. xxiii). The undergraduate classroom offers a new c ontext for A-I theory as well as the introduction of new variables. Although th ese variables were not discounted by A-I theory, up to this point, they had not been fully explored. Given the previous description of A-I th eory, the researcher found two studies of interest. First consid ering the undergraduate classroom, Pu ccio, Talbot and Joniak, (1993) found that students’ stress increased with the perceived adaptive ness of the course. Specifically, the rule/group c onformity construct of the KAI was most important in determining the relationship between cognitive style and stress. It se ems logical that an innovative student would find an adaptive cour se stressful, but why would an adaptive student find an adaptive c ourse stressful? The research er found the study conducted by Puccio, Talbot and Joniak one of the few inve stigations regarding Kirton’s A-I theory applied to the undergraduate classroom.

PAGE 47

29 In the second study, Barnhart (2002) set out to determine if increased cognitive gap led to aviation flight instruction ine fficiencies. No significant differences in effectiveness were found between total cognitive style gap between students and flight instructors, however significant effects we re found by analyzing KAI construct gaps. Specifically, a significant weak negative correlation of the efficiency cognitive style gap and time spent preparing for the first solo f light (r=-.23, p<.05) and time spent preparing for certification (r=-.26, p<.05). Furthermore, there was a weak negative correlation of rule/group conformity style ga p and time spent preparing for certification (r=-.29, p<.05) (Barnhart). Although these correlations were considered low (Mill er, 1998), one would expect a positive correlation in this study. That is, cognitive style gap should increase with the amount of time spent preparing for the first solo flight and certification. Both studies previously mentioned f ound low negative correlations between cognitive style gap and measures of stress a nd effort. Although these findings contradict with the theoretical model of this study, it is important to recognize them as two of the few published studies found by the researcher that applied A-I theory to the context of the learning environment. Learning and Learning Styles To understand learning styles, it is not only necessary to define learning, but also to provide learning theory to give evidence of different learning types, processes and styles. Gagne (1965) proposed ei ght types of learning: 1) si gnal learning; 2) stimulusresponse learning; 3) chaining; 4) verbal-assoc iate learning; 5) multiple discrimination; 6) concept learning; 7) principl e learning; and 8) problem so lving (p. 33). The types of learning are hierarchical as the la tter is a prerequisite of the former for a type of learning to take place (p. 60). Additionally, learning can be any one of these eight types which

PAGE 48

30 delineates the learning process of a situation and giving reason for a learner to exhibit a preferred style and a chosen level (Gange). Research conducted in the domain of l earning styles was categorized by Curry (1987) into four dimensions which formed th e “onion” model. Personality trait types of learning styles were categori zed at the core of the onion, as they were found the most stable within a learner. The second layer of learning styles represents information processing, while the third layer includes the so cial and situational interactions of the learner. The most unstable category of l earning styles includes the preference of instructional methods, found on the outside laye r of the “onion.” Curry placed Kirton’s measure of learning style at the core of the onion. An individual’s cognitive style defined by Kirton (2003) does not change as a person ag es, and is correlated with many measures of personality (Kirton, 1999). Nor does it change as the individual solves any particular type problem. However, in problem solvi ng, an individual may cope outside of one’s cognitive style as the situation requires. Shuell (1986; as cited in Schunk, 2000) defines learning as “an enduring change in behavior, or in the capacity to behave in a given fashion, which results from practice or other forms of experience” (p. 2). It is plausi ble that each of the eight types of learning provided by Gagne (1965) is fitting with this definition. That is, each type of learning may result from an experience and produce a behavioral change. Evidence of this may be found by examining a type of learning proce ss. Using stimulus-response learning as an example, a student may be stimulated to increase focus during independent study time. The discriminating stimulus is the al lowance to use background music during independent study time. This stimulus triggers a student response to either use or not use

PAGE 49

31 background music. If the student uses background music to help focus during independent study time, another stimulus (prais e) is utilized to reinforce the behavior (Skinner, 1953). Using problem solving learning as an example, a student is stimulated by a felt need to solve a problem of comple ting all homework assignments before the end of the night. The student knows he may use independent study time, but it is hard for him to focus. He generates solutions to solve his problem of lack of focus by playing music in the background and executes a plan of acti on (Wallas, 1926). For both types of learning, the increase of these experiences (practice) results in a changed behavior (increasing focus). Given the definition of learning used in this study, it seems paradoxical that one would have a preferred learning style that is evident by a stab le and innate behavior not altered by instruction (Kirton, 2003). However, learning styles are a preferred approach to learning, or a perception of the learni ng environment to which the learner has a preference. Claxton and Ralston (1978, p. 7) define learning st yle as “a consistent way an individual responds to, and uses stimuli in the cont ext of learning,” which is similar to Messick’s (1984) definition of cognitive style. Furthermore, a person may cope outside of a preferred learning style given a different situation, adequate motivation, and learned techniques (Jones, Reichard & M okharti, 2003; Kirton, 2003; Kolb, 1984). Ford and Chen (2001) found that matchi ng a student’s learning style with a faculty member’s learning style improved stude nt learning. However, Ford and Chen also recognized that type of learni ng and other variables interact with the matching of similar learning styles. Jensen (2000) connected scient ific studies of the brain to teaching and learning and wrote, “The whole notion of lear ning styles becomes irrelevant when we

PAGE 50

32 consider how much variety th e brain thrives on” (p.138). Howe ver, Jensen further wrote that learners do have preferences of many lear ning styles and operate in those styles at different levels. It is useful to understand aspects of a preferred style in order that learning differences are realized. This belief is concurrent with the most complete and thoroug h literature review concerning learning styles to date (Coffiel d, Moseley, Hall & Ecclestone, 2004). In this literature review, 71 learning styl es were held up to criteria of psychometric properties, theoretical backing and practical usage in th e classroom. The authors agreed that even though most learning style measures fall short of these criteria, each have strengths and weaknesses. Realizing the comple x nature of the brain, it is conceivable that more than 71 different learning styles do exist and further research wi ll bring greater understanding as to how these styles associate with learni ng. It is interesting to note that of the 71 learning styles reviewed by Coffield, Mose ley, Hall and Ecclestone, only a number of authors relied on related theories and con cepts of learning and all used different terminology, yet the term learning style and cognitive style encompass all measures of preferred behaviors in learni ng. Coffield, Moseley, Hall and Ecclestone argue for more collaborative research between learning st yle authors as findings may bring greater understanding to the process of learning and student engagement. It may be possible that a sp ecific learning style is congrue nt with one of the eight types of learning noted above by Gange (1965). For example, Kolb’s (1984) measure of learning style and process is harmonious with the learning of a principle. To learn a principle a learner engages in a concrete experience, followed by self-reflection and abstract conceptualization and ending w ith experimentation of that principle.

PAGE 51

33 Furthermore, Kirton’s (1976) measure of lear ning style is harmoni ous with learning from problem solving. To learn from problem solv ing, a learner manipulates a principle and interprets the meaning from the manipulat ion. For effective learning from a problem, Pask (1976) claims the learner must use both holist and serialist cognitive styles. Although Pask and Kirton measure cognitive style differently, they are much the same when considering description and categorizat ion (Cassidy, 2004). Both of these measures of cognitive style are representati ve of problem solving style. Other types of learning defined by Gange ( 1965) may be compatible with specific learning styles. For example, the Group Embedded Figures Test (GEFT) (Witkin, Oltman, Raskin & Karp, 1971) measures whether an individual is eith er field independent or field dependent by the individual’s ab ility to identify embedded figures within geometric shapes. The cognitive processes used to complete the GEFT gives some evidence that field independence and field dependence is a style of concept learning (Gange, 1965). Concept learning is defined as comparing and distingu ishing properties of similar items based on knowledge of those it ems (Gange). However, Witkin (1973) has found that his measure of learning style does ha ve implications for problem solving with regards to the need for structur e. Field dependent lear ners tend to prefer more structure in learning while field independent learners prefer less stru cture. This corresponds to Kirton’s (2003) adaptive indi viduals who favor structure and innovative individuals who prefer less structur e in the effort of problem solving. Learning Style and Learning Process Most psychologists concur that learni ng occurs through a process or cycle. Research of the brain has found five st ages for effective learning (Jensen, 2000), including preparation–background recall of information, acquisition–inputting

PAGE 52

34 information, elaboration–experimentation and feedback, memory formation–encoding through emotion and context, and functiona l integration–applica tion of knowledge (Jensen). It is easy to see these stages w ithin other learning processes. Kolb (1984) defines the process of learning a principle as a concrete experience with the principle to be learned, followed by self-reflection and ab stract conceptualiza tion and ending with active experimentation of that principle. Teaching and learning the problem solving process includes the stages of problem iden tification, idea generati on, idea evaluation and implementation (Swartz & Perkins, 1990). Still, the process the student uses to gather information dictates how the information is stored (Jensen, 2000; Pi aget, 1970). That is, a learner studying a principle will acquire know ledge of the principle, but not how to manipulate the principle to solve a problem ; knowledge to solve a problem must be acquired from the process of problem solving learning. Kirton (2003) argues that le arning style is independent from the learning process, a contradiction of Kolb’s experiential lear ning theory and Kolb’s measure of learning style. Kirton empirically tested (de Ci antis & Kirton, 1996) his claim by conducting a factor analysis on the Learning Style Questionnaire (Honey & Mumford, 1992), a learning style measure congruent with principle learning. The Learning Style Questionnaire was found as a more reliable ve rsion of Kolb’s Learning Style Inventory. Following factor analysis, comparisons were made with Kirton’s Adaption-Innovation Inventory to make the argument of independe nce between style and process. However, it may not be sensible to compare a learning style of principle and a learning style of problem solving as the two are of different l earning processes. That is, a learning style of problem solving may be independent of pr inciple learning, but not independent of

PAGE 53

35 problem solving learning. Furthermore, Kirton (2003) claims that ad aptive learners have preferences of inductive thinking methods lead ing to better problem representation, while innovative learners tend to us e deductive thinking methods to proliferate more ideas (p. 172). Again, both problem representation a nd idea generation are components of the problem solving process. The process of learning and learning style may not be completely independent as Kirton suggests. In addition to this, e ffective learning from problem solving requires the learner to us e cognitive techniques that correspond with adaptiveness and innovativeness (Pask, 1976) wh ile manipulating a principle to solve a problem. If coping behavior is required for an individual to be more adaptive or innovative, there is theoretical evidence that in problem solving learning stress is inevitable and motivation is needed to comp lete the problem solving learning process. Learning Style and Learning Level Kirton (1994) further separates problem solving in terms of cognitive level and cognitive style. Whereas cognitive style is fixed, innate and guides preferences of structure for solving problems, cognitive level is the capability to solve problems at a degree of complexity (Kirton). Examples of cognitive level include intelligence, creativity, academic achievement, etc. Again, the separation of cognitive style from cognitive level is the strength of A-I theor y. The distinction and accu rate measurement of cognitive style allows for the precise meas ure of other variables such as student intelligence, engagement, and motivation. According to A-I theory, all people ca n creatively solve problems and have original ideas (Goldsmith, 1994). However wh en solving problems, people who are more adaptive tend to produce a smaller number of ideas with a narrower perspective that tend to be more relevant, sound and useful while utilizing given rules. People who are more

PAGE 54

36 innovative generate a larger number of origin al ideas of which ma ny are different with widely changing perspectives and tendencies to alter rules (Kirt on, 1994). Both adaptive and innovative people can adequately solve a ny problem, just within their own style. Kirton’s Cognitive Style From the definitions of learning styl e (Claxton and Ralston, 1978) and cognitive style (Messick, 1984) used in this study, Kirton’s cognitive style was determined as a style of learning. Additionally, Kirton’s Ad aption-Innovation Inventory (KAI) simply measures style of problem solving and is not convoluted with factors of cognitive level, cognitive affect or other types of lear ning (Gange, 1965). Three factors located on a continuum of more adaptive to more innovativ e have been identified and determined as preferential behaviors of an i ndividual. Table 2-1 disp lays the factor trai ts of adaptors and innovators. Table 2-1. Factor Traits of Adaptors and Innovators Construct Adaptor Innovator Sufficiency of originality Efficiency Rules/Group conformity Tendency to produce fewer ideas aimed at being sound, useful and relevant to solving the problem. Preference for attention to detail to further develop ideas while narrowly focused on improving. Often applies group accepted rules in idea formation to solve problems. Tendency to produce more ideas that are both relevant and irrelevant to the problem with knowledge that some ideas are inappropriate. Preference for discontinuous thought to modify the problem with broad approaches to make different. Often challenges group accepted rules in idea formation to solve problems. Note. Source: Kirton, 2003, p. 58-60. The factor traits of A-I theory are “inna te, fixed and reliably measured” (Kirton, 2003, p. 76). Each of the three factor traits le nds to a perceived structure of the problem

PAGE 55

37 at hand which determines the approach needed to solve the problem The more adaptive tend to work within the frame of structure. A more adaptive cognitive style rarely challenges the boundaries of structure and so lutions to the problem are focused with detail to derive a better so lution that improves efficienc y. A more innovative cognitive style breaks boundaries as if they never ex isted and solutions generated are often unrelated in an effort to deri ve a different solution that transforms the structure itself (Kirton). The perception of the structure from the framed problem determines the response of the learner, providing for th e use of a preferred cognitive style. The Scale of Adaption and Innovation A benefit of A-I theory and the KAI is th e interval measure of difference between two cognitive styles, providing a specific de gree of separation between an individualÂ’s cognitive style from another i ndividualÂ’s cognitive style. This is opposed to a learning style measured nominally and determined as categorically different; a model for many learning style inventories. The KAI measures cognitive style by scoring adaptiveness and innovativeness along a continuum (Kirton, 2003 ). Anchoring the more adaptive on the left and the more innovative on the right of th e continuum, an interv al scale is created amid a theoretical range of 32 to 160 (K irton). However, the actual range is approximately 45 to 145 with a mean of 95 (Kirton). It is important to note that the terms ad aptors and innovators are used only to distinguish two differing cogniti ve styles, when in actuality, cognitive styles refer to being more adaptive or more innovative in the context of a situ ation (Kirton). For example, an individual may have a cognitive score of 125 distinguishing them as more innovative, but this individual wi ll be perceived as adaptive to an individual with a score of 140.

PAGE 56

38 Kirton (2003) suggests that a 10-poin t difference in score between two individual’s cognitive style, as measured by the KAI, develops a “noticeable difference” when working together on a problem. The difference found between the two cognitive styles is defined as cognitive style gap (Kir ton). A 20-point cognitive style gap leads to difficulties in mutual understanding and colla boration which may have a definitive effect on communication when problem solving (Kirton) For an individual with this amount of cognitive gap between one’s preferred cognitive style and the style re quired to work with a problem or another problem solver, coping be havior is needed and can exist in both intensity and duration (Kirton). Coping beha vior requires motivation (Walling, 1987) and is psychologically expensive (Kirton, 2003). When solving a problem, more adaptive individuals can use techniques such as brainstorming which can be taught, however, th ey are not preferred and less likely to use it in their natural envi ronment. Innovative individuals, li kewise can be taught to master details and operate within rules, but woul d prefer the unstructured situation. Kirton further suggests that practice of coping beha vior does not make coping easier for that person (Kirton, 2003, p. 255). This suggestion does limit the definition of learning used in this study; however the s uggestion is supported in the creativity literature. Few researchers have been able to provide hard ev idence that an individual can be taught to be more innovative or more adaptive (Nickerson, 19 99). Furthermore, there is evidence that teaching an individual to use different problem solving techniques has little long term effect (Nickerson). Cognitive style (Kirton, 2003) is fixed, innate, and determines an individual’s preference for stru cture within problem solving. There is no association with

PAGE 57

39 cognitive style and nationality or age, however females tend to be five points more adaptive then the general population mean (Kirton). Summary of Learning and Learning Styles Gange (1965) described eight different types of learning that are hierarchical in nature. Of the eight types, l earning from problem solving is th e highest form of learning with evidence of a problem solving process a nd problem solving style or cognitive style. KirtonÂ’s measure of cognitive style is a preferred approach to solve problems and independent from cognitive level (Goldsmith, 1994). However, both cognitive style and cognitive level are utilized to complete the problem solving process. There is evidence that other measures of learni ng style are specific to other types of learning defined by Gange (1965) with evidence found in the learning process. Coffield, Mosley, Hall and Ecclestone (2004) identified 71 different types of cognitive styles with varying degrees of reliability and validity. KirtonÂ’s measure of cognitive style was determined a reliable a nd valid measure of cognitive style (Kirton, 1999). Furthermore, the measure was found i ndependent from measures of motivation and capability (Kirton). KirtonÂ’s cognitive style determines if a person is more adaptive or more innovative in oneÂ’s preference to solv e problems and is stable characteristics of the individual. Along a continuum of adap tion anchoring the le ft and innovation anchoring the right an individualÂ’s cognitive styl e score is placed in a theoretical range of 32 to 160 (Kirton). Dissimilarity between th e cognitive styles of two individuals is determined by calculating the difference be tween two cognitive style scores. This cognitive style gap indicates differences betw een preferred methods of solving problems with stress as a consequence. Motivation may be utilized by the student to operate in a different cognitive style than preferred a nd bridge cognitive gap (Kirton, 2003). A

PAGE 58

40 cognitive gap larger than 20 points may i nhibit communication (K irton) between the student and faculty member and may e xplain lack of student engagement. Learning and Stress College students may be influenced by stress positively or negatively (Sandler, 2000). That is, stress as a stimulus can provoke a negative or positive reaction to the situation at hand. If a student ha s a positive reaction it is the f ear of failure that motivates this student to succeed, but a negative reacti on will likely be detrimental to classroom success. Given this variability of outcome from stress, little resear ch has been conducted with attention to stress asso ciated specifically concerni ng the undergraduate classroom. A word of caution is given regarding terminology and definition in discussing stress. Coping behavior specific to the reaction to a stressor can be defined as a conscious effort to deal with an event to which an individual has not yet developed automatic patterns of response (Shields, 2001). In this document, reaction is used instead of the word coping so that the reader does not conf use coping as a reaction to stress with coping behavior defined by Kirton (2003). A studentÂ’s reaction to the stressor may de termine if the stress is beneficial or detrimental to the academic achievement of the student. In a study conducted to examine this hypothesis, Aspinwall and Taylor (1992) surveyed 672 freshman undergraduates. Their findings provide evidence that a pos itive mood, higher optimism and an active reaction to stress orientation had positive eff ects on adjusting to new college experiences. Furthermore, the reaction to stress construc t avoidance had detrimental effects on the adjustment to new college experiences. Hock ey (1979) considered th e role of stress in aiding the selectivity of atte ntion and response in performi ng a cognitive task. Hockey

PAGE 59

41 elaborated that there may be an optimal le vel of stress for cognitive function although determining this level is dependent on the components of the task. Researchers have studied the detrimenta l effects of anxiety, a correlate of academic stress (Misra and McKean, 2000). In countless studies, test anxiety has been found to be detrimental to academic achie vement and self-esteem (Pintrich & Schunk, 2002). Specifically, worry is a factor of anxiet y that causes distractions resulting in time lost being able to perform at the level the student is capable of achieving. Most of the literature concerning anxiety focuses on this worry component and its influence on studentsÂ’ attention and time on task in comp leting an assignment or test. Like stress, anxiety is an affec tive component of KirtonÂ’s A-I theory. Undergraduate students face many stresscausing experiences throughout the time spent in completing a bachelorÂ’s degree. Among these experiences are moving away from home, academics, finances and work and relationships (Pfeiffer, 2001). The experiences themselves do not cause stress but the studentÂ’s perception and reaction turns the experience into a stressor (Romano, 1992). Studies have been conducted to determine the effects of specif ic interventions of student stress, such as better time manage ment. Again, none of the studies found by the researcher dealt with techniques specifically used to improve academics or studentsÂ’ relationships with faculty members. Stress in the Classroom The college environment is much differ ent than the work environment, making the study of college stress unique in the li terature. Research of undergraduate student stress has been conducted concerning the globa l stress of a college student as well as academic stress restricted to the classr oom. Academic stressors for undergraduate

PAGE 60

42 students may include homework, unclear as signments, uncomfortable classrooms and time pressure set by the faculty member (R oss, Niebling & Heckert, 1999). The most common stressors of college students may be intrapersonal, including the change in sleeping habits, change in eating habits and new respons ibilities (p.317). However, changes in the environment may also change types of stressors. That is, measuring stress at different times during the ye ar and at different times leading to the completion of a degree may find different levels of stre ssors. Shields (2001) found that stress had detrimental effects on grade point average du ring the fall semester, but not the winter semester. Shields postulates that the effects of a stressor diminishes once students begin to activate positive reactions to the stressor. Regarding academic stress, Ross, Niebling and Heckert (1999) found that increased cl ass workload and lower grades than anticipated were the two most frequent re ported stressors among undergraduate students. Pfeiffer (2001) believes that these academic stressors can cause depression and low selfconfidence (p. 13). In a study conducted to find the relationship between stressors, learning strategies and test anxiety, Gadzella, Masten, and Stacks (1998) found that frustration was negatively correlated (r=-.20, p<.05) with the learning strategy deep processing (Schmeck, Ribich, Ramanaiah, 1977). That is, students do not become as frustrated when they become more organized (G adzella, Masten & Stacks, 199 8). Surprisingly, this study found no significant correlations with fact ors of stressors and test anxiety. Using GadzellaÂ’s Student-Life Stress I nventory (SSI), Misr a and McKean (2000) found that practices of time management lowe red college student stress levels. However, the Misra and McKean study contradicted th e previously mentioned study by Gadzella,

PAGE 61

43 Masten and Stacks (1998) and found signi ficant positive correlations (p<.05) among constructs of stress and general anxiety among the population of full-time undergraduate students. Correlations between stress and anxi ety should be positive as they both have a worry component (Misra and McKean, 2000). Felsten and Wilcox (1992) conducted a st udy to find the impact of stress on 146 college maleÂ’s academic performance. Afte r analysis they found that stress was a detrimental effect on maleÂ’s grade point average. Shields (2001) conducted a study to examine the effects of stress and reactions to stress among persisting and nonpersisting undergraduate student s. A persister is defined as a full-time student enrolled in fall and wint er semesters, while a non-persisting student is one that enrolls in the fall semester but not the subsequent wint er semester. The group of study consisted of 220 persisters a nd 110 non-persisters and findings found that persisters had significantly higher levels of stress than nonpersis tors with no gender difference. Stress was positively associated with reactive efforts to stress, and age among persistors, but not nonpersisters This gives evidence that fu ll-time status or traditional students have developed reactive efforts to stressors, but non-tradi tional students do not develop these same reactive efforts (Shields). Finally with the use of regression, stress, age and number of credit hours taken in the se mester each had significant explanation of student reaction to stress (S hields). Recognizing that stress affects students in the classroom, it is necessary to examine the demographic variables related to stress. Demographic Variables Related to Stress Gadzella and Guthrie (1993) found that women undergraduate students tend to have significantly higher leve ls in experiencing stress of both pressure and change. Considering age, self-imposed stress wa s found to be highest among younger students

PAGE 62

44 (16 to 18 years-old) and lowest in older students (31 to 57 years old) (Gadzella & Fullwood, 1992). Moreover, young women experien ced more self-imposed stress and pressure than any age and gender combination (Gadzella & Fullwood). Measurement of Stress In the task of measuring stress, two types of instruments were found in the literature. The first type lists possible stressors relevant to the situa tion of the student and provides the research with eith er a count of stressors, or an indexed stress score with certain stressors weighted more than othe rs. Although this measur e of stress objectively attributes stress to the specifi c stressor, it ignores the intera ction of the student with the environment (Cohen, Kamarck & Mermelsei n, 1983). That is, students may attribute stress to a stressor without considering av ailable resources to react to the stress. Furthermore, for this type of stress instrument to be accurate a list of all possible stressors must be presented to the student. The second type of stress instruments meas ure perceived stress with regards to the stressor, situation and ability to react to the stressor. This type of measure broadly accounts for all stressors and provides a behavi oral construct. Howe ver, perceived stress instruments are limited as they may inadverten tly measure chronic stre ss, or other sources of stress (Cohen et al.). Research concerni ng perception of stress ha s found evidence that “students often misattribute stress to the appropriate stressor” (p. 387). Given this limitation, the measure of perceived stress is co ngruent with the definition of stress used in this study. Again, stress is not a corollary of a particular stressor, but a result of the student’s perception and reaction to an experience (Romano, 1992). To meet the objectives of this study, st ress was measured specific only to the classroom environment to which the instrume nt was administered. Again, the classroom

PAGE 63

45 environment for problem solving is described as the social context to which individuals interact with during the probl em solving process (Kirton, 2003 ). This gave reason to use an instrument that measured stress as perc eived by the student and experienced in the environment. The researcher found that Gadzella’s Stude nt Life Stress Inventory (SSI) a valid and reliable measure (Gadzella & Balogl u, 2001) of an undergraduate’s perceived classroom stress that could be adopted for use in this study. The theoretical framework for the SSI comes from Morris (1990) and treat s stress as a student experience, not just stressor variables affecting students (Gadze lla & Masten, 2005). Five stressors were identified as experiential to undergraduate students and used in the SSI including: 1) frustration–delays and failures in reaching academic goals, social acceptance and disappointing grades, 2) c onflict–one’s choices among desirable and undesirable outcomes in goal setting, 3) pressure–assessing class competition, deadlines, and overload of assignments, 4) changes–unpleas ant experiences and disruptive schedule changes and 5) self-imposed–desire to compet e, be loved by all, procrastinate finishing assignments and anxiety during tests (p. 4). Each stressor forms a construct with items using a five point Like rt scale with one signifying never and five signifying most of the time. Each stressor construct can be used i ndividually or cumulatively for determining total stress of the student. Chapter 3 further discusses how the SSI was modified to meet the needs of the objectives for this study. Summary of Learning and Stress Stress is defined as an individual’s per ception and reaction to an experience that requires an adjustment from normal life (R omano, 1992). Undergraduate students have many unique stressors including moving away fr om home, academics, finances, work and

PAGE 64

46 new relationships (Pfeiffer, 2001). Comm on academic stressors include increased workload in a course and lower than expected grades received in a course (Ross, Niebling and Heckert, 1999). Although stress in the form of anxiety was found to be detrimental to academic achievement, an optimistic attitude may allow students to positively react to stress (Aspinwall & Taylor, 1992). Ev idence was also found that young women undergraduate students tend to experience mo re stress than other undergraduate students (Gadzella & Fullwood, 1992). The measurement of stress can be categorized by instruments measuring a count of stressors or a perceived experience whic h accounts for the reaction to a situation. Both types of measurement have strengths and lim itations (Cohen et al.). This study utilized the SSI (Gadzella & Baloglu, 2001), a perceive d measure of stress which is limited in ability to determine source of stress but provides a measured level of stress in the form of a behavioral construct (Gadzella & Baloglu). The measurement of stress as an experience rather than a specific stressor was found to be more compatible with KirtonÂ’s A-I theory. Learning and Motivation Concerning KirtonÂ’s (2003) theoretical framework applied to the undergraduate classroom, motivation is identified as part of cognitive affect. Motivation and its influence on learning is a well-researched phenomena with many more questions still unanswered. Research has found that l earning is dependent on motivation, but interestingly the achievement of learning goals increases motivation to learn more (Schunk, 1991). Pace (1980) found that knowledg e gained from a course was positively correlated with the amount of effort st udents provided for learning. Theories of motivation in education that are considered pr ominent at the time of this writing can be grouped by expectancy-value theo ries, attribution theories, so cial cognitive theories, goal

PAGE 65

47 theories, and intrinsic-extrinsic theori es, (Pintrich & Schunk, 2002). Although, these theories each have their strengths, this st udy used an expectancy-value theory for the motivation theoretical framework (Eccles, 1983). Expectancy-value theory was found to be compatible with KirtonÂ’s (2003) definition of the environment. For example, a student may perceive himself as capable (expectancy) of solving a problem and receive some kind of reward (value) from execution of the solution. However, the faculty member also provides feedback on the studentÂ’s effort and solution to which the student will react. If the studentÂ’s perceived ability and reward are great enough, the student will be able to solve a problem despite cognitive style gap (Kirton, 2003). Furthermore, there is evidence that expectancy beliefs are related to student engagement and value beliefs are related to opportunities of future stude nt engagement (Pintrich & Schunk, 2002). The following is a short description of the expectan cy-value theory with empirical studies to support its importance to learning. Expectancy-Value Theory Current expectancy-value models have been adapted from AtkinsonÂ’s (1957) expectancy-value theory of achievement motivation and now provide the clearest explanation to studentsÂ’ motivation in the classroom (Pintrich & Schunk, 2002). However, much is still to be learned about motivation as no motivation theory has been universally accepted by educational psychologi sts. The premise of expectancy-value theory is that the learner is motivated by tw o components: 1) their perceived ability and effort expected to be needed in accomplis hing a goal (expectancy) and 2) the amount of value or reward the learner places on a ttaining the same goal (value). These two components will be further broken down and discussed theoretica lly and empirically.

PAGE 66

48 The expectancy component In this study the expectancy component c onsists of control of learning beliefs and self-efficacy. The belief is that a positive in crease of an expectancy component will promote learning. A studentÂ’s beli ef of their control in learni ng, or locus of control, is a main component of many expectancy-value th eories. Students who believe they are in control (internal) of the l earning succeed over students w ho explain away their control (external) of learning to other factors (Eccles & Wigf ield, 2002). Research has found some evidence that control of learning beliefs change from externally to internally while attending college, with the gr eatest change found in student s involved in honors programs (Pascarella & Terenzini, 1991) Research conducted to unders tand control of learning beliefs was prominent between the 1960Â’s and 1970Â’s. After this pe riod of time authors moved on to answer other questions concer ning motivation. Still, control of learning belief is an important part of many expectan cy-value theories and the measurement of motivation. Self-efficacy (Bandura, 1977) is accepting one has the capability to complete a task or goal. Bandura (1994) states the four sources of self-efficacy include mastery of challenging experiences, vicarious observati on of a peer, verbal praise and reducing stress. As it relates to so cial-cognitive theory (Bandura, 1986), self-efficacy affects behavior and environment, but is also an effect of behavior and environment. Furthermore, there is evidence that self-e fficacy may influence the studentÂ’s choice of which problem to solve (Schunk, 2000). Referring to expectancy-value motivation in the context of student engagement, student self -efficacy may affect a studentÂ’s choice to solve a problem while expectations of the st udent and the teacher raise the value placed on the solution.

PAGE 67

49 Research has shown that higher levels of self-efficacy have a significant effect on studentsÂ’ persistence in comp leting a challenging task, acad emic performance and selfregulated learning (Zimmerman, 2000). Li nnenbrink and Pintrich (2003) reviewed several studies regarding self-efficacy and cl aim a student with high self-efficacy will be more cognitively engaged through the use of higher order th inking skills and metacognitive strategies. In other research, AbouserieÂ’s (1994) findings supported the negative relationship between self-esteem and the variables of stress and anxiety. He elaborated that the major cause of test anxiety is the possibility of being found incompetent; a bash to a learnerÂ’s self-efficacy. The literature indicates there is a domain specific nature to studentsÂ’ self-efficacy. For example, academic self-efficacy was found to account for 21% of the variance in predicting studentsÂ’ course grades (Choi, 2005); however global self-efficacy was not found to be a significant predictor of academic achievement (p. 202). Many authors have examined self-efficacy, self-esteem, and locus of control and implications these traits have on learni ng. Though each of these expectancy components have their own attributes (Ajzen, 2002) they have many of the same core constructs and may be more alike than different (Judge, Erez, Bono & Thoresen, 2002). The value component The value components of expectancy-val ue theory consists of intrinsic motivation, extrinsic motivation and task value constructs, al l of which assign reward on attainment of a goal. The belief regard ing the relationship between these value components and learning is simply that an in crease of value in completing a learning goal leads to an increase in learning.

PAGE 68

50 Intrinsic motivation is defined as accomp lishing a task for no reward except for internal feelings that result from completi ng a task (Deci, 1975). Extrinsic motivation is accomplishing a task for the benefit gained from a separate outcome (Deci & Ryan, 1985). Although intrinsic and extrinsic motivation is separate and distinct there is some evidence that the two operate on a co ntinuum. For example, a young boy may be extrinsically motivated to learn how to play the violin as his music teacher rewards successful learning with grades, extra-credit points or time off from their regularly scheduled practice. As time moves on, the boy starts to find enjoyment in playing the violin and will play more just for the sake of playing. The boy moved from being extrinsically motivated to intrinsically motiv ated. This was empirically tested by Ryan and Connell (1989) and concluded that th is continuum does exist by examining relationships among different types of extrinsic motivati on and change in specific attitudes. Generally in the literature extrinsic motiv ation has been considered detrimental to learning and the level to which students solve problems. However, extrinsic motivation can be beneficial to learning if used in the correct manner. For example in problem solving, if a student is told how a solution will be eval uated and rewarded, extrinsic motivation positively influences engagement in the problem solving process (Collins & Amabile, 1999). Furthermore, an extrinsic rewa rd offered to a student before beginning the problem solving process may increase le vel of problem solving if there is no contingency contract (Collins & Amabile). In another study, Hennessey and Zbikowski (1993) attempted to educate students about th e negative effects of extrinsic motivation. The results found that after these students were extrinsically rewarded to solve a problem,

PAGE 69

51 there was no negative influence from the extrinsic reward. More research is needed to understand the positive influences of extrinsic motivation, especially if students are often motivated extrinsically and gradually become more intrinsically motivated over time. A review of the literature finds disagreement in the definition of task value motivation and how it is utili zed in different expectancy-v alue motivation models. Task value motivation sometimes incorporates intrinsic and extrinsic interests and is sometimes referred to as incentive mo tivation (Pintrich & Schunk, 2002). Other researchers refer to task value as a level of importance placed on a task (Battle, 1965) which is how task value is defined in this study. Considering the importance of task value to learning, Feather (1988) found task value to be specific to a college studentÂ’s academic major. For example, a student with a high task value in math had higher perceptions of their ability in math, as well as an increase of personal values concerning achievement in math. Affective motivation Affective motivation is not a component of expectancy value theory, but is becoming more common in recent models of mo tivation as researchers explore the role of emotion in motivating students. It is incl uded here because the measure of motivation chosen by the researcher included an affec tive component identified as test anxiety. Throughout the literature, there is evidence that test anxiety is detrimental to academic achievement and learning (Pintrich, Smith, Garcia & McKeachie, 1991). The two main components of test anxiety are worry and em otion. Some evidence suggests that emotions cause worry making test anxiety a behavior al construct (Hembree, 1988). Treatment of studentsÂ’ test anxiety is best provided by reducing worry and the emotionality associated with the test. This has shown to in crease academic performance (Hembree).

PAGE 70

52 Demographic Variables Related to Motivation Motivation is an unstable consequence of the situation perceived by the individual and therefore considered cont extual or situati onal (Paris & Turner, 1994). Motivation continues to change as a learner assesses and reassesses their abil ity to accomplish the task and the value of the reward upon acco mplishing the task. Demographic variables such as age, gender, ethnicity, etc. are alwa ys factored in the individualÂ’s assessment of the situation and their capability of succe ss in the situation. The environment of the classroom does allow for all learners to be in the same situ ation, but the learner continually constructs motivation (Paris & Turner) making motivation highly individualized and difficult to gene ralize to specific demographics. Measurement of Motivation The wide variety of motivational theori es has led to the development of many instruments to measure motivation. To meet the objectives of this study and to remain consistent with expectancy-value theor y, the Motivated Strategies for Learning Questionnaire (MSLQ) (Pintrich, Smith, Ga rcia & McKeachie, 1991) was used to determine motivation of each student specific to the class to which the instrument was administered. The MSLQ was designed specificall y to measure undergraduate student motivation for self-regulated learning. The th ree components of the instrument include an expectancy component with items related to student beliefs of ability in performing a task, a value component with items concerni ng goals and reasons for completing a task, and an affective component which includes ite ms regarding emotional reactions to the task (Pintrich & De Groot, 1990).

PAGE 71

53 The three components of the MSLQ can be further broken down into six scales of motivation that can be used independently or cumulatively (Pintrich, Smith, Garcia, McKeachie, 1991). The value component co mprises three scales: intrinsic goal orientation–participating in a task for reasons of challenge, curiosit y or mastery, extrinsic motivation–participating in a task for good grades, rewards or competition and task value–an evaluation of the importance of the task (Pintrich et al.). The expectancy component of the MSLQ has two scales which include control of learning beliefs–a belief that results hinge on self and effort will provide positive results, and self-efficacy for learning and performance–an appraisal and judgment of ability to complete a task (Pintrich et al.). The affective component has one scale identified as test anxiety which includes student performance and the resulting emotion (Pintrich et al.). It is important to note that the MSLQ includes an affective co mponent in measuring student motivation, but there is no affective component in expect ancy-value theory. A ffective motivation, the motivation scales that identify emotions, has increasingly been recognized in the literature during the last 20 years and found to be incorporated into more cognitive motivation models. More research is needed to understand how emotions play a part in student motivation to learn. In a study conducted to determine how th e components of the MSLQ relate to cognitive engagement of students, Pintrich and De Groot (1990) found that self-efficacy and intrinsic orientation were both positively correlated with cognitive engagement and academic performance. Test anxiety was not found to be correlated with cognitive engagement, but did have a significant ne gative correlation with self-efficacy. The affective scale of test anxiety was found to be either detrimental or show no relationship

PAGE 72

54 with effort given to the task at hand (Pintr ich & De Groot). Given that test anxiety is composed of worry and resulting emotion, one can easily see the similarities between stress and test anxiety. Furthermore, positive re actions to test anxiety may have the same effect that they have on student stressors. Chapter 3 gives more detail as to how the MSLQ was used to meet the objectives of this study. Summary of Learning and Motivation Expectancy value theories of motiva tion provide insight to how students are motivated to achieve academic success (P intrich & Schunk, 2002). This theory of motivation was established as compatible w ith KirtonÂ’s (2003) c ognitive affect as it relates to the environment in which indivi duals interact and react during the problem solving process. The expectancy component of the motivation theory refers to a perceived ability in accomplis hing a learning goal while the value component refers to the reward placed on accomplishing the same goal (Eccles, 1983). As a student consistently assesses and reassesses the ab ility to accomplish this goal and the value placed on achieving success, motivation becomes situational and indivi dualized (Paris & Turner, 1994). Evidence supports the assertion that mo tivation is necessary for learning and motivation increases after a learning goa l is achieved (Schunk, 1991). Research has indicated that student engage ment was positively correlate d with measures of selfefficacy and intrinsic motivation (Pintr ich & De Groot, 1990). Although extrinsic motivation was considered in the literature as detrimental to learning and student engagement, extrinsic motivation was found to positively influence engagement in the problem solving process when used without a contingency contract (Collins & Amabile,

PAGE 73

55 1999). More research is warranted to dis tinguish between the harmful and useful practices of providing extrinsic motivati on to facilitate st udent engagement. This study measured student motivation wi th the MSLQ (Pintrich, Smith, Garcia & McKeachie, 1991) which included the following motivation scales: control of learning, self-efficacy, intrinsic motivation, extrinsic motivation, task value and test anxiety. Although test anxiety was not part of expectan cy-value theory, it was a scale found in the MSLQ and was included in the study given that affective motivational theories have been increasingly found in the literature. Learning and Engagement Astin (1984) developed a theory of stude nt involvement to help guide future research in the area of student developmen t and learning in completing an undergraduate degree. In his writing, Astin explained th e tie between student engagement and motivation as “sharing a common psychologi cal construct” (p. 522). Realizing the importance of motivation, he also discussed st udents’ time as a resource. Astin’s theory states that student effort and time are two resources required for student engagement in learning. That is, an increase in effort and time by the student to accomplish a learning goal will increase student learning. Astin conc ludes that “students’ time and effort are limited and exhausted” (p. 523) on activities re lated to college and also life. Astin’s theory also identifies contributors to stude nt engagement, which guided the development of the National Survey for Student enga gement (NSSE), the measure of student engagement in this study. The roles that faculty members play as advisors and teachers have incredible influence on undergraduate students. Even in informal situations, faculty members can become positive role models for students, in fluencing attitudes and values (Pascarella,

PAGE 74

56 1980). The importance of the faculty member ca nnot be denied in contributing to student engagement in the classroom and therefore facilitating academic success. The study of student engagement involves identifying the practices used by faculty members to increase learning and personal development a nd then surveying students to determine if those practices are being used. The following is a review of the lite rature regarding how the faculty member can influence learning. The Influence of Faculty Members Umbach and Wawrzynski (2005) conclude th at “the single most important” factor in student development and classroom engagement is the faculty member. Whittington (1998) however, has found that faculty members in colleges of agriculture typically teach with methods that only require lower le vel thinking skills of the students and is detrimental to student classroom engagement However, many colleges and universities now provide resources to help faculty members improve their instructional discourse. A replication of Whittington’s study is needed to determine if faculty me mbers still teach at lower cognitive levels. In a study conducted to examine the re lationship between student-faculty interaction and students’ critic al thinking skills, Smith’s (1977) findings provide evidence that there is a positive relati onship. Specifically, increases in critical thinking ability were positively correlated among variables of class pa rticipation with the faculty member as well as participation with other st udents in the class (Smith). Increasing student engagement is more than just keeping students busy in the learning process. Schlechty (2002) claims that the amount of tim e a faculty member dedicates to improving the quality of assi gnments is directly related to student engagement. Many studies have been conduc ted examining different instructional

PAGE 75

57 techniques, distance education methods of instructional delivery, and uses of technology and their impact on student academic achievement and engagement while controlling for learning style. Most of these studies conclude that student s have different styles of learning and that teachers should use a variet y of teaching approaches to meet learning style needs of all students. As most good teach ers incorporate a variety of instructional techniques already, the signifi cance of learning style theo ry has found little practical application in the classroom (Coffiel d, Moseley, Hall & Ecclestone, 2004). The literature revealed similar conclu sions comparing the problem solving teaching approach to the content based appr oach. Dyer and Osborne (1996) examined the use of the two teaching approaches in a quasi-experimental design studying secondary students enrolled in agriculture. Using th e Group Embedded Figures Test (Witkin, Oltman, Raskin & Karp, 1971) to measure lear ning style, they found that controlling for teaching approach, neither field independent or field dependent students had significantly higher achievement scores. However, they also found that field independent learners had higher problem solving scores on non-agric ulturally based problems. Though their findings were limited, they also concluded that different instructional strategies should be used to facilitate the learning of all students (Witkin et al.). Pascarella and Terenzini (1991) summari ze teacher/teaching attributes that are effective for student learning. They include: command of subject matter, enthusiasm of subject, clear explanation of concepts, effici ent structure of class time, use of examples and analogies in presenting new concepts avoiding vagueness, and having good rapport with students (Pascarella and Terenzini).

PAGE 76

58 Realizing the tie between motivation and student engagement, Paris and Turner (1994) identified four teaching techniques to increase student motivation to learn, which include: allowing students to choose a course of action, pr oviding academic challenge, giving students control of achieving the le arning objective, and allowing students to collaborate in the problem solving process. Demographic variables related to engagement Pascarella and Terenzin i (1978) conducted a study to examine the academic performance and social development of stude nt-faculty informal interaction (N=528). Concerning the relationship betw een gender, academic major, ethnicity and parentsÂ’ level of education, no significant correlati ons were found among measures of academic performance, personal development and in tellectual development. The study was replicated in 1980 and conclusions were c onsistent (Terenzini and Pascarella, 1980). However, there is evidence that student engagement may be related to student demographics. For example, a female stude nt may be more influenced by a woman faculty member than by a male faculty memb er (Pascarella & Terenzini, 1991). Pike (2004) measured engagement with the NSSE and found that females exhibit a higher level of academic challenge and students of diffe rent disciplines have different levels of engagement. Pike did not find studentsÂ’ age a correlate of student engagement when examining students over that age of 24. Howeve r, as students progres s in attaining their degree, student engagement increases (Kuh, 2001). That is, as studentsÂ’ classification moves from freshman to senior, student s tend to be more engaged in learning. Measurement of Engagement The operational framework of this study places student engagement as a product or end result of the intera ction between student and f aculty member behavior. Many

PAGE 77

59 studies of learning styles use final grade or academic achievement as the dependent variable in order to demonstrate the practic al contribution that a specific learning style has in the classroom. However, as underg raduate classrooms become more student centered (Acharya, 2002; Lunde, Baker, Buelow & Hayes, 1995; Study Group on the Conditions of Excellence in American Higher Education, 1984), there is greater need to identify variables related to student engageme nt. Student engagement is often determined with constructs of academic challenge, collaborative learning and student-faculty interaction (Kuh, 2001). Each construct is de pendent on the teaching practices of the faculty member and requires communication between the faculty member and student (Umbach & Wawrzynski, 2005). If a dissim ilar learning style between the faculty member and the student inhibits communica tion during the process of learning, evidence should be found in the variance of student enga gement scores. That is, a large cognitive style gap may explain why students do not show evidence of engagement in the classroom. Also, if a faculty member uses subjective evaluations (rubrics) of student projects and assignments–which are commonly used to evaluate the solution to a problem–the final course grade is also s ubjective and biased. That is, a faculty member may prefer solutions to assigned problems that are congrue nt with their learning style. The researcher found no evidence in the literat ure concerning instructors’ evaluation biases based on learning style, but that does not suggest that this may exist. Finally, the use of academic achievement as a dependant variable, whether in the form of a percentile or le tter grade assigned by the faculty member, may subject the study to measurement error. There is evidence that faculty members inflate students grades for

PAGE 78

60 the purpose of receiving tenure and prom otion (Germain & Scandura, 2005). Furthermore, student engagement is highly co rrelated with grade point average and other measures of academic achievement (Kuh, 2001). Given the previous discussion, the researcher found student engagement a more objective and accurate measure of learning and interaction between the student and faculty member. A remarkable increase in the study of st udent engagement in the undergraduate classroom came after the release of Involvement in Learning (Study Group on the Conditions of Excellence in American Highe r Education, 1984), which called for the use of teaching strategies that increase student learning in higher e ducation. Since then, researchers have sought to id entify empirically based practi ces that impact learning and develop effective measures to determine st udents’ level of engagement. The National Survey of Student Engagement (NSSE) was designed to assess student engagement among colleges across the nation. Scales of the NSSE form “benchmarks” in order that colleges can compare the teaching practices of their faculty members to the teaching practices of other faculty members acro ss the nation (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). The instrument was f ound to be stable and psychometrically solid with evidence of validity and reliabi lity (Kuh, 2001). Additionally, the NSSE has been positively correlated with undergraduate student grade point average, critical thinking, quality of learning and GRE scores (Kuh). The NSSE was developed through an exte nsive literature re view of classroom activities and teaching practices that lead to valuable student outcomes (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Three of the five benchmarks measured by the NSSE were specific to learning and engagement in a classroom. They include

PAGE 79

61 academic challenge, active learning and student-faculty interaction. Two other benchmarks measured engagement releva nt to the college, including enriching educational experiences and supportive cam pus environment. As the latter two benchmarks were relevant to the college and not the classroom, they were not utilized in this study. Said differently, they were not measuring engagement practices directly attributed to the studen t and the faculty member. The NSSE benchmarks of interest are described as follows: level of academic challenge–time spent preparing for class, amount of reading and writing, and use of higher order thinking skills; active lear ning–participating in class and working collaboratively with other students inside and outside the classr oom and student-faculty interaction–talking with the faculty member, discussing ideas outside of class with the faculty member, getting feedback from the faculty member, and conducting research with the faculty member (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001, p. 5). More detail of the NSSE is provided in Chapter 3. Summary of Learning and Engagement Student’s have a limited amount of time and effort to apply in accomplishing a learning goal (Astin, 1984). However, facu lty members may be the “single most important” factor in engaging the student to learn. Although the student is responsible for engaging in learning, the faculty member is accountable for facilitating student engagement. The study of student engagement primarily entails identifying teaching practices that increase student engagement and surveying students to determine if teaching practices are being used. Many researchers have examined the relationship learning styles have with academic achievement. However, an emphasi s on increasing student engagement (Lunde,

PAGE 80

62 Baker, Buelow & Hayes, 1995) gave reason to examine if dissimilar learning styles between faculty member and student can contri bute to the literatu re regarding student engagement. Furthermore, student engagement as measured by the NSSE may serve as a more objective dependent variable than acad emic achievement. Student engagement is closely tied to learning and academic achievement (Kuh, 2001). Summary Chapter 2 explains KirtonÂ’s Adaption -Innovation theory as applied to the undergraduate classroom providing a theore tical framework for this study. The framework was derived from KirtonÂ’s Adap tion-Innovation theory which may provide insight to how dissimilar c ognitive styles explain stude nt engagement. Other major sections of this chapter included: Learni ng and Learning Styles, Learning and Stress, Learning and Motivation, and Learning and Engagement. This review of the literature found tie s between problem solving learning and KirtonÂ’s measure of cognitive style. Although KirtonÂ’s cognitive style has been applied in many contexts including business and manage ment of change, little research has been conducted to examine if it is appropriate in explaining student engagement in the undergraduate classroom. There is a need for more research consid ering the variable stress as it relates to learning and student engagement in the underg raduate classroom. Specifically, why do more adaptive courses have higher levels of stress and how does this contribute to student engagement? However, research has consis tently found motivation to significantly contribute to learning, yet fe w studies have examined both stress and motivation while controlling for an instructor with a dissim ilar learning style from that of students.

PAGE 81

63 Student engagement in a course is larg ely dependent on teachi ng practices of the faculty member. The amount of structur e preferred by faculty members can be determined by identifying their cognitive style. However, students may have different cognitive styles than the faculty member a nd prefer either more or less structure in learning from problem solving. As student e ngagement is closely tied to studentsÂ’ academic success, it is critical to consider if dissimilar cogni tive styles between students and the faculty member can explai n variance in student engagement.

PAGE 82

64 CHAPTER 3 METHODS Chapter 1 introduced the theory of learni ng styles, the appropriateness of KirtonÂ’s Adaption-Innovation (A-I) theory for the classr oom, and the need for increasing student engagement in undergraduate classrooms. A point was also made that the literature contains many types of learning styles with varying degrees of re liability and validity. However, KirtonÂ’s (2003) measure of learning st yle is valid and reliab le and distinct from cognitive level. Finally, definitions of key terms, purpose of the study, and objectives of the study were presented. Chapter 2 presented the theoretical and conceptual frameworks of the study. Included were the following topics with atten tion given to theory a nd empirical evidence: a) learning and learning styles b) learning and stress, c) learning and motivation, and d) learning and student engagement Theoretical evidence was pr ovided that KirtonÂ’s (2003) measure of learning style is compatible w ith problem solving learning (Gange, 1965) and may provide insight to why some students lack classroom engagement. This chapter provides information concer ning the methodology of the study. More specifically, details were gi ven concerning the research design, target population, instrumentation, data collection procedur es, and data analysis procedures. The purpose of this study was to dete rmine if relationships exist between cognitive style gap, student stress, student motivation, student engagement and selected demographic variables of undergraduate student s at the University of Florida in the College of Agriculture and Life Sciences. The objectives of the study were to: 1)

PAGE 83

65 Describe selected faculty and students accordi ng to their selected demographic variables; 2) Determine the cognitive style, studen t stress, student motivation, and student engagement of undergraduate students; 3) Determine the cognitive style gap between faculty and students and to expl ore its relationship with stude nt stress, student motivation, student engagement and specific demographi c variables of undergraduate students; 4) Explain undergraduate student stress and st udent motivation based on cognitive style gap and selected student demographic variable s and 5) Explain undergraduate student engagement based on cognitive style gap, student stress, student motivation and selected student demographic variables. Research Design The researcher used an ex post facto design one (Ary, Jacobs & Razavieh, 2002) to accomplish the objectives of the study. This design allows for the control and measure of the independent variable(s) to test hypot heses concerning varia tion in the dependent variable(s). Cognitive style is inherent, stable, and measurable as an interval score along a continuum allowing for the precise measure of dissimilarity between two individualÂ’s cognitive styles. The literature gave little gu idance concerning A-I th eoryÂ’s application to the context of the undergraduate classroom concerning relationships with dependent variables. Thus, this empirical desi gn was deemed most appropriate. For objective one, student demographic variab les were measured with items in the National Survey of Student Engagement 2005 (NSSE) (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001) and included: age, gender, major, college classification, fulltime status, and number of classes taken with si milar content as the class. Lastly, students were asked to give the number of problem sets that had taken more than an hour to complete during a typical week. This question wa s not used in data analysis, but was used

PAGE 84

66 to validate that problem solving does exist in the classes used in this study. Faculty members were asked questions regarding age, gender, years of teaching experience, and academic department. Faculty members also pr ovided the researcher with syllabi and additional information to descri be the course. Finally, partic ipating faculty members were administered KirtonÂ’s Adaption-Innovation I nventory (KAI) (Kirton, 1999) to determine their cognitive style. Descrip tive statistics were used to report determined values. For objective two, four instruments were administered to nine undergraduate student classes to determine each studentÂ’ s cognitive style, stress, motivation and engagement in their respective course. The KAI provided a measure of each studentÂ’s cognitive style. Level of student stress e xperienced in the classroom was determined using the Student-life Stress Inventory (SS I) (Gadzella & Baloglu, 2001). The Motivated Strategies for Learning Questionnaire (MSL Q) (Pintrich, Smith, Garcia & McKeachie, 1991) was administered to m easure student motivation in the classroom. The NSSE was used to determine student engagement. De scriptive statistics were used to report determined values. For objective three, cognitive style gap was calculated by subtracting faculty membersÂ’ cognitive style scores from each studentÂ’s cognitive style scores as measured by the KAI. The same procedure was used to determine gaps within the constructs of cognitive style. These calculated values were used to find correlations with constructs of stress, motivation, engagement and selected student demographic variables. Objective three was accomplished us ing bivariate correlation. For objective four, depende nt variables including student stress, student motivation, and student engagement were each separately explained by the independent

PAGE 85

67 variables including cognitive st yle construct gaps and stude nt demographic variables. Objective four was accomplished using backward stepwise multiple linear regression For objective five, independent variables including cognitive style construct gaps, student motivation, student stress and selected student demographic variables were used to explain the dependent variable student engagement. Objective five was accomplished using backward stepwise multiple linear regression. Population This study utilized nine purposefully chosen class groups in the College of Agriculture and Life Sciences at the University of Florida. Since direction and degree of cognitive style gap was dependent on the cogni tive style of the faculty member, courses were selected based on a faculty memberÂ’s cognitive style: a) three faculty members with similar adaptive cognit ive style scores, b) three faculty members with similar cognitive style score approximating 95, and c) three faculty memb ers with similar innovative cognitive style scores. Researchers have found little evidence th at class size influences undergraduate student achievement (Williams, Cook, Qui nn & Jensen, 1985). However, Cuseo (2004) found the optimum class size for student enga gement at approximately 15 students with no evidence provided in the literature to s upport a linear or asympt otic relationship. The literature supports small class sizes for the retention of knowledge, student participation, and critical thinking (McKeachie, Chism, Menges, Svinicke & Weinstein, 1994), but there is no agreement on what number of student s constitutes a small class. Most studies consider 50 or more students as a large class (Cuseo, 2004). A sample size of 46 is needed to attain a power of .70 ( =.05, R2=.20) with the use of four independent variables (Ary, Jacobs & Razavieh, 2002) The researcher found from previous

PAGE 86

68 experience of administering the KAI to stude nts at the University of Florida that mortality rate approximates 40%, which is lower than KirtonÂ’s (1999) estimated 20% mortality rate for this type of population. The researcher determined a cognitive style score as unusable using guidelines provided by Kirton (1999). Given the high mortality rate of the KAI, the researcher sought out classes with nearly 65 students. However the researcher found that many courses in the Colleg e of Agriculture and Life Sciences at the University of Florida were either belo w 50 students or approximated 100 students. Therefore the researcher c hose classes that approximat ed 100 students realizing the limitations of large class sizes having less student engagement (McKeachie et al., 1994). Finally, instructional techniques employe d by the faculty members often promote student centered classr ooms providing for student engage ment (McKeachie et al.). For this reason, a review of pros pective course syllabi and as signments was conducted by the researcher to purposefully select courses c onducive for student engagement. Specifically, the researcher examined these documents fo r evidence that courses required students to attend class in order to be successful in the course. The researcher was not concerned with a percentage of the final grade requiri ng attendance, but examined if assignments and projects provided reason for students to attend class. Fu rthermore, since the KAI is a measure of problem solving style, courses selected for this study required the use of problem solving assignments. A problem was defi ned in this study as a desire or felt need for a solution, but not immediately knowing the mental operations to arrive at the solution (Soden, 1994).

PAGE 87

69 Instrumentation Four instruments were utilized to collect the data and complete the objectives of the study. The following is a description of the instrument and its constructs, as well as assessments of validity, and reliability. Kirton Adaption-Innovation Inventory The Kirton Adaption-Innovation Inventor y (KAI) was developed by Kirton (1976) as a measure of cognitive style unrelated to cognitive ability and cognitive affect. Appendix D provides example items concerni ng the KAI and contact information to acquire permission to use the instrument. The KAI is a self-r eported instrument consisting of 33 items, with one item not rele vant in determining cognitive style. In distinguishing more adaptive individuals from more innovative individuals, items delineate a theoretical total range of 32 (highly adaptive) to 160 (highly innovative) (Kirton, 1999). However, the actu al total range is approximate ly 45 to 145 with a general population mean of 95 (Kirton). For possible ranges of cogn itive style construct scores, Kirton reported the following: sufficiency of originality (17 to 63), efficiency (7 to 33) and rule/group conformity (14 to 56). Individuals taking the KAI were asked “How easy or difficult do you find it to present yourself, consistently, over a long period as:” followed by a list of personal characteristics. Individuals then mark an “X” on a continuum of “Very Hard” and “Very Easy”. Construct scores were calculated by summing respective items and total scores were calculated by summing construct scores. The KAI measured the cognitive style of the students and faculty members through the constructs of suffic iency of originality, efficien cy, and rule/group conformity. Cognitive style gap was then calculated using the difference of total and construct

PAGE 88

70 cognitive style scores of the faculty member from the student. Reliability of the KAI was determined with Cronbach’s alpha coeffici ent (Kirton, 1999). Validity of the KAI was established through correlations with personali ty assessments (criteri on validity), lack of correlations with assessments of cognitive leve l (criterion validity), and factor analysis (construct validity) (Kirton). Th e researcher is a certified practitioner of the KAI. See Table 3-1 for construct and reliabi lity information concerning the KAI. Table 3-1. Construct Reliability of Kirton Adaption-Innovation Inventory Construct No. of items Mean SD Sufficience of originality Efficiency Rule/Group conformity Total KAI 13 7 12 32 41 19 35 95 9 6 9 18 .83 .76 .83 .88 Note. Source: (Kirton, 1999). Student-life Stress Inventory The second instrument administered to students was the Student-life Stress Inventory (SSI) developed by Gadzella (1994). Appendix E provides example items concerning the SSI and contact information to acquire permission to use the instrument. The SSI is a self-reported inst rument designed for college st udents and comprised of two parts: perceived stress (5 scal es, 23 items) and reactions to stress (4 scales, 28 items). Items were in the form of statements of whic h participants responded to five-point Likert scales with numbers signifying responses of “never”, “seldom”, “occasionally”, “often”, and “most of the time”. Construct scores of stressors were calculated by summing the items per each scale and include frustrations conflicts, pressures, changes, and selfimposed. The total stress scor e had a possible range from 23 to 115 and possible ranges of constructs were frustration (7 to 35), conf licts (3 to 15), pressure s (4 to 20), changes (3

PAGE 89

71 to 15) and self-imposed (6 to 30). Total stress scores were ca lculated by summing constructs scores. The researcher determined constructs related to reactions to stress including physiological, emotional, behavioral, and cogni tive appraisal, unimpor tant for this study as they measured physical and psychological responses to stress and not level of stress. Therefore, these were removed from the instru ment prior to data collection. Furthermore, item number six was removed from the frustra tion construct as it was not related to stress between the faculty member and the student. St ems of the items were altered to include the words “in this course” to facilitate the m easurement of stress specific to the course of which the instrument was administered. The SSI was used with permission. The SSI was assessed for reliability with Cronbach’s alpha coefficients and validity through confirmatory factor analysis (construct validity) (Gadzella & Balogl u, 2001). See Table 3-2 for descriptive information regarding constructs and reliability of the SSI. Table 3-2. Construct Reliability of Student-life Stress Inventory Construct No. of items Mean SD †Total Stressors †Frustrations †Conflicts †Pressures †Changes †Self-imposed Total Reactions to Stressors Physiological Emotional Behavioral Cognitive appraisal Total SSI 23 7 3 4 3 6 28 14 4 8 2 51 70.09 17.65 8.43 14.08 8.03 21.82 67.19 31.07 12.17 11.47 6.01 137.28 11.53 3.87 1.91 2.83 2.39 3.78 13.32 8.32 3.51 4.32 1.98 20.98 .92 .70 .68 .80 .86 .63 .75 .86 .82 .71 .82 .92 Note Source: (Gadzella & Baloglu, 2001). † Signifies construct wa s used for this study.

PAGE 90

72 Motivated Strategies for Learning Questionnaire The third instrument administered to st udents was the Motivated Strategies for Learning Questionnaire (MSLQ) developed by Pintrich, Smith, Garcia, and McKeachie (1991). Appendix F provides example ite ms concerning the MSLQ and contact information to acquire permission to use th e instrument. The MSLQ is a self-reported instrument consisting of two components: mo tivation orientation (6 scales, 31 items) and use of learning strategies (9 scales, 50 items). Students taki ng the questionnaire are asked to rate themselves on seven point Likert scal e items anchored by “not at all true of me” and “very true of me”. Constr ucts of the MSLQ motivation or ientation included scales of value components, expectancy components, a nd affective components. The total possible range of the MSLQ motivationa l orientation was 6 to 42 with all construct scores ranging from 1 to 7. Scores were calculated by summing items related to the construct and dividing by the number of items. Learning strategy scales of the MSLQ include cognitive and meta-cognitive strategies, and resource management strate gies. The researcher determined that the learning strategy scales were unnecessary fo r this study and were removed from the instrument prior to data collection. The le arning strategy scales were a measure of cognitive engagement (Pintrich et al.) which is highly correlated with student engagement. Furthermore, the researcher found the items biased towards being more adaptive in respect to A-I theory. The MSLQ was specifically designed to be administered to college students enrolled in a college course and to be flexible to the needs of the researcher. The 15 scales of the MSLQ can be used either cu mulatively or individua lly. It was developed over nine years with the devel opers assessing validity through factor analysis (construct

PAGE 91

73 validity) and correlations with ot her instruments and final grad e scores (criterion validity) (Pintrich et al., 1991). Internal reliabili ty was assessed by th e developers with Chronbach’s alpha coefficients (Pintrich et al.). The MSLQ was used with permission. See Table 3-3 for information describing the constructs and reliability of the MSLQ. Table 3-3. Construct Reliability of Motivat ed Strategies for Learning Questionnaire Construct No. of items Mean SD †Total Motivation Scale Value †Intrinsic goal motivation †Extrinsic goal motivation †Task value Expectancy †Control of learning beliefs †Self-efficacy for learning Affective †Test anxiety Total Learning Strategies Scale Cognitive and meta-cognitive strat. Rehearsal Elaboration Organization Critical thinking Meta-cognitive self regulation Resource management strategies Time and study environment Effort regulation Peer learning Help seeking Total MSLQ 31 4 4 6 4 8 5 50 4 6 4 5 12 8 4 3 4 81 30.46 5.05 5.03 5.54 5.74 5.47 3.63 39.13 4.53 4.91 4.14 4.16 4.54 4.87 5.25 2.89 3.84 69.59 1.41 1.23 1.25 0.98 1.14 1.45 1.35 1.08 1.33 1.28 0.90 1.05 1.10 1.53 1.23 * .74 .62 .90 .68 .93 .80 .69 .76 .64 .80 .79 .76 .69 .76 .52 Note. Source: (Pintrich, Smith, Garcia & McKeachie, 1991). † Signifies scale was used for this study. Signifies value was not reported. National Survey of Student Engagement The fourth instrument administered to the students was the National Survey of Student Engagement 2005 (NSSE) develope d by the National Center for Higher Education Management Systems (NCH EMS) (Kuh, 2001). Appendix G provides example items concerning the NSSE and contact information to acquire permission to use

PAGE 92

74 the instrument. The NSSE is a self-reported instrument consisting of 85 items measuring the constructs of college activ ities, educational and persona l growth, opinions about your school, educational program char acteristics and other measures of academic performance. An additional 15 items were designated to determine student demographic information. However, only 42 of the 85 construct items fo rm “benchmarks of effective educational practice” (Kuh, Hayek, Carini, Ouimet, G onyea & Kennedy, 2001) from which academic institutions calculate year ly progress towards increasing student engagement. Benchmarks of the NSSE include academic ch allenge, active and coll aborative learning, student faculty inter action, enriching edu cational experiences, and supportive campus environment. Only three benchmarks were used as a measure of student engagement in this study, including academic challenge, active and collaborative learning, and student faculty interaction. The benchm arks of enriching educati onal experiences and supportive campus environment were not applicable to this study as many items were relevant to the campus and not the classroom. Stems of items we re altered to include the words “in this course” in order that students answer items specific only to the class the instrument was administered. All items were f our-point scale items except fo r one specific to the number of hours spent on homework; it was a seven-poin t scale and changed to a four-point scale to better determine the number of hours spen t preparing specifically for the course used in this study. Anchors of the scale typi cally resembled “very often”, “often”, “sometimes”, and “never”, while other items were anchored with numbers representing frequencies of comple ting a specific task.

PAGE 93

75 Possible benchmark ranges included academic challenge (11 to 44), active and collaborative learning (7 to 28) and student faculty interaction (6 to 24). Benchmark scores were calculated by summing items rela ted to each construct. In this study these three benchmark scales were added to form a total possible engagement range of 24 to 96. However, since two NSSE benchmark scales were removed in th is study, this total score range does not correspond with th e total range reported by the NSSE. Validity of the NSSE was assessed by the NSSE design team through welldefined items (face validity), research of the literature (content valid ity), factor analysis (construct validity), and review of the collected data (c oncurrent validity) (Kuh, 2001). Reliability was assessed utilizing Spearman’s rho correlations between yearly tests (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). The NSSE was used with permission. See Table 3-4 for a description of cons tructs and reliability of the NSSE. Table 3-4. Construct Reliability of National Survey of Student Engagement Construct No. of items Mean SD Spearman’s rho †Academic challenge †Active and collaborative learning †Student-faculty interaction Enriching educational experiences Supportive campus environment Total NSSE 11 7 6 12 6 42 33.18 17.92 12.64 15.57 23.06 102.37 * * * .77 .76 .83 .89 .85 .83 Note. Source: (Kuh, Hayek, Carini, Ouimet, G onyea & Kennedy, 2001). Values reported regarding the senior populat ion of college students. † Signifies scale was used for this study. Signifies value was not reported. Data Collection Procedures The Institutional Review Board (IRB) at the University of Florida conducted a formal review of the research proposal fo r ethical soundness. The research proposal was approved and assigned the I RB protocol number 2006-U-0078.

PAGE 94

76 Before the administration of instruments to collect data, all participants received a letter of informed consent which was also re ad by the researcher. Th e letter of informed consent described the study, identified the res earcher, addressed potential risks associated in the participation of the study and approxima te amount of time expected to complete the instruments. Participants were informed of the voluntary nature of the study with the opportunity to quit the study at any time with no consequences. Individuals who volunteered to participat e in the study signed the letter of informed consent as acceptance of the terms and returned the letter of info rmed consent with the completed instruments utilized in the study. Instruments were pack aged together with corresponding codes to administer all instruments at the same time and ensure each instrument corresponded with the same student. Additionally, directions we re given orally and in written form to students to complete the SSI, MSLQ and NSSE in regards to the specific class the students received the instrument and not to ot her classes enrolled in that semester. The students were also informed that these direct ions did not apply to the KAI as it measures their style of solving problems regardless of class. Sixty-four faculty members in the College of Agriculture and Life Sciences at the University of Florida were contacted duri ng February 2006 to participate in this study. Faculty members were identified by contac ting individuals serving as undergraduate coordinator of each department in the colle ge. Undergraduate coordinators provided the researcher with contact information of f aculty members teaching courses that used problem solving assignments and had enough stude nts enrolled in the course to meet the requirements of this study. Although many time s the identifie d faculty members believed their course did not match the requirements of the study, they often provided contact

PAGE 95

77 information of another faculty member within the same department whose course better matched the requirements of this study. In th e end, 21 faculty members were willing to participate in the study, but only 15 faculty me mbers were administered the KAI as their courses were deemed suitable by the research er. From this group, nine faculty members were chosen based on their cognitive style score. The selection of faculty members provided a range along the adap tive-innovative continuum as three faculty members had similar adaptive cognit ive style scores, thr ee faculty members had similar innovative cognitive style scores and three faculty me mbers had similar cognitive style scores located in the middle of the adaptive-innovative continuum. Kirton (2003) states that total cognitive style scores less than 10 points are considered the same, between 10 and 19 points are considered similar a nd scores of 20 points or more are considered different. Each of the nine faculty members’ course s was coded with a letter to maintain confidentiality in data pres entation. In order of adaptiven ess to innovativeness faculty members were coded beginning with the letter “A” and ending with the letter “I”. The grouping of classes by faculty member cognitiv e style resulted in an adaptive teacher group (Classes A, B and C), a middle score teacher group (Classes D, E and F) and an innovative teacher group (Classes G, H and I). Student data were collected from each faculty members’ class after midterm examinations during the spring 2006 semester specifically March 27 to April 12, 2006. This time of administration was selected to ensure students had opportunity to interact with faculty members in the classroom as th e interaction of dissimilar cognitive styles causes stress (Kirton, 2003). Students were admi nistered the instruments during regularly scheduled class times, however, it was the f aculty member’s choice if the instruments

PAGE 96

78 were to be completed during class time. If not completed during class time, students completed the instruments outside of cla ss and brought them back during the next scheduled class time. Specifically, students en rolled in Classes C, D, G and H were allowed to complete the instruments during re gularly scheduled class time while students enrolled in Classes A, B, E, F and I took the questionnaires out of the classroom and brought them back completed the next re gularly scheduled class meeting time. Faculty members were given the choice to allow students to complete instruments during class time or outside of class time out of respect for the faculty membersÂ’ instructional time with students. The estimated time for students to complete all four instruments was 40 minutes. Some faculty memb ers were reluctant to give up this amount of time and would not have pa rticipated if instruments were required to be completed during class time. Authors of the instruments used in this study ga ve no indication that students must complete the instruments in the classroom. Also, it was the faculty memberÂ’s choice to offer an incentive of extra-credit to encourage student participation. This was done primarily to encourage students to come to class on the day instruments were administer ed and to prompt return of the instruments if students completed instruments outside of class. Only studen ts enrolled in Class E were not given extra-credit for partic ipating in this study, which may have influenced response rate. A limitation of this study was the lack of control for met hod of administering instruments. Method of instrument administ ration is confounded with the faculty member giving no way to defend if either completion of instruments or use of extra-credit affected studentsÂ’ scores concerning cognitive styl e, stress, motivation or engagement.

PAGE 97

79 Furthermore, differences of scores between classes were expected and supported by authors of the instruments. For example, Kirton (2003, p. 75) provides evidence that selfselected courses may be more adaptive or innovative depending on the nature of the course. If students in each course were di sproportional regarding the number of young women, there may be higher levels of stress (Gadzella and Guthrie, 1993). Authors of the MSLQ (Pintrich, Smith, Garcia, McKeachie, 1991) claim that “responses to the questions might vary as a function of different courses” (p. 5). Finally, student engagement is an outcome of teaching practices and may va ry with individual differences found among faculty members (Kuh, Hayek, Carini, Ou imet, Gonyea & Kennedy, 2001). Given the above reasons, the data were not analyzed to examine the effect of different types of instrument administration. Data Analysis Procedures Data were analyzed according to th e following objectives and presented in chapter 4 with respect to study objectives and classes. Research Objective One Research objective one was to describe selected faculty and students according to their selected demographic variables. Stude nt participants were asked demographic questions found on the NSSE including age, gender, college classification, full-time status, number of similar courses taken and ma jor. Student participants were also asked how many problem sets were assigned during a typical week that took more than an hour to complete. Faculty participants were as ked to respond to demographic items including age, gender, and years of teaching experience. To achieve objective one, data were entered into SPSS statistical software for Windows and an alyzed using descriptive statistics (frequencies and m easures of central tendency).

PAGE 98

80 Research Objective Two Research objective two was to determine student cognitive style, student stress, student motivation, and student engagement. Student participan ts were asked to respond to the four self-reported in struments: the KAI, the SSI, the MSLQ and the NSSE. All instruments were hand-scored by the resear cher. To meet objec tive two, data were entered into SPSS statistical software for Windows. The researcher used descriptive statistics (frequencies and meas ures of central tendency) to analyze the data. For further analysis, tests of significance were used to compare groups. Research Objective Three Research objective three was to determine the cognitive style gap between faculty and students and explore rela tionships with undergraduate student stress, motivation, engagement and selected student demographi c variables. Cognitive style gap was found by subtracting each faculty memberÂ’s total cognitive style score from each of their studentÂ’s total cognitive style score. Specifi c constructs of cognitive style gap scores were calculated in the same manner. Therefore, cognitive style gap described students in terms of being more adaptive or more innova tive in relation to their faculty member. Utilizing SPSS statistical software, the research er used PearsonÂ’s correlation coefficient to determine the strength and direction of the relationship between cognitive style gap with student stress, student motivation, st udent engagement and student demographic variables identified in objective one. Research Objective Four Research objective four was to expl ain both undergraduate student stress and student motivation based on cognitive styl e gap and student demographic variables identified in objective one. Using SPSS statis tical software, backward stepwise multiple

PAGE 99

81 linear regression was used twice for each cl ass to achieve objec tive four. Data were analyzed first to determine relationships between constructs of cognitive style gap (independent variable), student demographics (independent variable ) and student stress (dependent variable); and second, to determ ine relationships of the same independent variables with student motivation (dependent variable). For further analysis, tests of significance were used to compare groups. Research Objective Five Research objective five was to explai n undergraduate student engagement based on cognitive style gap, student stress, st udent motivation and selected student demographic variables identified in objectiv e one. Employing SPSS statistical software, backward stepwise multiple linear regression wa s used to explain the dependent variable student engagement based on i ndependent variables including cognitive style gap, student stress, student motivation and student demogr aphic variables identified in objective one. For further analysis, tests of signifi cance were used to compare groups. Summary This chapter provided details of the me thodology of this study with regards to research design, population, instrumentation, da ta collection, and data analysis. The study utilized an ex post facto Design 1 methodology to study th e relationships of KirtonÂ’s cognitive gap with stress, mo tivation, engagement and se lected student demographic variables. The population consis ted of undergraduate students enrolled in the College of Agricultural and Life Sciences at the University of Flor ida. To ensure variance of studentsÂ’ cognitive style gap, both in size and direction, faculty members were selected based on their adaptiveness and innovativeness of cognitive style. From this group of

PAGE 100

82 faculty members, nine courses were purposefu lly selected to meet the objectives of this study. Four instruments were utilized to collect data from the participants. They were the following: Kirton Adaption-I nnovation Inventory (KAI) (Kir ton, 1976), National Survey of Student Engagement 2005 (NSSE) w ith demographic questions (Kuh, 2001), Motivated Strategies for Learning Questionn aire (MSLQ) (Pintrich, Smith, Garcia, and McKeachie, 1991), and Student-life Stress Inventory (SSI) (Gadzella, 1994). All instruments were assessed for validity and reliability prior to conducting the study. Posthoc reliabilities for this study were presented in chapter 4. In chapter 4, data were analyzed per cl ass and organized by objectives. This study primarily used descriptive statistics, corre lation statistics, and backward stepwise multiple linear regression.

PAGE 101

83 CHAPTER 4 RESULTS Chapter 1 provided the introduction of KirtonÂ’s Adaption-Innovation (A-I) theory as it applies to learning styles of undergraduate students in the classroom. Attention was also given to KirtonÂ’s separation of cognitive level from cognitive style. The chapter also presented a need for facilitating student e ngagement in the unde rgraduate classroom. A dissimilar learning style (cognitive gap) betw een the university faculty member and the student may give insight to why students ar e not engaged. The purpose of this study was to determine if significant rela tionships exist between cognitive style gap, student stress, student motivation, student engagement and se lected student demographic variables of undergraduate students at the University of Fl orida in the College of Agriculture and Life Sciences. The specific objectives of this study were to: 1) describe selected faculty and students according to their sele cted demographic variables; 2) determine the cognitive style, student stress, student motivation, and student engagement of undergraduate students; 3) determine the cogni tive style gap between facult y and students and to explore its relationship with student stress, student motivation, student engagement and specific demographic variables of undergraduate student s; 4) explain undergraduate student stress and student motivation based on cognitive style gap and selected demographic variables; and 5) explain undergraduate student engage ment based on cognitive style gap, student stress, student motivation and selected dem ographic variables. Furthermore, chapter 1 listed definitions, basic assumptions, and cited limitations of the study.

PAGE 102

84 Chapter 2 discussed the theoretical and c onceptual framework of the study as well as reviewing the literature concerning cogniti ve style, stress, motiv ation, and engagement as they relate to learning. Special attenti on was given to KirtonÂ’s cognitive style as a measure of problem solving learning. Furt hermore, dissimilar cognitive style, as measured by cognitive style gap, may cause faculty members and students to have difficulties communicating and working together in the classroom environment. The reliable and valid measure of cognitive styl e gap may explain why some students are not engaged in the course. Chapter 3 provided the research methodology of this study with attention given to research design, population and instrumenta tion. Procedures were outlined for data collection and data analysis to comple te the five objectives of this study. This chapter presents the results of this study which are organized by the objectives listed previously. Instrument Post-hoc Reliability CronbachÂ’s alpha was used to determine post-hoc reliability of the four instruments administered to participating stude nts. Considering all students in this study, the total cognitive style score measured with KirtonÂ’s Adaption-Innovation Inventory (KAI) had an alpha coefficient of .85, which comprised of the constructs: sufficiency of originality ( =.80), efficiency ( =.66) and rules/group conformity ( =.76). Regarding the measurement of student stress, the Studentlife Stress Inventory (SSI) had a total score alpha coefficient of .88 which included five constructs: frustrations ( =.80), conflicts ( =.84), pressures ( =.73), changes ( =.89) and self-imposed ( =.69). Motivation in this study was measured with the Motivated Stra tegies for Learning Questionnaire (MSLQ) which had a post-hoc total score alpha coeffi cient of .86 which included the constructs:

PAGE 103

85 intrinsic ( =.70), extrinsic ( =.62), task value ( =.91), control of learning belief ( =.70), self-efficacy for learning ( =.91) and test anxiety ( =.83). The National Survey of Student Engagement (NSSE) was used to meas ure student engagement in the classroom and the total score alpha coefficient for this measure was .79. Three scales were used to determine student engagement: academic challenge ( =.70), active and collaborative learning ( =.62) and student-faculty interaction ( =.62). Reliability coefficients are considered acceptable at a level above .70 (Schmitt, 1996). Although some construct scales pr esented previously had post-hoc alpha coefficients below .70, it should be noted that only total scores for stress, motivation and engagement were used in data analysis to meet the objectives of this study. However, the cognitive style construct effi ciency was used despite its low post-hoc reliability. The efficiency construct is a m easure of preferred method to solve problems (Kirton, 2003) and was expected to have relationships with stress and motivation (Barnhart, 2002). Furthermore, Kirton (1999) provided abundant evidence that all three constructs of cognitive style are reliably measured by the KAI. Objective One Describe selected faculty and students accord ing to their selected demographic variables. This study comprised nine faculty members each of whom taught a course of approximately 100 undergraduate students. Faculty members were asked questions regarding age, gender, years of teaching e xperience, and academic department. Faculty members also provided the researcher with sy llabi and additional information to describe the course. Undergraduate students were asked demographic questions relevant to student engagement found on the National Survey fo r Student Engagement (NSSE) (Kuh, 2001)

PAGE 104

86 which included: age, gender, major, college classification, full-time status, and number of classes taken similar to the subject area of th e course. Lastly, student s were asked to give the number of problem sets that had taken more than an hour to complete during a typical week. The last question was not us ed in data analysis, but serv ed to provide evidence that problem solving assignments did exist in the courses examined in th is study and will be considered in the concluding discussion in chapter 5. The data are presented with each faculty member’s class separated as an intact group from the total group of respondents. E ach class was coded by letter according to the faculty member’s cognitive style score along the continuum of adaptiveness to innovativeness. Therefore, the most adaptive faculty member was assigned the letter “A” and the letter “I” was assigned to the most innovative faculty member. See Table 4-1 for a description of faculty members’ cognitive style scores determined by Kirton’s Adaption-Innovation Inventory (KAI). Table 4-1. Listing of Classes designated by Faculty Members’ Total Cognitive Style Score (N=9) Total Cognitive Style Sufficiency of Originality Efficiency Rule/Group Conformity Class A Class B Class C Class D Class E Class F Class G Class H Class I All Faculty 64 68 83 94 95 103 116 132 134 98.78 29 25 40 48 41 51 49 53 60 44 10 19 12 15 18 9 21 29 20 17 25 24 31 31 36 43 46 50 54 37.78 Kirton (2003) states that total cognitive style scor es less than 10 points are considered the same. Cognitive style scores between 10 and 19 points are considered

PAGE 105

87 similar and scores that are more than 20 poi nts apart are considered different. Therefore from this group of faculty members an adap tive teacher group was formed from Classes A, B and C, a middle score teacher group was formed from Classes D, E and F and an innovative teacher group was formed out of Cl asses G, H and I. This allowed for further analysis of student variables based on similar cognitive styles of faculty members. Class A This three hour course concentrated on improving writing skills for academic settings and business applicati ons. The course was practical in nature requiring students to prepare a resume and application packet, bo th of which were utilized in a mock jobinterview. A research report was also de veloped as the faculty member led students through the process of conducti ng research. The course is not required for undergraduate students, but is suggested as a general writi ng course in the College of Agriculture and Life Sciences. The instructor of Class A was a 25 year-old female graduate assistant. She had three years of teaching experience at the college level, but no prior teaching experience in any other setting. The instructor chose to have instruments administered in class, completed outside of class, and then returned at the next normally scheduled class meeting. The instructor gave extra-credit points to students who returned instruments. There were 100 students enrolled in Class A, of which 70 responded. All participating students were considered full-ti me (taking more than 12 credit hours). Of the respondents, 44.3% (n=31) were male a nd 55.7% (n=39) were female. For academic majors, 28 different majors represented the cl ass. No specific major held a majority, but the three most predominant majors were family youth and consumer science (14.3%, n=10), nutritional science (11.4%, n=8), a nd food science and human nutrition (10%, n=7).

PAGE 106

88 For Class A, the student mode age wa s 20 years old (32.9%, n=23). There were 20 students (28.6%) who reported their age as 21years of age and 10 students (14.3%) who were 19 years of age. Almost half of Class A were juniors (48.6%, n=34), while 30% (n=21) were seniors and 18.6% (n=13) were sophom ores. There were only two (2.9%) freshman enrolled in the class that participated in this study. Students were asked how many courses they had taken in the s ubject area similar to Class A. Respondents repor ted that 17.4% (n=12) had taken no courses, 68.1% (n=47) had taken one to two courses, 13.0% (n=9) had taken three to four courses and 1.4% (n=1) had taken more than four courses in the subject area. There was one student that did not answer this specific question. Regarding the number of problem sets taking more than an hour to complete in a typical week 52.9% (n=37) stude nt participants in Class A claimed that none of the assigned problem sets had taken more than an hour. However, 38.6% (n=27) claimed one or two problem sets and 8.6% (n=6) of the resp ondents claimed three to four problem sets assigned in class had taken more than an hour to complete. Class B Class B was a three hour course focused on the principles of quantitative genetics as applied to domesticated animals. Partic ular attention was given to farm animal industries including beef, da iry, equine and poultry. Although the final grade was calculated from exams and quizzes, mathematical problem sets were used throughout this course in the form of homewo rk assignments. The course wa s required for students in the animal science department. The faculty memb er providing instruction for Class B was a male at 54 years of age. He had taught at the University of Florida for 29 years and had no other teaching experiences. Instruments were administered in class, but completed

PAGE 107

89 outside of class and then re turned at the next normally scheduled class meeting. The instructor gave extra-credit points to students who returned instruments. There were 100 students enrolled in Cl ass B, of which 72 responded. Of the respondents, 95.7% (n=66) were full-time students while 4.3% (n=3) were part-time students. Three respondents did not complete the demographic questions. The majority of the respondents in Class B were female (81.2%, n=56). Male st udents comprised 18.8% of the respondents (n=13). Class B students we re members of five different majors, of which 97.1% (n=67) were within the animal science department. In Class B, the mode age was 21 years old (37.7%, n=26). There were 14 (20.3%) students that were 22 years of age and 9 ( 13.0%) students that were 20 years of age. There were only 3 (4.3%) stud ents who were 19 years old. The class was only composed of juniors (50.7%, n=35) and seniors (49.3% n=34). Again, three respondents did not answer the demographic questions. Students were asked how many courses were taken in the subject area similar to Class B. Respondents reported 23.2% (n=16) had taken no courses, 37.7% (n=26) had taken one to two courses, 13.0% (n=9) had ta ken three to four courses and 26.1% (n=18) had taken more than four courses in the subj ect area. There were three students that did not answer this specific question. For problem sets assigned in class that ha d taken students longer than an hour to complete in a typical week, 51.4% (n=36) claimed none, 44.3% (n=31) said one to two, 1.4% (n=1) said three to four and 2.9% (n=2) declared more than five problem sets. Two students did not answer the question describing number of problem sets taking longer than an hour.

PAGE 108

90 Class C This course consisted of three cr edit hours focusing on the study of food marketing. Four case studies were used to teach concepts of marketing, product development, distribution and pricing. Students were asked to identify issues, make decisions, and give recommendations based upo n the information presented in the case study. The course was required for students majoring in food and resource economics. The faculty member instructing Class C was a 36 year-old female. She had taught 12 years at the college level, but had no pr ior teaching experience. Instruments were administered and completed during class time. The faculty member provided extra-credit points for students that participated. There were two sections of Class C, each taught by the faculty member with the same syllabus. Each section included 45 st udents. In the first section, 33 students participated, while 38 students participated in the second section. The two groups were combined to form Class C. Of the respondent s, 98.6% (n=70) were full-time students. One student (1.4%) reported being a part-time student. The majority of the students in Class C were male (71.8%, n=51) with fema les making up 28.2% (n=20) of the group. Respondents in Class C were working toward s five different majors of which 93% (n=66) were food and resource economics, a major within the department. The remainder of the responding students majored in agri culture (2.8%, n=2), business administration (1.4%, n=1) and community recreation (1.4%, n=1). The mode age of Class C was 22 year s old (35.2%, n=25). There were 13 respondents (18.3%) reporting an age of 21 years and anot her 13 respondents (18.3%) reporting an age of 23 years. Class C comprise d mostly of seniors (91.5%, n=65) with the remainder of the class be ing juniors (8.5%, n=6).

PAGE 109

91 Students were asked how many courses were taken in the subject area similar to Class C. Of the respondents, 2.8% (n=2) ha d taken no courses, 23.9% (n=17) had taken one to two courses, 26.8% (n=19) had taken th ree to four courses and 46.5% (n=33) had taken more than four courses in the subject area. Participating students also answered one item asking how many problem sets assigned in a typical week ha d taken more than an hour to complete. Of the student respondents in Class C, 19.7% (n=14) repor ted none, 63.4% (n=45) stated one to two problem sets, 14.1% (n=10) asse rted three to four problem sets, and 2.8% (n=2) declared five or more problem sets assigned in a t ypical week had taken longer than an hour to complete. Class D Three credit hours were assigned to Class D, in which students learned about embryonic, prenatal and postnatal growth a nd development of domesticated animals. Even though the content learned was primar ily lecture based, students were asked to analyze, synthesize and apply f our journal articles of the studentÂ’s choice. The purpose of the assignments were to give students the experience of finding speci fic journal articles and be able to apply the information to topi cs discussed in class and to their life. The course was required for students in the animal science department. The faculty member for Class D was a 62 y ear-old male with 38 years of teaching experience at the college level. He had no prior teaching experience before becoming a university faculty member. Instruments were administered and completed during class time. The faculty member provided extra-cr edit points for students that participated. Class D enrolled 116 students of which 108 participated in this study. In this course, 97.2% (n=103) were considered fulltime students and 2.8% (n=3) were part-time

PAGE 110

92 students. Two students did not answer th e question regarding full-time status. The majority (81.3%, n=87) were female while males consisted of 18.7% (n=20). One student did not complete the ques tion specifying gender. Students in Class D made up five diffe rent academic majors of which three majors offered within the department cons tituted 97.2% (n=104) of the group. The most predominant major in the course was animal science (67.3%, n=72). The academic major animal biology had the second highest number of students at 23.4% (n=25). The remaining academic majors were pre-vete rinarian, wildlife conservation and biology which together accounted for 9% (n= 10) of students in the class. The mode age of Class D was 21 years old (36.4%, n=39). Students who were 20 years old numbered 28 (26.2%) with students reporting 19 years of age constituting 10.3% (n=11) of the class. On e student reported being 17 ye ars of age. One student did not answer the question specifying age. The co llege classification of students in Class D comprised 60.4% (n=64) juniors, 32.1% (n=34) seniors, 6.6% (n=7) sophomores, and one freshman. Two students did not answer the question regarding college classification. Students were asked how many courses were taken in the subject area similar to Class D. Of the respondents, 5.7% (n=6) ha d taken no courses, 32.1% (n=34) had taken one to two courses, 28.3% (n=30) had taken th ree to four courses and 34.0% (n=36) had taken more than four courses in the subject area. There were two students that did not answer this specific question. Concerning the number of problem sets that require more than hour to complete during a typical week, 83.3% (n =90) of student respondents asserted none, 15.7% (n=17) stated one to two problem sets and 0.9% (n=1) claimed three to four problem sets.

PAGE 111

93 Class E Class E was a two-hour, stand-alone laboratory class in which students participated in laboratory exercises. Content of the course entailed st ructure, nutrition and growth of microorganisms. The students were also required to write reports discussing the findings of laboratory exerci ses. The course was mandatory for all students wishing to take advanced courses in the Depart ment of Microbiology and Cell Science. The faculty member of Class E was a 54 year-old female. She had 23 years of teaching experience at the college level; howev er she had no prior teaching experience at any other level. The faculty member insisted that instruments were to be administered during class, but completed outside of class time. Students were aske d to bring completed instruments to the next normally scheduled class meeting time. The faculty member did not provide extra-credit for stude nts participating in the study. There were three sections of Class E, each taught by the faculty member with one syllabus. Each section consisted of 35 stude nts of which 16 responded in each section. The three sections were combined to form Cl ass E with a total of 48 student respondents. Among respondents, 97.9% (n=47) were full-time students with one student reporting as part-time. Female students held the major ity (77.1%, n=37) with males making up 22.9% (n=11) of the class. There were 18 academic majors reported in Class E, of which only four constituted more than 10% of the group. Thos e academic majors were nutritional science (18.8%, n=9), microbiology (14.6%, n=7), an imal biology (12.5%, n=6) and biology (12.5%, n=6). Student age in Class E was bimodal with 14 students (29.2%) being 21 years of age and 22 years of age. Three students (6.3%) were 19 years old and two students

PAGE 112

94 (4.2%) were over 30 years old. Class E was made up of mostly seniors (64.6%, n=31) and juniors (33.3%, n=16). There was one sophomore enrolled in Class E. Students were asked how many courses were taken in the subject area similar to Class E. Of the respondents, 8.3% (n=4) ha d taken no courses, 54.2% (n=26) had taken one to two courses, 14.6% (n=7) had taken th ree to four courses and 22.9% (n=11) had taken more than four cour ses in the subject area. In this class, students were asked how many problem sets assigned during a typical week that require more than an hour to complete. In Class E, 91.7% (n=44) stated no problem sets were completed that met this criteria while 6.3% (n=3) declared one or two problem sets and 2.1% (n=1) claimed three to four problem sets assigned had taken longer than an hour to complete. Class F Two credit hours were assigned to Class F in the subject area of human nutrition. Concepts learned included nutritional requirements during different stages of the human life cycle as well as socio economic, cultural and psychological aspects of behaviors toward food. Critical thinking teaching met hods were utilized throughout assignments, exams, and class discussions. Assignments resembled real-life situations in which students were asked to solve problems. Furthe rmore, students enrolle d in the class were made aware of the research process to be tter understand the lite rature pertaining to nutrition. The course was required for nutriti onal science academic majors and dietetics majors. The faculty member who taught Class F wa s a 46 year-old female who had taught 11 years at the University of Florida. Previous to that, she had worked as a dietician for 7 years in which she occasionally educated pa tients in group settings. The faculty member

PAGE 113

95 allowed administration of instruments during class time, but students were not allowed class time to complete the instruments and were asked to return the completed instruments at the next normally scheduled class meeting time. Students were given extra-credit for their participation. There were 150 students enrolled in Cl ass F of which 115 participated. All respondents of this course were full-time stude nts. The majority of the course was female (69.6%, n=80) with 30.4% of the class being male (n=35). Class F was composed of 12 academic majors of which 81.8% were held by two majors, nutritional science (55.7%, n=64) a nd food science and hu man nutrition (26.1%, n=30). Eight students (7%) were worki ng towards majors in dietetics. Class F students had a mode age of 21 years (43.5%, n=50). There were 30.4% (n=35) students reporting as 20 years old a nd 8.7% (n=10) students reporting as 22 years old. Of the respondents, juniors made up th e majority (58.3%, n=67) of Class F with seniors constituting 33.0% (n=38). Also enrolle d in Class F were six (5.2%) sophomores and four (3.5%) freshmen. Students were asked how many courses were taken in the subject area similar to Class F. Of the respondents, 1.7% (n=2) ha d taken no courses, 29.6% (n=34) had taken one to two courses, 44.3% (n=51) had taken th ree to four courses and 24.3% (n=28) had taken more than four courses in the subject area. Student participants were asked how many problem sets had taken more than an hour to complete during a typical week. In Class F, 81.7% (n=94) asserted no problem sets, 16.5% (n=19) claimed one to two problem sets and 1.7% (n=2) claimed three to four problem sets given during a t ypical week had taken longer than an hour to complete.

PAGE 114

96 Class G A three credit hour course, Class G cont ent revolved around principles of youth development. Priority was given to physical, cognitive, social and emotional changes between childhood and adulthood. Students were required to work in teams for the purpose of participating in a service learning project. This project allowed students to pick a childhood or teen group, identify curriculum needs and teach the curriculum to the group. The course was required for majors and minors in family youth and community science. The faculty member who taught Class G wa s a 35 year-old female. She had taught five years in her current position at the Univer sity of Florida. This faculty member also reported tutoring student athletes for eight ye ars and two years of educational experience as a cooperative extens ion agent. Instruments were administered and completed by students during class time. Stude nts were given extra-credit fo r participating in this study. Class G enrollment consisted of 110 student s of which 85 agreed to participate. Respondents in this course were 98.8% (n= 84) full-time students. One (1.2%) student was identified as being part-time. Howeve r, 11 respondents did not answer the question specific to full-time status. Most of the respondents were female (81.2%, n=69) with males making up 18.8% (n=16) of the class. Still, 11 respondents did not answer the question regarding gender. There were 13 reported academic majors in Class G. Of these majors, 75.3% (n=64) were family youth and consumer science and 10.6% (n=9) were psychology. None of the other academic majors totaled more than 3% (n=2) respondents. Again, 11 students did not answer the quest ion regarding academic major.

PAGE 115

97 Class G students reported a mode ag e of 21 years (34.1%, n=29) with 31.8% (n=27) reporting as 20 years old. Also, 11.8% (n=10) of respondents were 22 years of age. Class G respondents were 57.6% (n=49) juniors, 30.6% (n=26) seniors and 11.8% (n=10) sophomores. Still, 11 participating st udents did not answer the question regarding college classification. Students were asked how many courses were taken in the subject area similar to Class G. Respondents declared that 2.4% (n=2 ) had taken no courses, 23.8% (n=20) had taken one to two courses, 41.7% (n=35) ha d taken three to four courses and 32.1% (n=27) had taken more than four courses in the subject area. There were 12 students that did not answer this specific question. For the number of problem sets assigned during a typical week that had taken longer than an hour to complete, 50.0% (n= 42) claimed none, 45.2% (n=38) said one to two problem sets and 4.8% (n=4) said thr ee to four problem sets. There were 12 participating students who did not respond to the questions distinguishing number of problem sets completed during a typical week. Class H As a three credit hour course, the instructor of Class H taught topics relevant to agricultural and food ma rketing systems. Specifically, th e topics included identification of food marketing participan ts, functions, interdependenc ies and trends, as well as evaluating components of the system. Student final grades were based on four exams and two projects. Furthermore, effort was made by the faculty member to apply learned content to actual settings requiring a solution to a problem. Written a nd oral presentations were used to showcase student learning and gi ve opportunity to apply concepts to real-

PAGE 116

98 life situation. The course was required for students majoring in food and resource economics. The faculty member who taught Class H was a 42 year-old male. He had five years of collegiate teaching experience at the Un iversity of Florida. This faculty member also reported teaching experiences during his two years of serving in the Peace Corps and two years at an agricultura l technical college. Instrume nts were administered and completed by students during class time. Students were given extra-credit for participating in this study. Class H consisted of 122 students of which 70 responded. Of the participating students 98.5% (n=67) were enrolled as full-time students. One (1.5%) student was determined as a part-time student while two (2.9%) students did not answer the question specifying enrollment status. Males held a slim majority of th e respondents at 55.1% (n=38) with female respondents of Class H accounting for 44.9% (n=31). One student did not respond to the question regarding gender. Nine academic majors were reported in Cl ass H. The majority of the respondents (81.2%, n=56) majored in food and resource eco nomics. Four (5.8%) students declared a major in animal science. The academic majors of agricultural communication and mechanical engineering each had two responde nts (2.9%). The remaining five academic majors consisted of only one student each. Th ere was one student who did not answer the question on the subject of academic major. For ages of respondents in Class H the mode age was 21 years old (31.9%, n=22). There was an equal number of participants reporting 20 and 22 years of age (18.8%, n=13). Ten students (14.5%) reported the age of 23 years. One student did not answer the

PAGE 117

99 question specifying age. Class H respondents were primarily juniors (53.6%, n=37) and seniors (36.2%, n=25). Sophomores made up 7.2% (n=5) of the class and freshman accounted for 2.9% (n=2). One student did no t respond to the questi on identifying college classification. Students were asked how many courses were taken in the subject area similar to Class H. Respondents claimed that 15.9% (n= 11) had taken no courses, 34.8% (n=24) had taken one to two courses, 27.5% (n=19) had taken three to four courses, and 21.7% (n=15) had taken more than four courses in the subject area. There was one student that did not answer this specific question. Student participants were asked how ma ny problem sets were completed during a typical week that had taken more than an hour to complete. In Class H, 55.7% (n=39) claimed none, 40% (n=28) declared one to tw o problem sets, 2.9% (n=2) stated three to four problem sets, and 1.4% (n=1 ), asserted more than five problem sets had taken more than an hour to complete during a typical week. Class I This course was identical to Class A in content, assignments and exams. However, the syllabus, content and classroom policies were determined by the instructor. There was no team teaching between Class A and Class I; rather these were two completely separate sections of the same course with standardized assignments and exams. Class I was a three credit hour cour se focused on developing studentsÂ’ writing skills. Students were led th rough the process of applyi ng for a job which included creating a resume, completing j ob applications and mock in terviews. Students were led through the process of conducting research and worked with th e instructor to complete a

PAGE 118

100 research report. The course was not re quired for undergraduate students, but was suggested to undergraduate students in the Co llege of Agriculture and Life Sciences. The instructor who taught Class I was 26 y ear-old female and a graduate assistant with one year of collegiate teaching experi ence. She also had less than 6 months of leading non-formal educational programs at a community level. The instructor permitted administration of instruments in class, but students completed instruments outside of class time. Students were given extra-credit from the faculty member for returning completed instruments. There were 100 students enroll ed in Class I with 77 part icipating in this study. All students who responded (n=76) claimed to be full-time, however, one student did not answer the question concerning full-time status. Females constituted 62.3% (n=48) of the student respondents with ma les comprising 37.7% (n=29). Thirty different academic majors were represented among the respondents. The most predominate academic majors were nut ritional science (14.3%, n=11) and food and resource economics (14.3%, n=11). The academ ic major of family youth and consumer sciences constituted 11.7% (n=9) of the re sponding students. None of the other 27 academic majors accounted for more than 10% of the respondents. Concerning the ages of Class I, resp ondents were bimodal with 34.2% (n=26) each at 20 and 21 years of age. Eleven (14.5 %) students reported 22 years of age. One student did not report age. Cl ass I respondents determined th eir college classification as 48.1% (n=37) juniors, 26.0% (n=20) senior s, 20.8% (n=16) sophomores and 5.2% (n=4) freshman.

PAGE 119

101 Students were asked how many courses were taken in the subject area similar to Class I. In response, it was found that 34.2% (n=26) had taken no courses, 46.1% (n=35) had taken one to two courses, 13.2% (n=10) had taken three to four courses and 6.6% (n=5) had taken more than four courses in the subject area. Student participants were asked how ma ny problem sets were assigned during a typical week that had taken more than an hour to complete. In Class I, 47.4% (n=36) stated none, 43.4% (n=33) claimed one to two problem sets, 6.6% (n=5) declared three to four problem sets and 2.6% (n=2) stated mo re than five problem sets were completed during a typical week which re quired more than one hour. All Students The nine classes examined in this stu dy were combined to present demographic information of all students participating in this study. There were 993 undergraduate students enrolled in the nine courses ex amined in this study, of which 716 (72.1%) participated. Of the respondents, 98.6% (n=698) were classified as full-time students. Ten students (1.4%) considered themselves as pa rt-time students. Ther e were eight students who did not respond to the item regarding full -time status. Of the student respondents, 65.7% (n=467) were female while 34.3% (n=2 44) were male. Five students did not answer the question regarding gender. A total of 63 different academic ma jors were identified among the student participants. The five academic majors with the largest number of pa rticipating students were: food and resource economics (19.1%, n=136), animal science (17.5%, n=125), nutritional science (13.0%, n= 93), family youth and consumer sciences (11.6%, n=83) and animal biology (8.8%, n=63). Together, these five academic majors comprised 70.1%

PAGE 120

102 of the student respondents in this study. There were three students who did not answer the question specifying the studentÂ’s academic major. The mode age of student participants was 21 years (33.6%, n=239). There were 173 respondents (24.3%) with a reported age of 20 year s and 115 respondents (16.2%) reporting an age of 22 years. There were f our students who did not answer the item regarding student age. The majority of th e students were juniors (48.6%, n=345) and seniors (41.4%, n=294). Howeve r sophomores accounted for 8.2% (n=58) of the class and freshman accounted for 1.8% (n=13). Ther e were six students who did not report their college classification. Considering the number of courses student s had taken in the same subject area to the course of which instruments were admi nistered, 11.5% (n=81) had taken no courses, 37.2% (n=263) had taken one to two courses, 26.7% (n=189) had ta ken three to four classes and 24.6% (n=174) had taken more than four courses. There were nine students who did not answer the item regardi ng number of similar courses taken. The number of problem sets assigned duri ng a typical week for the entire group of student respondents that had ta ken more than one hour to co mplete was a question asked to help validate that prob lem solving based assignments were being used in the classroom. It was found that 60.7% (n=432) of students had no assignments as such, but 33.8% (n=241) stated one to two problem sets during a typical week had taken more than an hour to complete. There were 4.5% (n= 32) claiming three to four problem sets assigned during a typical week had taken more than an hour to complete and 1.0% (n=7) stated more than five such assignments ex isted. There were four students who did not answer the question regarding number of probl em sets completed dur ing a typical week.

PAGE 121

103 Summary of Findings for Objective One Faculty members were coded by their c ognitive style score with the letter “A” designating the most adaptive faculty member and the letter “I” designating the most innovative faculty member. This group of faculty members ranged from a graduate student employed as a lecturer to tenured faculty members with ages ranging from 25 to 62. Six of the nine faculty members were female. The group’s range of teaching experience was 1 to 38 years with few facu lty members having teaching experience outside of the classroom. Faculty members re presented six academic departments in the College of Agriculture and Life Sciences at the University of Florida. Faculty demographic information was not used in any further data analysis, but presented here simply to describe the faculty member group. See Table 4-2 for a summary of demographic information regarding participating faculty members. Table 4-2. Summary of Classes Regarding Reported Demographic Information of Participating Faculty Members Course Instructor Age Gender Experience Teaching (years) Academic Department Faculty member A Faculty member B Faculty member C Faculty member D Faculty member E Faculty member F Faculty member G Faculty member H Faculty member I 25 54 36 62 54 46 35 42 26 Female Male Female Male Female Female Female Male Female 3 29 12 38 23 11 7 7 1 Agricultural Education & Communication Animal Science Food & Resource Economics Animal Science Microbiology & Cell Science Food Science & Human Nutrition Family Youth & Consumer Science Food & Resource Economics Agricultural Education & Communication

PAGE 122

104 Students were asked demographic found on the NSSE (Kuh, 2001) which included: age, gender, major, college classifi cation, full-time status and number of classes taken similar to the subject area of the course. Students were also asked how many problems sets that had taken more than an hour to complete during a typical week. This question provided some evidence that problem based assignments were used in the class, albeit less than anticipated by the researcher. This will be further discussed in chapter 5. As most student participants were full-tim e students (taking more than 12 credithours) and were between the ages of 20 and 23, there is evidence that these students were mostly traditional undergraduate students. Seven of the nine classes were predominately female. The data suggests that the majority of students in the classes of study were either juniors or seniors in college classification indicati ng that courses were of higher level. Regarding the number of courses the student ha d taken with similar content varied across classes, however from partic ipantsÂ’ responses, 50.7% (n=363) of the students had taken at least three classes that were similar to th e course used in the study. See Table 4-3 for a summary of student demographic information. Table 4-3. Summary of Classes Regarding Reported Demographic Information of Participating Students Gender (No. of students) College Classification (No. of students) No. Similar Courses (No. of students) Course Mode Age (years) F M Fr. So. Jr. Sr. 0 1-2 3-4 4+ Class A Class B Class C Class D Class E Class F Class G Class H Class I All students 20 21 22 21 22 21 21 21 21 21 39 56 20 87 37 80 69 31 48 467 31 13 51 20 11 35 16 38 29 244 2 0 0 1 0 4 0 2 4 13 13 0 0 7 1 6 10 5 16 58 34 35 6 64 16 67 49 37 37 345 21 34 65 34 31 38 26 25 20 294 12 16 2 6 4 2 2 11 26 81 47 26 17 34 26 34 20 24 35 263 9 9 19 30 7 51 35 19 10 189 1 18 33 36 11 28 27 15 5 174

PAGE 123

105 Objective Two Determine the cognitive style, student stres s, student motivation and student engagement of undergraduate students. Student participants in e ach class were asked to res pond to four questionnaires. They included the following: Kirton Adap tion-Innovation Inventory (KAI) to measure cognitive style with constructs of sufficiency of originality, effici ency, and rule/group conformity; the Student-life St ress Inventory (SSI) to measur e classroom specific stress characterized by frustrations, conflicts, pressures, changes and self-imposed stress; the Motivated Strategies for Learning Questionnai re (MSLQ) to measure classroom specific motivation with components of intrinsic mo tivation, extrinsic motivation, task value, control of learning beliefs, se lf-efficacy for learning and te st anxiety; and the National Survey of Student Engagement (NSSE) to measure classroom specific engagement with scales of academic challenge, active and collaborative learning, and student faculty interaction. Chapter 3 discusse d how these instruments were modified to meet the needs of this study. The data are pr esented here with re sults of all four questionnaires grouped by the classes described above. Also, cognitive style, stress, motivation and engagement were determined through examining all participating students as a group. Class A There were 58 usable responses in Cl ass A (N=100, n=70) pertaining to the KAI. Cognitive style was coded with lower numbers signifying more adaptive and higher numbers signifying more innovative. The KAI has a range of 32 to 160 with a mean of 95. The total cognitive style mean for Class A student respondents was slightly more adaptive (M=90.60, SD=18.01, n=58) than the normali zed mean reported by Kirton

PAGE 124

106 (1999). Constructs of cognitive style were likewise slightly more adaptive than the general population (Kir ton): sufficiency of originality was 1.38 points more adaptive, efficiency was 2.10 points more adaptive and rule/group conformity was 0.91 points more adaptive. However, constructs scores we re consistent with the total mean of 90.60 (Kirton). The most adaptive student in Class A scored a 51 on the KAI and the most innovative student scored a 125. See Table 4-4 for findings specific to student cognitive style. Table 4-4. Class A Student Mean Scores of Cognitive Style Constructs (n=58) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 90.60 39.62 16.90 34.09 18.01 8.40 4.72 8.45 51 22 8 17 125 58 28 51 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. There were 68 usable scores for dete rmining Class A respondentsÂ’ perceived classroom total stress. The SSI measured pe rceived stress with 22 items which were coded one for low level of perceived stress and five indicating a high perceived stress level. The total stress mean of student respondents for Class A was approximately 18 points lower (M=52.44, SD=10.57, n=68) than the norm provided by Gadzella and Baloglu (2001), indicating a lower level of st ress. The student with the lowest level of stress scored 29 on the SSI and the highest le vel of perceived stress was 76. The construct mean for student frustrations stress wa s 5.62 points lower than the norm provided by Gadzella and Baloglu, however the researcher notes that one item was removed from this construct prior to data colle ction. The remaining stress cons tructs also indicated lower student stress levels than the instrument nor m: conflicts was 2.46 poi nts lower, pressures

PAGE 125

107 was 2.50 points lower, changes was 3.16 points lower, and self-imposed was 3.65 points lower in measuring student perceived level of stress in Class A. See Table 4-5 for findings regarding classr oom specific stress. Table 4-5. Class A Student Mean Scores of Stress Constructs Construct N Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 68 70 69 69 70 70 52.44 12.03 5.97 11.58 4.87 18.17 10.57 3.73 2.54 3.42 2.04 3.86 29 6 3 5 3 9 76 21 12 19 10 25 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. For total motivation of Class A there were 69 usable scores. The MSLQ was utilized to measure motivation and comp rised 31 items. Construct scores were standardized and ranged from 1 to 7 providi ng for a total score ra nge of 6 to 42. Items were coded with low scores signifying lo w motivation and high sc ores signifying high motivation. The least motivated student of Class A scored 21.03 for total motivation while the most motivated student scored 36.73. Class A respondentsÂ’ total motivation (M=29.31, SD=3.59, n=69) was 1.15 points lower than the norm reported by Pintrich, Smith, Garcia and McKeachie (1991). Three mo tivation constructs between Class A and the instrument norm (Pintrich, et al.) indicated lower levels of motivation: intrinsic motivation was 1.02 points lower, task va lue was 1.01 points lower and control for learning was 0.05 points lower. However three motivation constructs indicated Class A had higher levels of motivation than the in strument norm (Pintrich, et al.): extrinsic motivation was 0.28 points higher, self-efficacy was 0.56 points higher and test anxiety was 0.05 points higher. All scale means of motiv ation were within one standard deviation

PAGE 126

108 identified by Pintrich et al. See Table 4-6 for findings regarding classroom specific motivation for Class A. Table 4-6. Class A Student Mean Sc ores of Motivation Constructs Construct N Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 69 70 69 70 70 69 69 29.31 4.03 5.31 4.53 5.79 6.03 3.68 3.59 1.07 1.06 1.14 0.92 0.86 1.30 21.03 1.75 3.00 1.50 3.25 3.75 1.00 36.73 6.50 7.00 6.83 7.00 7.00 6.60 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. The NSSE was used to determine student engagement utilizing three constructs: academic challenge, active and collaborative learning and student-fa culty interaction. The sum of these three constructs provided 24 items with a range of 24 to 96. A higher engagement scores signifies a more engage d student. Regarding student engagement related to Class A, the summed mean of the three construct scores was 48.43 (SD=7.70, n=70), 15.31 points lower than the reported national mean for college seniors (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Construct scores of student engagement in Class A were also lower th an the national mean for college seniors: academic challenge was 8.39 points lower, active learning was 6.02 points lower and student-faculty inter action was 0.90 points lower. Exam ining the differences between mean scores in Class H and the average college senior mean scores (Kuh et al.) with respect to the standard deviati on, there is evidence that thes e students had lower levels of academic challenge and active learning, but comparatively had an average level of student-faculty interaction. See Table 4-7 for findings regarding Class A student engagement.

PAGE 127

109 Table 4-7. Class A Student Mean Scores of Engagement Constructs (n=70) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 48.43 24.79 11.90 11.74 7.70 3.76 3.09 2.82 32 14 7 6 69 33 20 21 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class B From Class B (N=100, n=72), there were 46 KAI responses considered usable by the researcher. The KAI determines cognitive style with lower scores signifying more adaptive and higher scores signifying more innovative. The KAI has 32 items with a range of 32 to 160 and a reported mean of 95 (Kirton, 1999). Cla ss B studentsÂ’ total cognitive style mean score (M=92.43, SD =15.05, n=46) was 2.57 points more adaptive than the general population norm (Kirton). Constr uct scores of cognitive style were less than one point more adaptive than cons truct norms reported by Kirton, except the construct efficiency which was 1.59 points more adaptive than the general population norm. However, all construct mean scores in Class B were consistent with a total cognitive score of 92.43 (Kirton). In Class B, the most adaptive student had a total cognitive style score of 46, while the most innovative student in Cl ass B scored 116. See Table 4-8 for results concerning stud ent cognitive style for Class B. Table 4-8. Class B Student Mean Scores of Cognitive Style Constructs (n=46) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 92.43 40.39 17.41 34.63 15.05 8.45 4.65 6.68 46 22 7 17 116 59 28 47 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative.

PAGE 128

110 For classroom stress, 68 responses were d eemed usable from Class B participants. The SSI was used to determine students total stress level with 22 Like rt scale items. The range for the SSI was 22 to 110 with higher scor es signifying more stress. The total stress mean of Class B (M=55.85, SD=12.95) was appr oximately 14 points lower than the total mean reported by Gadzella and Baloglu ( 2001). Also, Class B construct mean stress scores were lower than the norm (Gadzella & Baloglu): frustrations was 3.96 points lower, conflicts was 2.40 points lower, pr essures was 3.14 points lower, changes was 2.72 points lower and self-imposed was 1.94 points lower. Note that one item was removed from the frustrations construct prio r to data collection which would provide a slightly lower mean score. The student with th e lowest level of per ceived stress scored 31 for total stress and the student with the highe st amount of perceived stress had a total score of 87. See table 4-9 for result s regarding Class B student stress. Table 4-9. Class B Student Mean Scor es of Stress Constructs (n=68) Construct Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 55.85 13.69 6.03 10.94 5.31 19.88 12.95 4.98 2.84 3.79 2.60 4.23 31 6 3 4 3 8 87 27 12 20 11 29 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. There were 69 usable scores for calculat ing student motivation in Class B. The MSLQ was used to measure student motiv ation and consisted of 31 items. Students responded to items using a 7-poi nt Likert scale with one co ded to signify low motivation and seven coded to signify high motivati on. Construct scores were standardized providing for a total ra nge of 6 to 48 points. Of total student motivation, Class B was 0.19

PAGE 129

111 points higher (M=30.65, SD=12.95, n=68) than the reported sample reported by Pintrich et al. (1991). Construct mean scores were lik ewise similar with less than one point above or below the sample norm (Pintrich et al.) : Specifically, intrinsic motivation was 0.45 points lower, extrinsic motivation was 0.44 points higher, task value was 0.49 points lower, control of learning was 0.01 points higher, self-efficacy wa s 0.08 points lower and test anxiety was 0.78 points highe r. The least motivated student of Class B scored 17.43 for total motivation and the most motivate d student scored 38.58. See Table 4-10 for findings regarding classroom speci fic motivation of Class B. Table 4-10. Class B Student Mean Sc ores of Motivation Constructs Construct N Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 69 70 70 69 70 69 70 30.65 4.60 5.47 5.05 5.75 5.39 4.41 4.37 1.19 1.09 1.27 1.17 1.27 1.68 17.43 1.50 2.25 1.00 2.50 2.25 1.00 38.58 7.00 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. Class B had 70 usable scores for student engagement. The total student engagement score was calculated by adding th e constructs of academic challenge, active learning and student-faculty in teraction. Twenty four items each with a 4-point scale were coded one for low engagement and four for high engagement. The range of the three summed constructs was 24 to 96 with a mean of 63.74 for college seniors. Class B total student engagement mean (M=46.01, SD= 7.47, n=70) was 17.73 points lower than the national mean of college seniors considering the three constructs of study (Kuh et al., 2001). Construct mean scores were also lower than the reported national mean of college seniors (Kuh, et al.): active learning was 6.11 points lower an d student-faculty interaction

PAGE 130

112 was 0.68 points lower. Of particular note, the construct academic challenge was 10.94 points lower, indicating lower levels of student prepara tion and use of higher order thinking skills in Class B. The student with the lowest level of engagement in Class B had a total engagement score of 33 points wh ile the most engaged student in the class scored 63 for total engagement. See Table 411 for findings specific to Class B student engagement. Table 4-11. Class B Student Mean Scores of Engagement Constructs (n=70) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 46.01 22.24 11.81 11.96 7.47 3.63 2.73 3.21 33 16 7 7 63 31 19 20 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class C For Class C (N=90, n=71), there were 56 usable KAI responses considered acceptable by the researcher to determine cogn itive style. Lower scores signify a more adaptive cognitive style and higher scores si gnify a more innovative cognitive style with a range of 32 to 160 and mean of 95. The total cognitive style mean for Class C was 5.86 points higher (M=100.86, SD=14.34) than th e general population mean determined by Kirton (1999) which indicated a slightly innovative group. Cognitive style construct mean scores were likewise higher: suffici ency of originality was 3.59 points higher, efficiency was 0.61 points higher and rule /group conformity was 1.66 points higher indicating slightly more innovative students. However, all cognitive style construct mean scores in Class C was consistent with the to tal mean score (Kirton). The most adaptive

PAGE 131

113 student in Class C scored 65 on the KAI, wh ile the most innovative student scored 142. See Table 4-12 for findings concerning student cognitive style of Class C. Table 4-12. Class C Student Mean Scores of Cognitive Style Constructs (n=56) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 100.86 44.59 19.61 36.66 14.34 6.59 4.28 7.50 65 32 9 22 142 62 29 58 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. There were 70 usable responses to calcul ate total perceived st ress in Class C. The SSI was utilized to measure stress with 22 items providing a range of 22 to 110. Low perceived stress scores signify low stress and high perceived stress scores signify high levels of perceived stress. The total stress mean for Cla ss C participants was 15.18 points lower (M=54.91, SD=12.59, n=70) than reported means by Gadzella and Baloglu (2001). Construct mean scores for student stress also indicated lower levels of stress with respect to the norm (Gadzella and Baloglu): frustrations wa s 4.81 points lower, conflicts was 2.36 points lower, pressures was 3.63 points lower, changes was 3.40 points lower and self-imposed was 0.92 points lower. Note that one item was removed from the frustrations construct prior to data collection. Also note that total stress and constructs of stress were approximately one standard de viation lower than the sample reported by Gadzella and Baloglu, but self-imposed stress of Class C was less than one point from the sample norm. The student with the lowest le vel of total stress scored 29 on the SSI and the student with the highest level of total stress scored 83. See Table 4-13 for findings regarding classroom stress of Class C.

PAGE 132

114 Table 4-13. Class C Student Mean Scores of Stress Constructs Construct N Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 70 70 71 71 71 71 54.91 12.84 6.07 10.45 4.63 20.90 12.59 4.68 2.53 3.34 2.42 4.78 29 6 3 4 3 9 83 29 12 20 12 30 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. There were 71 usable responses to determine student motivation of Class C. To measure total motivation, 31 items were us ed from the MSLQ. Respondents answered items on a 7-point Likert scale with one si gnifying low motivation and seven signifying high motivation. Construct scores were standard ized and added to form a total motivation range of 6 to 48. Total motivation in Cl ass C was 0.40 points higher (M=30.86, SD=3.80) than that reported by Pintrich et al. (1 991), indicating higher levels of motivation. However, construct mean scores of motiva tion in Class C were both higher and lower than the test sample norm (Pin trich et al.). Intrinsic motiv ation was 0.26 points lower than the reported norm (Pintrich et al.) and the motivation construct task value was 0.45 points lower. However, extrinsic motivation was 0.37 points higher, control of learning was 0.01 points higher, self-efficacy was 0.45 points higher, and test anxiety was 0.27 points higher than the reported norm (Pintrich et al .), indicating higher levels of student motivation in Class C. The least motivated student of Class C respondents had a total score of 18.73 and the most motivated studen t scored 39.59. See Table 4-14 for findings regarding motivation in Class C. Total student engagement of Class C was determined with the NSSE which in this study comprised three constructs The possible range for total engagement was 24 to 96.

PAGE 133

115 Table 4-14. Class C Student Mean Scores of Motivation Constructs (n=71) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 30.86 4.79 5.40 5.09 5.75 5.92 3.90 3.80 0.91 1.15 1.35 0.90 0.70 1.52 18.73 2.75 1.50 1.00 2.75 3.75 1.00 39.59 6.50 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. The 4-point scale items were coded as one signifying low engagement and four signifying high engagement. The total stude nt engagement mean score for Class C (M=54.73, SD=9.15, n=71) was 9.01 points lower than the national mean for college seniors (Kuh et al., 2001). Construct mean scor es for student engagement in Class C were also lower than the national mean for colleg e seniors (Kuh, et al.): academic challenge was 5.19 points lower and active learning was 3.93 points lower. The exception to this was the construct student-faculty interaction which was 0.12 poi nts higher than the norm (Kuh, et al.). This indicated that students in Class C had an average level of studentfaculty interaction despite low levels of acad emic challenge and active learning. In Class C, the least engaged student had a total en gagement score of 37 and the most engaged student scored 80. See Table 4-15 for findings specific to Class C student engagement. Table 4-15. Class C Student Mean Scores of Engagement Constructs (n=71) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 54.73 27.99 13.99 12.76 9.15 4.46 3.53 3.54 37 16 8 7 80 37 23 21 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement.

PAGE 134

116 Class D For Class D (N=116, n=108), there were 73 usable KAI scores accepted by the researcher. The KAI uses 32 items to determ ine cognitive style and has a theoretical range of 32 to 160 and a mean of 95. Lower cognitive style scores signify adaptiveness while higher cognitive style scores signify innovativeness. The tota l cognitive style mean for Class D student respondents was 93.63 (SD=17.40). This was 1.37 points more adaptive than the standard mean of typi cal populations defined by Kirton (1999). The efficiency construct mean scores of cognitive style was also slightly more adaptive (1.93 points lower) than the genera l population mean (Kirton) in Class D. However, cognitive style constructs of sufficiency of original ity and rule/group conformity were 0.49 points higher and 0.07 points higher respectively. Rega rdless, all construct mean scores were consistent with a cognitive style total m ean score of 93.63 (Kirton). The most adaptive student in Class D had a tota l cognitive style score of 55 and the most innovative student had a cognitive style score of 135. See Table 416 for results regarding student cognitive style in Class D. Table 4-16. Class D Student Mean Scores of Cognitive Style Constructs (n=73) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 93.63 41.49 17.07 35.07 17.40 8.37 5.37 7.98 55 23 7 20 135 55 32 55 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. For total perceived stress in Class D ther e were 101 usable scores recorded by the researcher. The SSI was used to measure to tal perceived stress which included 22 items

PAGE 135

117 and a range of 24 to 96. Using a 5-point Li kert scale students re sponded by marking one to signify low perceived stress and five to signify high perceived stress. In Class D, the total stress mean of the class (M=52.80, SD=13.71, n=101) was 17.29 points lower than that reported by Gadzella and Baloglu (2001), indicating lower levels of perceived stress. Constructs mean scores of student stress in Class D were also lowe r than that reported by Gadzella and Baloglu: frustrations was 5.87 poi nts lower, conflicts was 3.13 points lower, pressures was 3.32 points lower, changes wa s 3.40 points lower and self-imposed was 1.40 points lower. Note that one item was rem oved from the frustra tions construct prior to data collection as it was not applicable to the study. See chapter 3 for an explanation for removing this item. Interestingly, all s cales of stress were one standard deviation lower than the normalized means (Gadzella a nd Baloglu) except for self-imposed stress. This provided evidence that students in Cla ss D had higher levels of self-imposed stress compared to the other stress constructs. The st udent in Class D with the lowest level of total perceived stress scored 26 points on the SSI while the student with the highest level of perceived stress scored 94 points. See Table 4-17 for fi ndings specific to stress of Class D. Table 4-17. Class D Student Mean Scores of Stress Constructs Construct N Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 101 103 102 103 102 103 52.80 11.78 5.30 10.76 4.63 20.42 13.71 4.25 2.48 3.71 2.46 4.34 26 6 3 4 3 9 94 26 12 20 13 30 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress.

PAGE 136

118 For total student motivation in Class D, there were 102 usable responses. The MSLQ determines level of motivation with si x standardized construc t scores providing a total motivation range of 6 to 48. Students respond to 7-point Likert scale items with one signifying low motivation and seven signify ing high motivation. In Class D, total motivation (M=32.42, SD=3.77, n=102) for the cl ass was 1.96 points higher than the reported total motivation mean from the in strumentÂ’s authors (Pintrich et al. 1991), indicating a higher level of mo tivation in this group. Motivat ion construct mean scores were also higher than the normalized construc t mean scores (Pintrich et al.): extrinsic motivation was 0.68 points higher, task value was 0.27 points higher, control of learning was 0.30 points higher, self-efficacy was 0.38 points higher and test anxiety was 0.49 points higher. The exception to this was intrinsic motivation, which was 0.16 points lower than the norm reported by Pintrich et al. However, all constructs mean scores were within one standard deviation as presented in the study conducted by Pintich et al. The student with the lowest level of motivation in Class D scored 21.26 and the student with the highest level of motivation scored 38.85. See Table 4-18 for reported motivation scales of Class D. Table 4-18. Class D Student Mean Sc ores of Motivation Constructs Construct N Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 102 103 103 102 103 103 103 32.42 4.89 5.71 5.81 6.04 5.85 4.12 3.77 1.00 1.07 1.05 0.83 1.08 1.43 21.26 1.75 2.00 2.00 3.50 2.75 1.00 38.85 6.75 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation.

PAGE 137

119 Concerning student engagement of Cla ss D, 108 participants provided usable responses for this study. In this study, three constructs with 24 items were used from the NSSE to provide a total student engagement score. Items used 4-point scales with one coded as low engagement and four coded as high engagement. The engagement score range was 24 to 96. In Class D, the total calculated student engagement mean was 48.33 (SD=7.89) which was 15.41 points lower than the national reported mean for college seniors (Kuh et al., 2001). Like wise, construct mean scores of student engagement in Class D were also lower than the norm (Kuh et al., 2001): academic challenge was 8.63 points lower, active learning was 6.25 points lo wer, and student faculty interaction was 0.53 points lower. Examining the standard devi ation of student engagement constructs, there was evidence that students in Class D had an average level of student-faculty interaction, but had low leve ls of academic challenge a nd active learning. The least engaged student had a total score of 32 a nd the most engaged student scored 69. See Table 4-19 for student engagement construct scores for Class D. Table 4-19. Class D Student Mean Scores of Engagement Constructs (n=108) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 48.33 24.55 11.67 12.11 7.89 4.22 3.30 3.03 32 16 7 6 69 33 24 22 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class E There were only 32 usable KAI responses from Class E (N=105, n=48). The high number of non-responders may be attributed to the faculty member not allocating extracredit to encourage student participation. The KAI uses 32 items to determine a total

PAGE 138

120 cognitive style score with a range of 32 to 160 and mean of 95. Lowe r scores indicated a more adaptive cognitive style and higher sc ores indicated a more innovative cognitive style. The most adaptive student in Class E had a total cognitive st yle score of 66 while the most innovative student scored 116. In Cl ass E, the total cognitive style score mean was more adaptive (M=89.41, SD=14.95) than th e general population mean reported by Kirton (1999). Likewise, construc t mean scores indicated sli ght adaptiveness with respect to the general population mean (Kirton): sufficiency of orig inality was 2.00 points lower, efficiency was 1.94 points lower and rule/g roup conformity was 1.66 points lower. All construct scores of cognitive style were cons idered consistent with the total cognitive style score (Kirton). See Tabl e 4-20 for findings concerning student cognitive style in Class E. Table 4-20. Class E Student Mean Scores of Cognitive Style Constructs (n=32) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 89.41 39.00 17.06 33.34 14.95 7.87 3.65 7.30 66 25 10 20 116 56 26 46 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, high score equals more innovative. There were 47 usable responses in dete rmining total perceived stress for students of Class E. Level of stress was determined with the SSI using 22 items and a total range of 22 to 110. Students responded to a 5-point Li kert scale with one indicating a low level of perceived stress and five indicating a high level of percei ved stress. The student with the lowest level of perceived stress in Class E scored a 27 for total stress while the student with the highest level of stress scored a 66. In Class E, the total stress score mean was more than two standard deviations lower (M=42.89, SD=11.28, n=47) than the test

PAGE 139

121 sample used by Gadzella and Baloglu (2001). C onstructs of student stress were also had lower mean scores than the test sample c onstructs (Gadzella a nd Baloglu) indicating lower levels of stress in Cl ass E: frustrations was 8.21 poi nts lower, conflicts was 3.75 points lower, pressures was 5.43 points lowe r, changes was 3.58 points lower and selfimposed was 5.19 points lower. It is important to remember that one item was removed from the frustrations construct prior to data collection; however it was still two standard deviations below the test sample (Gadzella and Baloglu). The other four construct mean scores of stress were more than one standa rd deviation less than the mean reported by Gadzella and Baloglu. See Table 4-21 for findings specific to st ress scores of Class E. Table 4-21. Class E Student Mean Scores of Stress Constructs Construct N Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 47 48 47 48 48 48 42.89 9.44 4.68 8.65 3.58 16.63 11.28 3.73 2.15 3.25 1.23 4.70 27 6 3 4 3 7 66 21 11 17 7 29 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. Class E had 48 usable MSLQ responses for determining student motivation. The MSLQ included six standardized constructs to provide a tota l motivation score range of 6 to 48 using 31 items. Students responded to 7-po int Likert scale items with a value of one indicating low motivation and a va lue of seven indicating high motivation. In Class E, the student with the lowest level of motivation scored 19.90 while the student with the highest level of motivation scored 37.57. Th e Class E total motivation score (M=31.20, SD=3.79) was 0.74 points higher than the total mean score reported by Pintrich et al. (1991). Four motivation scale mean scores meas ured by the MSLQ were also higher than

PAGE 140

122 the test sample (Pintrich et al.) included extrinsic motivation (0.19 points higher), task value (0.04 points higher), control of lear ning (0.28 points higher) and self-efficacy (0.77 points higher). Motivation scales means in Cla ss E that were lower than the test sample included intrinsic motivation (0.47 points lowe r) and test anxiety (0.06 points lower). All motivation scale mean scores measured in Cla ss E were within one st andard deviation of the test sample (Pintrich et al.). See Tabl e 4-22 for reported motivation mean scores of Class E. Table 4-22. Class E Student Mean Scores of Motivation Constructs (n=48) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 31.20 4.58 5.22 5.58 6.02 6.24 3.57 3.79 1.15 1.05 1.19 0.93 0.77 1.30 19.90 2.00 2.75 2.00 2.50 3.50 1.00 37.57 6.50 7.00 7.00 7.00 7.00 6.60 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. There were 48 usable NSSE responses in Class E to determine level of student engagement. In this study, constructs of academic challenge, active and collaborative learning and student-faculty interaction formed a total student engagement score with a range of 24 to 96 using 24 items. Items were c oded in order that lower scores signify less student engagement while higher scores si gnify more student engagement. The least engaged student in Class E had a total enga gement score of 32 and the most engaged student scored 59 for total engagement. The Cl ass E total student engagement mean score was 47.88 (SD=5.50), which was 15.86 points lowe r than the national reported college senior total engagement mean (Kuh et al ., 2001) indicating a lowe r level of student engagement in Class E. Constructs of stude nt engagement as meas ured with the NSSE

PAGE 141

123 were also lower than national average for seniors (Kuh, et al.): academic challenge was 9.14 points lower, active learning was 5.32 point s lower and student-faculty interaction was 1.41 points lower. Examining the standa rd deviation of the construct scores measured in Class E, there was evidence that students in Class E were less academically challenged. See Table 4-23 for findings dete rmining Class E student engagement. Table 4-23. Class E Student Mean Scores of Engagement Constructs (n=48) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 47.88 24.04 12.60 11.23 5.50 3.60 2.16 2.34 32 15 8 6 59 32 17 17 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class F For Class F (N=150, n=115), there were 71 KAI responses that were usable for this study. The KAI determined studentsÂ’ c ognitive style level with 32 items. The theoretical range of the KAI was 32 to 160 with a mean of 95 (Kirton, 1999). Items were coded in order that lower scores indicated a more adaptive cognitive style and higher scores indicated a more innovative cognitive st yle. For Class F, the total cognitive style mean (M=90.17, SD=16.28) was approximately 4.83 points lower than the general population reported by Kirton ( 1999). That is, responding students of Class F were slightly more adaptive. Cognitive style construct mean scores also indicated students were slightly more adaptive as those mean scores were lower than Kirton specifies. In Class F these constructs include: sufficiency of originality which was 1.80 points lower, efficiency which was 0.85 points lower and rule/group conformity which was 2.18 points lower than the general population (Kirton). All constructs of cognitive style were

PAGE 142

124 consistent with a total cognitive style scor e of 90.17 (Kirton). The most adaptive student in Course F had a total cognitive style scor e of 55 while the most innovative student scored 132 on the KAI. See Table 4-24 for fi ndings specific to Class F student respondent cognitive style. Table 4-24. Class F Student Mean Scores of Cognitive Style Constructs (n=71) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 90.17 39.20 18.15 32.82 16.28 8.89 4.01 6.81 55 21 10 17 132 60 26 51 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. Class F had 112 usable responses for de termining classroom stress. Level of perceived stress was determined with the SSI The SSI used 22 items incorporated into five constructs. Students completing the SSI responded to 5-point Likert scale items providing a range of 22 to 110 for a total perc eived stress score. The Class F total mean stress score was 49.29 (SD=13.20) which was 20.80 points lower than results provided by Gadzella and Baloglu (2001). All construc ts of perceived stress likewise indicated lower levels of student stress in Class F. Compared to the te st sample mean (Gadzella and Baloglu), Class F mean scores of stress c onstructs frustrations was 6.63 points lower, conflicts was 3.26 points lower, pressures wa s 4.43 points lower, changes was 3.48 points lower and self-imposed was 2.92 points lower. No te that the stress construct frustrations had one item removed by the re searcher prior to data collection. All the aforementioned stress constructs in Class F were more than one standard deviation lower than the test sample except the construct self-imposed stre ss. The student with the lowest level of

PAGE 143

125 perceived stress had a total SSI score of 25 wh ile the student with the highest level of perceived stress scored 89. See Table 4-25 for specific findings regarding Class F stress. Table 4-25. Class F Student Mean Scor es of Stress Constructs (n=112) Construct Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 49.29 11.02 5.17 9.65 4.55 18.90 13.20 4.02 2.64 3.69 2.53 4.60 25 6 3 4 3 8 89 27 12 20 13 28 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. Class F had 115 usable responses in de termining student motivation. Motivation was determined with the MSLQ which consis ted of 31 items forming six standardized motivation construct scores. Thus, the total motivation score range was 6 to 48 with higher scores signifying more mo tivation. The student in Class F with the lowest level of motivation had a total score of 20.40 points on the MSLQ, while th e student with the highest level of motivation sc ored a total 37.35 points. In Cl ass F, the total motivation mean (31.25, SD=3.71) was 0.79 points higher to that reported by Pintri ch et al. (1991). Four of the motivation construc ts had higher mean scores than reported by Pintrich et al. including extrinsic motivation (0.48 points higher), task value (0.26 points higher), control of learning (0.08 points higher) and self-efficacy (0.47 points higher). The other two constructs had lower mean scores than reported by Pintrech et al.: intrinsic motivation was 0.37 points lower and test a nxiety was 0.14 points lower. All motivation construct means for Class F were within one standard deviation of the instrument norm (Pintrich et al.). See Table 4-26 for re sults of Class F student motivation.

PAGE 144

126 Table 4-26. Class F Student Mean Scores of Motivation Constructs (n=115) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 31.25 4.68 5.51 5.80 5.82 5.94 3.49 3.71 1.10 0.96 1.13 0.88 0.80 1.39 20.40 1.25 2.25 1.33 3.75 3.75 1.00 37.35 6.75 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. Class F had 115 usable responses to th e NSSE for the purpose of determining level of student engagement. The NSSE co mprised of 24, 4-point scale items providing for a possible total engagement score range of 24 to 96. Items were coded so as higher scores indicated higher levels of student engagement. The stud ent with the lowest level of engagement had a total engagement score of 33 while the most engaged student had a score of 74. In Class F, respondentsÂ’ total student engagement mean was 47.87 (SD=7.81), a total score that was 15.87 points lower than the national mean for college seniors (Kuh et al., 2001). Construct mean scor es of student engagement in Class F were also lower than the average college senior (Kuh et al.): academic challenge was 7.32 points lower, active learning was 6.69 points lo wer and student-facu lty interaction was 1.87 points lower. This provides evidence that students in Class F had lower levels of total student engagement with emphasis placed on the construct of academic challenge. See Table 4-27 for reported means of student engagement in Class F.

PAGE 145

127 Table 4-27. Class F Student Mean Scores of Engagement Constructs (n=115) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 47.87 25.86 11.23 10.77 7.81 4.28 2.98 3.09 33 14 7 6 74 38 22 20 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class G In Class G (N=110, n=85), the researcher found 65 of the KAI scores acceptable for the purpose of determining students cogniti ve style. The KAI consists of 32 items and three constructs. It has a theoretical range of 32 to 160 and a mean of 95. Lower scores indicated that an individual was more adap tive while higher scores indicated that an individual was more innovative. The most ad aptive student in Class G had a total KAI score of 52, while the most innovative stude nt scored 125. Class G had a total cognitive style mean score of 93.42 (SD=15.04) which was 1.58 points lower than the general population according to Kirton (1999). All construct mean scores of student cognitive style in Class G were less than one poin t lower than the gene ral population construct means (Kirton). See Table 4-28 for findings of student cognitive style in Class G. Table 4-28. Class G Student Mean Scores of Cognitive Style Constructs (n=65) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 93.42 40.18 18.31 34.92 15.04 7.67 4.76 7.85 52 21 9 16 125 56 29 54 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. Considering student stress for Class G, there were 84 usable responses. Total perceived level of stress was determined with the SSI which used 22 items to form a total

PAGE 146

128 level of perceived stress range of 22 to 110 points. Students completed the SSI by responding to 5-point Likert scale items with one signifying low stress and five signifying high stress. The student with the lo west level of percei ved stress in Class G had a total stress score of 27 points while the student with the highest level of perceived stress scored a total of 92 points on the SSI. The Class G total mean stress score was 18.55 points lower (M=51.54, SD=13.72) than th e results published by Gadzella and Baloglu (2001). Construct mean scores of st udent stress also indicated lower student stress levels as mean scores were lower than that reported by Ga dzella and Baloglu: frustrations was 6.14 points lower, conflicts was 2.51 points lower, pressures was 3.31 points lower, changes was 2.86 points lower and self-imposed was 3.65 points lower. Note that one item was removed from the frus trations stress cons truct prior to data collection. See Table 4-29 for reported mean scores of student stress of Class G. Table 4-29. Class G Student Mean Scor es of Stress Constructs (n=84) Construct Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 51.54 11.51 5.92 10.77 5.17 18.17 13.72 3.93 2.61 3.95 2.94 4.70 27 6 3 4 3 6 92 23 13 20 15 30 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. Considering student motivation in Class G, 85 student responses were considered usable. The MSLQ was used to determine le vel of studentsÂ’ motivation and totaled 31 Likert scale items with a valu e range of 1 to 7. Constructs were standardized providing a total MSLQ range of 6 to 48. A response of one indicated low motiv ation and a response of seven indicated high motivation. The stude nt in Class G with the least amount of

PAGE 147

129 motivation had a total score of 12.60 while the student with the most amount of motivation scored 39.15 for total motivation. For Class G, the total motivation score mean was 30.01 (SD 4.27), which was 0.45 points lower than the tota l motivation norm provided by Pintrich et al. (1991) In this class, three constr uct mean scores were lower than the norm construct mean scores (Pintr ich et al.): intrinsic motivation was 0.45 points lower, task value was 0.43 points lower a nd control for learning was 0.28 points lower. The remaining three constructs were highe r than the norm construct mean scores (Pintrich et al.) indicating s lightly higher levels of motivation: extrinsic motivation was 0.07 points higher, self-efficacy was 0.51 point s higher and test anxiety was 0.13 points higher. All motivation construct score means we re within one standard deviation from the normalized values. See Table 4-30 for scale mean scores of Class G participants. Table 4-30. Class G Student Mean Scores of Motivation Constructs (n=85) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 30.01 4.60 5.10 5.11 5.46 5.98 3.76 4.27 1.15 1.24 1.28 1.14 0.92 1.39 12.60 1.75 1.00 1.00 1.00 3.25 1.00 39.15 6.75 7.00 7.00 7.00 7.00 6.80 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. There were 24 items used from the NSSE to determine level of student engagement in Class G. St udents responded to items using a 4-point scale with one signifying low engagement a nd four signifying high engagement. The range for total student engagement was 24 to 96. The least e ngaged student in Class G had a total score of 34 points while the student with the highest level of engagement scored 79 on the NSSE. For Class G, there were 85 usable re sponse providing a to tal engagement mean

PAGE 148

130 score of 50.05 (SD=7.80) which was 13.69 points lower than the national average for college seniors (Kuh et al., 2001). Note that the total scores for student engagement consisted of the three constructs measured in this study. Construct m ean scores of student engagement were also lower than the nationa l college senior average (Kuh et al.) as academic challenge was 8.66 points lower, active learning was 3.64 points lower and student-faculty inte raction was 1.39 points lower. This pr ovides evidence that students in Class G had low levels of academic challeng e. See Table 4-31 for findings specific to Class G student engagement. Table 4-31. Class G Student Mean Scores of Engagement Constructs (n=85) Construct Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 50.05 24.52 14.28 11.25 7.80 4.22 2.99 2.87 34 16 9 6 79 37 25 18 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. Class H For Class H (N=122, n=70), 50 KAI scores were found acceptable to determine studentsÂ’ cognitive style. The KAI has 32 item s with a theoretical range of 32 to 160 and mean of 95. A lower score signifies a more adaptive cognitive st yle and a higher score signifies a more innovative cogni tive style. The most adapti ve student in Class H had a total cognitive style score of 67, while the most innovative st udent had a total cognitive style score of 135. The total cognitive style mean of Class H was 95.30 (SD=15.01) which was 0.30 points higher than the genera l population mean determined by Kirton (1999). Cognitive style construct score means were not more than one point lower or

PAGE 149

131 higher than norms reported by Kirton. See Ta ble 4-32 for results specific to student cognitive style in Class H. Table 4-32. Class H Student Mean Scores of Cognitive Style Constructs (n=50) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 95.30 41.24 18.64 35.42 15.01 8.65 3.75 7.04 67 22 10 19 135 60 27 51 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. There were 70 students providi ng usable scores to determine stress in Class H. Level of perceived stress was determined us ing the SSI, an instrument consisting of 22 items. Students responded to items with a 5point Likert scale w ith one signifying low stress and five signifying high stress. In Cla ss H, the student with the lowest level of perceived stress had a total score of 30 while the student with the highest level of perceived stress scored 78. Class H had a to tal student perceived stress mean of M=50.73 (SD=11.54) which was 19.36 points lower than the total stress mean reported by Gadzella and Baloglu (2001). Likewise, c onstructs of student stress ha d mean scores lower than normalized values (Gadzella and Baloglu): fr ustrations was 5.85 poi nts lower, conflicts was 2.99 points lower, pressures was 4.29 poi nts lower, changes was 3.61 points lower and self-imposed was 2.55 points lower. Note th at the frustrations stress construct had one item removed prior to data collection wh ich lowers the measured construct in the study in comparison to the work conducted by Gadzella and Baloglu. Interestingly in Class H, all stress construct mean scores we re more than one standard deviation lower than test norms (Gadzella and Baloglu) excep t for the construct self-imposed stress. See Table 4-33 for Class H responses concerning student stress.

PAGE 150

132 Table 4-33. Class H Student Mean Scor es of Stress Constructs (n=70) Construct Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 50.73 11.80 5.44 9.79 4.42 19.27 11.54 3.33 2.53 3.08 2.24 5.10 30 6 3 4 3 8 78 22 15 17 12 30 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. Considering student motivation in Class H, 70 responders returned usable MSLQ instruments. The MSLQ consists of 31 items which form six standardized constructs. Constructs were then summed to provide a total motivation score. Students respond to 7point Likert scale items with one signifyi ng low motivation and seven signifying high motivation. The least motivated student in Class H had a total motivation score of 17.25 while the most motivated student scored 39.80 for total motivation. In Class H, the total motivation mean was 31.05 (SD=4.60), wh ich was 0.59 points higher than the normalized mean reported by Pintrich et al. (1991). None of the motivation construct mean scores were more than 0.50 points higher or lower than the normalized mean scores (Pintrich et al.). See Table 4-34 for reported mean scores of motivation of Class H. Table 4-34. Class H Student Mean Scores of Motivation Constructs (n=70) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 31.05 4.85 5.50 5.56 5.69 5.48 3.96 4.60 1.16 1.10 1.18 1.02 1.08 1.49 17.25 2.00 2.50 1.50 1.75 2.25 1.00 39.80 7.00 7.00 7.00 7.00 7.00 6.60 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation.

PAGE 151

133 There were between 69 and 70 usable N SSE responses for determining student engagement in Class H. Twenty-four items were used from the NSSE to form three constructs and a total student engagement sc ore. Items were coded with one indicating low engagement and four indicating high enga gement using a 4-point scale. The total possible range for student engagement was 24 to 96. The student least engaged in Class H had a total engagement score of 34 while the st udent most engaged in the class scored 72 for total student engagement. For Class H, the total student enga gement mean (M=52.28, SD=8.59, n=69) was 11.46 points lower than th e national reported mean for college seniors (Kuh et al., 2001). Considering the cons tructs of student engagement measured in this study mean scores were also lower than the national mean for college seniors (Kuh et al.): academic challenge was 6.29 points lower, active learning was 4.38 points lower and student faculty inter action was 0.70 points lower. Examin ing the standard deviation of student engagement constructs measured in Class H, there is evidence that in Class H there was higher levels of student-faculty interaction and lower levels of academic challenge and academic learning. See Table 435 for reported student engagement mean scores of Class H. Table 4-35. Class H Student Mean Sc ores of Engagement Constructs Construct N Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 69 70 70 69 52.28 26.89 13.54 11.94 8.59 3.61 3.28 3.60 34 18 9 6 72 34 23 21 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement.

PAGE 152

134 Class I For Class I (N=100, n=77), there were 60 usable KAI responses to determine studentsÂ’ cognitive style. The KAI has 32 items providing for a theoretical range of 32 to 160 and a mean of 95. Lower scores indicate d that a person was more adaptive while higher scores indicated that a person was more innovative. Th e most adaptive student in Class I had a total cognitive st yle score of 50 while the mo st innovative student had a total score of 124 on the KAI. For Class I, the total cognitive style mean was 92.90 (SD=14.25), which was 2.10 points lower than th e general population reported by Kirton (1999). Comparing Class I c ognitive style construct mean scores to the general population construct mean scores found suffici ency of originality 0.13 points higher, efficiency 1.37 points lower and rule/group conformity 0.87 points lower. However all cognitive style construct scores were consis tent with a total scor e of 92.90 (Kirton). See Table 4-36 for reported cognitive style scale means determined for Class I. Table 4-36. Class I Student Mean Scores of Cognitive Style Constructs (n=60) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 92.90 41.13 17.63 34.13 14.25 7.61 4.77 7.42 50 25 7 18 124 58 29 53 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. Class I had 77 acceptable responses to the SSI in determining student perceived stress. The SSI used 22, 5-point Likert scal e items to which stude nts responded with a one to signify low stress and five to signi fy high stress. The possible range for total student stress was 22 to 110. In Class I, the student with the least amount of perceived stress had a score of 29 while the student with the highest level of perceived stress had a

PAGE 153

135 total score of 80. For Class I, the total mean stress score was 49.82 (SD=11.96) which was 20.27 points lower than the sample fi ndings of Gadzella and Baloglu (2001). Construct scores of student stress were also lower than the test sample (Gadzella and Baloglu): frustrations was 6.77 points lower, conflicts was 2.49 point s lower, pressures was 4.26 points lower, changes was 3.39 point s lower and self-imposed was 3.27 points lower. The frustrations construct was more than one standard devi ation lower than the norm, however one item was removed from this summative scale prio r to data collection. The student stress constructs conflicts, pr essures and changes were also one standard deviation lower than the norm, however sel f-imposed stress was within the instrument reported standard deviation. See Table 4-37 fo r findings specific to Class I respondentsÂ’ stress. Table 4-37. Class I Student Mean Scor es of Stress Constructs (n=77) Construct Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 49.82 10.88 5.94 9.82 4.64 18.55 11.96 4.15 2.58 2.87 2.35 4.87 29 6 3 4 3 7 80 22 14 16 14 27 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. For motivation, 77 responses were considered usable. The MSLQ was used to determine motivation and consisted of six sta ndardized constructs and totaled 31 items. The possible range for total motivation was 6 to 48. The least motivated student of Class I had a total score of 19.76 while the most mo tivated student scored 37.75 points for total motivation. In Class I, the total motiva tion mean was 29.86 (SD=3.74) which was 0.60 points lower than the total motivation mean re ported by Pintrich et al. (1991). Construct

PAGE 154

136 scores of student motivation in Class I vari ed positively and negatively but differed less than one point from the norm (Pintrich et al.). Specifically compared to the norm, intrinsic motivation was 0.74 points lower, extrinsic motivation was 0.35 points higher, task value was 0.69 points lower, control of learning was 0.21 points lower, self-efficacy was 0.66 points higher and test anxiety was 0.03 points hi gher. See Table 4-38 for reported findings of motivation in Class I. Table 4-38. Class I Student Mean Scores of Motivation Constructs (n=77) Construct Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 29.86 4.31 5.38 4.85 5.53 6.13 3.66 3.74 1.00 1.15 1.22 0.99 0.87 1.66 19.76 1.50 2.50 1.00 2.50 3.75 1.00 37.75 6.25 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. Student engagement was determined using 24 items from the NSSE which formed three constructs. The possible range for tota l engagement was 24 to 96. Items were coded with one signifying low engagement and f our signifying high engagement. The least engaged student of Class I had a total enga gement score of 36 points while the most engaged student of the class scored 71 points for total engagement. Class I had 76 responses in determining student total e ngagement. The total engagement mean was 51.58 (SD=7.72) which was 12.16 points lower than the national mean for college seniors (Kuh et al, 2001). Construct scores of stude nt engagement were also lower than the national average for college seniors (Kuh et al.): academic challenge was 7.69 points lower, active learning was 4.06 points lower and student-faculty interaction was 0.43 points lower. This provides evidence that Cl ass I students had lower levels of academic

PAGE 155

137 challenge and active learning, but had an averag e level of student-f aculty interaction. See Table 4-39 for reported construct scores of Class I student engagement. Table 4-39. Class I Student Mean Sc ores of Engagement Constructs Construct N Mean SD Min Max Total student engagement Academic challenge Active learning Student-faculty interaction 76 76 77 77 51.58 25.49 13.86 12.21 7.72 4.47 2.65 2.53 36 17 9 7 71 36 22 19 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement. All Students Students were grouped together (N=993, n= 716) to determine values of student cognitive style, stress, motivation and engage ment for all students participating in the study. There were 511 acceptable KAI responses to be used in determining student cognitive style. There are 32 items in the KAI providing a theoretical range of 32 to 160 with a mean of 95. Lower scores indicated an individual was more adaptive while higher scores indicated an individua l was more innovative. The most adaptive student in this study had a total KAI score of 46 while the mo st innovative student had a cognitive style score of 142. The mean of the total group was 93.28 (SD=15.95) which was 1.72 points lower than the general popul ation defined by Kirton (1999). Likewise, construct mean scores for cognitive style also indicated slight adaptiveness as scores were lower than the general population mean (Kirton): sufficien cy of originality was 0.19 points lower, efficiency was 1.11 points lower and rule/g roup conformity was 0.42 points lower. All construct scores of cognitive style were cons istent with a total cognitive style score of 93.28 (Kirton). See Table 4-40 for the total student c ognitive style for all participants of this study.

PAGE 156

138 Table 4-40. All Student Mean Scores of Cognitive Style Constructs (n=511) Construct Mean SD Min Max Total cognitive style Sufficiency of originality Efficiency Rule/Group conformity 93.28 40.81 17.89 34.58 15.95 8.19 4.58 7.52 46 21 7 16 142 62 32 58 Note. Cognitive style measured by the KAI with 32 items. Theoretical range: Total (32160), Sufficiency of Originality (13-65), Effi ciency (7-35) and Rule/Group Conformity (12-60). Coded: lower score equals more ad aptive, higher score equals more innovative. There were 697 usable SSI scores in de termining total perc eived stress for all participants. The SSI consisted of 22 summate d 5-point Likert scale items coded as one signifying low stress and five signifying high st ress. The student with the lowest level of perceived stress in this study scored 25 while the student with the highest level of perceived stress scored 94 for total percei ved stress. The total stress mean was 51.35 (SD=12.91), which was 18.74 points lower than the reported norm (Gadzella & Baloglu, 2001). Construct mean scores of student stress were also lower than the norm (Gadzella & Baloglu): frustrations was 5.96 points lower, conflicts wa s 2.82 points lower, pressures was 3.77 points lower, changes was 3.35 point s lower and self-imposed was 2.72 points lower. Note that one item was removed from the frustrations construct prior to data collection, which was discussed in chapter 3. This modification affects the comparison between norm values and the a ssessment of studen t stress in this st udy by lowering the summated scale for the frustrations stress construct as well as the total stress score. Of the constructs assessing student stre ss, all were more than one standard deviation lower than norm values except for self-imposed stress. See Table 4-41 for findings of stress of all participants of this study. There were 706 usable MSLQ scores fo r determining student motivation of all participants. The MSLQ consists of six standard ized constructs summed to provide a total

PAGE 157

139 Table 4-41. All Student Mean Sc ores of Stress Constructs Construct N Mean SD Min Max Total stress Frustrations Conflicts Pressures Changes Self-imposed 697 702 700 702 702 703 51.35 11.69 5.61 10.31 4.68 19.10 12.91 4.23 2.59 3.57 2.43 4.69 25 6 3 4 3 6 94 29 15 20 15 30 Note. Perceived stress was measured with th e SSI using 22 summated items. Possible range: Total (22 to 110), Frustr ations (6-30), Conf licts (3-15), Pressu res (4-20), Changes (3-15), Self-imposed (6-30). C oded: higher score equals higher level of perceived stress. motivation score and possible range of 6 to 48. The instrument uses 31, 7-point Likert scale items. The least motivated student par ticipating in the study had a total score of 12.60 while the most motivated student in th e study scored 39.80 in determining total motivation. Considering all part icipating students, the total mean motivation score was 30.81 (SD=4.04, n=706), which was 0.35 points higher than the norm reported by Pintrich et al. (1991). Comparing constructs of the mean scores with norm values (Pintrich et al.), motivation constructs of participating students varied positively and negatively within one standard deviation of the norm construct value. Specifically participating students reported the following construct mean sc ores with respect to values reported by Pintrich et al.: intrinsic motivation was 0.44 points lower, extrinsic motivation was 0.39 points higher, task value was 0.23 points lower, control of learning was .02 points higher, self-efficacy was 0.41 points higher and test anxiety was 0.20 points higher. See Table 4-42 for findings specific to student stress of all participants. For total student engagement there were 712 responses to the NSSE considered usable for this study. Twenty-four items were used from the NSSE to measure total student engagement. To complete the NSSE students completed items using a 4-point scale with one signifying low engagement and four signifying high engagement. The

PAGE 158

140 Table 4-42. All Student Mean Scor es of Motivation Constructs Construct N Mean SD Min Max Total motivation Intrinsic motivation Extrinsic motivation Task value Control of learning Self-efficacy Test anxiety 706 709 708 707 709 707 708 30.81 4.61 5.42 5.31 5.76 5.88 3.83 4.04 1.10 1.10 1.26 0.99 0.97 1.48 12.60 1.25 1.00 1.00 1.00 2.25 1.00 39.80 7.00 7.00 7.00 7.00 7.00 7.00 Note. Motivation was measured with the MS LQ using 31 items with standardized constructs. Possible range: Total motivation (6 -48), all constructs (1-7). Coded: higher score equals higher level of motivation. least engaged student in this study had a total score of 32 wh ile the most engaged student had a total engagement score of 80. The total me an for student engagement of all students was 49.58 (SD=8.20, n=712) which was 14.16 points lower than the national mean for college seniors (Kuh et al., 2001). Constructs of student engagement we re also lower than the national mean for college seniors (Kuh et al.): academic challenge was 7.99 points lower, active learning was 5.26 points lower and student-faculty interaction was 0.90 points lower. These construct mean score di fferences provide evid ence that student respondents had low levels of academic cha llenge and active learning, but an average level of student-faculty interaction. See Tabl e 4-43 for specific findings regarding student engagement of all participants of this study. Table 4-43. All Student Mean Scor es of Engagement Constructs Construct N Mean SD Min Max Range Total student engagement Academic challenge Active learning Student-faculty interaction 712 713 714 713 49.58 25.19 12.66 11.74 8.20 4.33 3.22 3.08 32 14 7 6 80 38 25 22 48 24 18 16 Note. Engagement was measured by the NSSE with 24 summated items. Possible range: Total Engagement (24-96), Academic Cha llenge (11-44), Active Learning (7-28), Student-Faculty Interaction (6-24). Coded: higher score equals higher level of engagement.

PAGE 159

141 Comparisons Between Classes The literature provided evidence that student s had more stress in courses requiring characteristics of adaptiveness (Puccio, Talbot & Joniak, 1993). To test this, a one-way ANOVA was conducted to examine significant di fferences of student stress levels between students taught by adaptive faculty members (Classes A, B, & C), middle score faculty members (Classes D, E & F) and i nnovative faculty members (Classes G, H & I). With total stress designated as the dependent variable, a significant difference was found between the three groups (F=8.92, p=.00). A Bonf erroni post-hoc test was conducted to further examine between group differences. Th e adaptive teacher group had student stress scores (M=54.51) that were significantly high er than the student stress scores of the middle score teacher group (M=49.49, p=.00) and si gnificantly higher than student stress scores of the innovative teacher grou p (M=50.72, p=.01). However, there was no significant difference between st ress scores for students in the middle score teacher group and the innovative teacher group (p=.87). Th is finding confirms the results found by previous research (Puccio et al., 1993) A one-way ANOVA was conducted to examin e significant differences of student motivation scores between c ourses taught by adaptive teacher group (Classes A, B & C), courses taught by the middle score teacher gr oup (Classes D, E & F) and courses taught by the innovative teacher group (Classes G, H & I). With total motivation as the dependent variable, a significant diff erence was found between the three groups (F=10.41, p=.00). A Bonferroni post-hoc test was conducted to further examine the significant differences between each of th e three groups. Students in the innovative teacher group had total motivation scores (M =30.27) that were significantly lower than student motivation scores in the mi ddle score teacher group (M=31.69, p=.00), a

PAGE 160

142 difference of 1.42 points on a 36 point scale. St udent respondents in the adaptive teacher group had motivation scores (M=30.28) that were also significantly lower than motivation scores in the middle score teach er group (p=.00). There was no significant difference for student motivation scores between the adaptive teacher group and the innovative teacher group (p=.99). This indicated that in these classes, students were less motivated in classes taught by more ad aptive faculty members and more innovative faculty members. To determine if student engagement va ried among adaptive faculty members and innovative faculty members, a one-way AN OVA was conducted to examine significant differences between an adaptive teacher group (Classes A, B & C), a middle score teacher group (Classes D, E & F) and an innovative teacher group (Classes G, H & I). Using total student engagement as the depe ndent variable, a signi ficant difference was found between groups (F=9.61, p=.00). A Bonfe rroni post-hoc test was conducted to further examine group differences. Students in the innovative teacher group had total engagement scores (M=51.22) significantly hi gher than engagement scores of students enrolled in the middle score teacher gr oup (M=48.05, p=.00). However, students in the innovative teacher group did not have signi ficantly higher engagement scores than students in the adaptive teacher grou p (M=49.75, p=.17). There was no significant difference for student engagement scores be tween the middle score teacher group and the adaptive teacher group (p=.07). This finding i ndicated that students in these classes taught by innovative faculty members tended to have higher engagement scores. In the nine classes utilized in this st udy, total student engagement mean scores were lower than the national average for co llege seniors (Kuh et al., 2001). However,

PAGE 161

143 only 41.06% of the total responde nts were seniors. It may be that freshman, sophomore and junior students lowered student engagement mean scores in each of the classes since college freshman tend to have lower levels of engagement than co llege seniors (Kuh et al.). A one-way ANOVA was conducted to exam ine if significant differences of student engagement scores existed between college freshman, sophomores, juniors and seniors. Significant differences were found among th e four groups of students (F=2.75, p=.04). However the assumption of homogeneity of variance was not met as LeveneÂ’s Test of Equality of Error Variances indicated to re ject the null hypothesis, (F=3.00, p=.03) and conclude that the four groups of students were not equal regarding variance. Because variances among the four groups of student s were not homogenous and sample sizes differed among the four groups, a Games-Howell post-hoc test was conducted to account for these differences. In this analysis, college senior student engagement scores were not significantly different from j unior student engagement scores (p=.07), sophomore student engagement scores (p=.99) or freshman student engagement scores (p=.87). The researcher concluded that low scores of st udent engagement found in this study were not attributed to colle ge classification. Summary of Findings for Objective Two In determining studentsÂ’ cognitive style seven classes had total score means that were slightly adaptive (below the mean of 95), one class had total cognitive style score means that were slightly innovative (above th e mean of 95), and one class had a cognitive style score approximating 95. Although self-selec ted courses may have a tendency to be more adaptive or more innovative, the total m ean cognitive style scores of these classes were indicative of the percentage of females in the course. Kirton ( 1999) states that the total cognitive style mean score for females is 90, which is five points lower than the

PAGE 162

144 general population mean. In th is study, the overall cognitive style mean score for females was 91.49 (n=339), while the cognitive style mean score for males was 96.81 (n=170). Therefore, classes with a larger percentage of females tend to have more adaptive mean scores. Note that Class C was 72% male and had the most innovative total cognitive style mean score (M=100.86) of the nine classes examined. In all classes, cognitive style construc t mean scores were consistent for the respective total score (Kirton, 1999) However, it is interesting to note that the efficiency cognitive style construct was the most adaptive construct in Classes A, B, C, D, H and I. Among these six classes the difference between the efficiency cognitive style construct and the next most adaptive cognitive style construct ranged from .60 (Class H) to 2.00 (Class D). Non-response error limited the findings of this study as not every student completed all four instruments. Particularl y, non-response error may be higher in Class E given that less than 50% of the students responded. The low re sponse rate may be attributed to the faculty member not provi ding extra-credit points to participating students. No assessment for non-response erro r was conducted as explained in chapter 3. The KAI suffered a 29% mortality rate in this study. The KAI was scored by the researcher with the guidelines presented by Kirton (1999). Many of these scores were discarded as students tended to answer most of the items as easy or very easy indicating that all tasks were easy to perform over a l ong period of time. This left few or no marks to signify a specific task as hard or very hard to exhibit over a long period of time. Because everyone has difficulty with some aspect of solving a problem, students marking

PAGE 163

145 all items as easy or very easy were not providi ng a true measure of cognitive style, but were confounding the measure with cognitive level (Kirton). For student stress, there wa s a tendency for perceived stress mean scores to be lower than the reported norm by Gadzella and Baloglu (2001). One sm all contributor to this was the removal of one fr ustrations stress construct it em prior to data collection. Another contributor may be the time of year at which the instrument was administered. There was no indication of the semester that Gadzella and Balogl u conducted their study which provided instrument norms, but Shield s (2001) found that stud ent stress decreases during the second semester of the academic y ear. This may be why student stress mean scores in this study were low when compared to Gadzella and Balogl u as these data were collected during the spring semester. A one-way ANOVA was conducted to examin e significant differences of student stress levels between students taught by ad aptive faculty members, middle score faculty members and innovative faculty members. With total stress designate d as the dependent variable, a significant difference was found between the three gr oups (F=8.92, p=.00). A Bonferroni post-hoc test was conducted to further examine between group differences. The adaptive teacher group had student stress scores (M=54.51) that were on average significantly higher than the student stress scores of th e middle score teacher group (M=49.49, p=.00) and on average significantly higher than student stress scores of the innovative teacher group (M=50.72, p=.01). Th is finding confirms the results found by previous research (Puccio, Ta lbot, & Joniak, 1993) and indi cates a relationship between student stress and courses taught by adaptive faculty members.

PAGE 164

146 The MSLQ was used to determine level of student motivation in each course. Throughout the nine courses total motiva tion means ranged from 29.31 (Class A) to 32.42 (Class D) with no construct mean sc ore in any class exceeding one standard deviation from the norm (Pintrich et al., 1991). A one-way ANOVA was conducted to examin e significant differences of student motivation scores between courses taught by adaptive teacher group, middle score teacher group and innovative teacher group. With total motivation as the dependent variable, a significant difference was found between the three gr oups (F=10.41, p=.00). A Bonferroni post-hoc test was conducted to fu rther examine the significant differences between each group. Students in the innovative teacher group had a mean score of 30.27 which was significantly lower than student mo tivation mean scores in the middle score teacher group (M=31.69, p=.00), but not signifi cantly lower than the adaptive teacher group (M=30.28, p=.99). The evidence suggests th at students were less motivated in classes taught by more adaptive faculty me mbers and more innovative faculty members. Mean scores of total student engagement were low compared to national score means of college seniors. Kuh (2001) states th at engagement scores tend to increase with college classification. However, this was not found to be true in this study as no significant differences were found between st udent engagement scores of seniors and engagement scores of freshman, sophomores or juniors. That is, there was no evidence of this relationship among these data. A one-way ANOVA was conducted to examin e significant differences of student engagement between the adaptive teac her group, middle score teacher group and innovative teacher group. A significant di fference was found between groups (F=9.61,

PAGE 165

147 p=.00). A Bonferroni post-hoc test was conducted to furt her examine group differences. Students in the innovative teacher group had total engagement scores (M=51.22) significantly higher than engagement scores of students enrolled in the middle score teacher group (M=48.05, p=.00). This finding in dicated that students in these classes taught by innovative faculty members tended to have higher engagement scores. See Table 4-44 for a summary of student to tal mean scores for cognitive style, stress, motivation and student engagement per class. Table 4-44. Summary of Classe s Regarding Student Cognitive Style, Stress, Motivation and Engagement Mean Scores Course Total Cognitive Style Mean Total Stress Mean Total Motivation Mean Total Engagement Mean Class A Class B Class C Class D Class E Class F Class G Class H Class I All Classes 90.60 92.43 100.86 93.63 89.41 90.17 93.42 95.30 92.90 93.28 52.44 55.85 54.91 52.80 42.89 49.29 51.54 50.73 49.82 51.35 29.31 30.65 30.86 32.42 31.20 31.25 30.01 31.05 29.86 30.81 48.43 46.01 54.73 48.33 47.88 47.87 50.05 52.28 51.58 49.58 Objective Three Determine the cognitive style gap between fa culty and students and to explore its relationship with student stress, stude nt motivation, student engagement and specific demographic variables of undergraduate students. Student and faculty member participants in each class were asked to respond to the KAI to determine their cognitive style. The faculty memberÂ’s total cognitive style score and construct scores we re subtracted from studentsÂ’ total cognitive style and construct scores which provided cognitive st yle gap scores for each individual student. Once cognitive style gap scores were cal culated between each student and their

PAGE 166

148 respective teacher, PearsonÂ’s correlation coeffi cient was used to determine relationships between cognitive style gap and student stress, motivation, engagement and selected demographic variables. Gender was the only categorical variable and was dummy coded one for males and two for females. Mill er (1998) provided a classification for determining the magnitude of a correlation w ith r=0.01 to 0.09 described as negligible, r=0.10 to 0.29 described as low, r=0.30 to 0.49 described as moderate, r=0.50 to 0.69 described as substantial, r=0.70 to 0.99 desc ribed as very high, and r=1.0 described as perfect. Class A There were 58 usable KAI responses fo r determining cognitive style gap of students in Class A. The faculty member had a total cognitiv e style score of 64, indicating that she was more adaptive. The reported total cognitive style gap mean in Class A was 26.60 (SD=18.00), indicating that students on average were 26.60 points more innovative than the faculty member. In Class A, the most adaptive student had a 13 point lower cognitive style score than the faculty memberÂ’s cognitive style score. The most innovative student in the class scored 61 points higher th an the faculty member for the total KAI score. Note that stress become s apparent to individua ls with a cognitive style gap at 20 points or hi gher (Kirton, 2003). In Class A, 38 (65.5%) of the 58 student respondents had a cognitive styl e gap at or above 20 points. See Table 4-45 for findings specific to cognitive style gap of Class A. Once cognitive style gaps were calculat ed, PearsonÂ’s correlation coefficient was used to find relationships with total scores and constructs scores of stress, motivation, engagement and demographics.

PAGE 167

149 Table 4-45. Class A Student Mean Scores of Cognitive Style Construct Gaps (n=58) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap 26.60 10.62 6.90 9.09 18.00 8.40 4.72 8.45 -13 -7 -2 -8 61 29 18 26 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Examining internal correlations for th e KAI, very high correlations were found between the total cognitive st yle gap score and the sufficiency of originality cognitive style gap (r=.87, p<.05) as well as the rule/group conformity gap (r=.93, p<.05). Sufficiency of originality gap was very highly correlated with rule/group conformity gap (r=.70, p<.05). However, efficien cy cognitive style gap had a substantial correlation with total cognitive style gap (r=.66, p<.05), but only a moderate correlation with sufficiency of originality (r=.36, p<.05). This data suggests that in Class A, sufficiency of originality gap and rule/group conformity gap are closely associated, while the efficiency cognitive style gap is less associated w ith total cognitive style gap. No te that direction and degree of construct gap correlations were dependent on cognitive style scores of faculty member in relation to studentsÂ’ cognitive style mean scores. Another explanation to the lower efficiency gap correlations may include the lower post-hoc reliability coefficient for the efficiency scale. Total cognitive style gap had a significant moderate relationship with total stress (r=.34, p<.05), indicating an association w ith an innovative cognitive style gap and higher total stress scores. Al so note that efficiency cogni tive style gap had moderate correlations with stress constructs: frustra tions (r=.37, p<.05), c onflicts (r=.35, p<.05), and pressure (r=.38, p<.05). Furt hermore, rules group/conform ity cognitive style gap had

PAGE 168

150 moderate correlations with stress constructs : frustrations (r=.34, p<.05), and changes (r=.33, p<.05). These findings indicate a re lationship between higher stress construct scores and both constructs, efficiency c ognitive style gap and rule/group conformity cognitive style gap. Total cognitive style gap had a significant low relationship with total motivation (r=-.29, p<.05). This indicated that in this course taught by an adaptive faculty member, studentsÂ’ innovative total cognitive style ga p was associated with studentsÂ’ lower motivation scores. Also of interest, efficien cy cognitive style gap scores had moderate correlations with total motivation (r=-.33, p< .05), extrinsic motivation (r=-.36, p<.05) and self-efficacy (r=-.44, p<.05) indicating that a higher innovative gap was associated with lower stress among these constructs. Consid ering the rule/group conformity cognitive style gap, moderate correlations were found w ith total motivation (r=.31, p<.05) and task motivation (r=-.33, p<.05) signi fying that higher innovative rule/group conformity gaps were associated with lower motivation levels among these constructs in Class A. Total cognitive style gap had a negligib le correlation with total engagement (r=.04, p>.05) signifying little a ssociation between the two vari ables. No other significant correlations were found between constructs of cognitive style gap and constructs of student engagement. For internal correlations of the SSI, student total stress significantly correlated with stress constructs: frust rations (r=.83, p<.05), conflic ts (r=.59, p<.05), pressures (r=.73, p<.05), changes (r=.48, p<.05) and se lf-imposed (r=.60, p<.05). As expected, these correlations indicated th at the scales used to meas ure stress in this study were moderately to highly related with the total stress measured by the SSI.

PAGE 169

151 Total stress scores of students were negatively correlated wi th self-efficacy (r=.37, p<.05) indicating lower self-efficacy was associ ated with higher stress in this class. Self-efficacy was also negatively correlated with stress constructs frustrations (r=-.40), pressures (r=-.41, p<.05) and changes (r=-.31, p<.05). These correlati ons also indicated that low self-efficacy was associated with higher stress in this class. Examining student stress and student engagement, a moderate correlation was found between total stress and total student engagement (r=.37, p<.05). Total stress was also correlated with the engagement cons tructs: academic cha llenge (r=.30, p<.05) and student-faculty interact ion (r=.35, p<.05). The stress constr uct frustrations moderately correlated with total student engagement (r=.38, p<.05), academic challenge (r=.36, p<.05) and active learning (r=.32, p<.05). All of these correlations provide evidence that associations exist between stude ntsÂ’ higher stress scores and studentsÂ’ higher engagement scores. For Class A, the MSLQ was used to meas ure motivation of students. Moderate to very high correlations were found between the total motivation score and construct scores: intrinsic motivation (r=.73, p<.05), extrinsic motivation (r=.62, p<.05), task motivation (r=.77, p<.05), control of lear ning (r=.31, p<.05) and self-efficacy (r=.33, p<.05). However, the correlation between total motivation scores and test anxiety scores was not significant, but did have a low correlation (r=.24, p>. 05). The data suggests that constructs used to measure motivation were related to the total motivation score, less test anxiety. Examining correlations between motivation and engagement in Class A, no significant relationship was found with tota l motivation and total engagement (r=.24,

PAGE 170

152 p>.05). However extrinsic motivation was mode rately correlated with total engagement (r=.32, p<.05) and student-facu lty interaction (r=.43, p<.05). Also task motivation was moderately correlated with total engage ment (r=.31, p<.05) and academic challenge (r=.33, p<.05). These correlations indicated that higher scores in st udent motivation were associated with higher scores in st udent engagement among the aforementioned constructs. However, a correlation was found in the opposite direction when considering motivation constructs contro l of learning and self-efficac y. Control of learning was negatively correlated with total student engagement (r=-.32, p<.05) and academic challenge (r=-.30, p<.05). These two correlations indicated that higher levels of student control of learning were associated with lo wer levels of student engagement and lower levels of academic challenge. Self-efficacy wa s also negatively correlated with active learning (r=-.32, p<.05) which indicated that lo wer levels of student self-efficacy were associated with higher levels of active learning. Considering the internal correlations for the NSSE found in this class, very high correlations existed between total student engagement and the constructs: academic challenge (r=.79, p<.05), active learning (r=.85, p<.05) and student-faculty interaction (r=.76, p<.05). These correlations signify a cl ose association between these constructs. See Table 4-46 for relationships between c ognitive style gap, stress, motivation and engagement.

PAGE 171

153Table 4-46. Class A Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=56) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .87* -3. Gap-efficiency .66* .32* -4. Gap-rule .93* .70* .54* -5. Total stress .34* .24 .37* .29* -6. Frustrations .34* .20 .37* .34* .83* -7. Conflicts .21 .14 .35* .13 .59* .47* -8. Pressures .29* .14 .38* .28* .73* .57* .19 -9. Changes .35* .28* .28* .33* .48* .30* .20 .36* -10. Self-imposed .01 .10 -.09 -.03 .60* .28* .18 .24 .05 -11. Total motivation -.29*-.13 -.33*-.31*-.06 -.06 -.18 -.16 -.27* .30* -12. Intrinsic motivation -.13 .01 -.12 -.22 -.07 -.12 -.08 -.23 -.01 .18 .73* -13. Extrinsic motivation -.31*-.20 -.36*-.26 .02 .02 -.04 .01 -.10 .10 .62* .22 -14. Task motivation -.28*-.19 -.15 -.33*-.06 .06 -.08 -.12 -.21 .04 .77* .69* .37* 15. Control of learning .01 .06 -.11 .02 -.17 -.19 -.03 -.24 -.39* .15 .31* .07 .12 16. Self-efficacy -.13 .01 -.44*-.05 -.37*-.40*-.29* -.41*-.31* .11 .33* .06 .26 17. Test anxiety -.09 -.05 -.01 -.14 .27* .22 -.11 .27* .02 .34* .45* .24 .11 18. Total Engagement .04 -.02 .12 .04 .37* .38* .19 .29* .18 .16 .24 .27* .32* 19. Academic Challenge .16 .14 .17 .12 .30* .36* .16 .29* .17 -.01 .20 .28* .12 20. Active Learning -.02 -.12 .07 .04 .25 .32* .10 .17 .02 .13 .13 .17 .23 21. Faculty Interaction -.06 -.09 .03 -.06 .35* .21 .20 .23 .23 .28* .25 .17 .43* M 26.8 10.8 6.77 9.25 53.0 12.1 5.95 11.8 4.77 18.4 154 15.9 21.2 SD 18.0 8.32 4.70 8.32 10.4 3.82 2.64 3.34 1.92 3.80 17.4 4.26 4.10 Note. Cases excluded listwise. All constr ucts coded: high scores equals increased levels. signifies p<.05

PAGE 172

154 Table 4-46 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning -.05 -16. Self-efficacy .11 .44* -17. Test anxiety .24 -.21 -.36*-18. Total Engagement .31* -.32*-.29* .27* -19. Academic Challenge .33* -.30*-.24 .24 .79* -20. Active Learning .20 -.24 -.32* .19 .85* .51* -21. Faculty Interaction .20 -.22 -.13 .21 .76* .32* .54* -M 27.2 23.3 48.6 17.9 49.0 25.2 12.0 11.9 SD 7.01 3.57 6.23 6.81 7.37 3.40 2.92 2.93 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Demographic questions were asked of st udents regarding age, gender, college classification and number of classes taken simila r to the content of Class A. The variable student age was found positively associated w ith total cognitive style gap (r=.31, p<.05) indicating that younger students were associated with sma ller cognitive style gaps. Student gender was found moderately rela ted with total cognitive style gap (r=-.32, p<.05), indicating that being female was related to lower total cognitive style gap with the faculty member. Considering the number of colle giate courses taken related to the subject area of Class A, a correlation was found with total engagement (r=.39, p<.05), indicating that more courses relevant to the subject matter of Class A was coupled with higher

PAGE 173

155 engagement. Also, college classification had a moderate association (r=.30, p<.05) with total student engagement, indicating that a hi gher class rank was a ssociated with higher engagement. Class B For cognitive style gap of Class B, there were 46 usable responses. The instructing faculty member of Class B scor ed 68 on the KAI providing evidence that his cognitive style was more adaptive. Cognitive style gap was calculated by subtracting the faculty member cognitive style score from the student cognitive style score. The reported mean of total cognitive style gap between th e faculty member and the students was 24.43 (SD=15.05) indicating that stud ents on average were 24.43 points more innovative than the faculty member. A cognitive style gap of 20 points has been identified as the threshold at which stress becomes apparent to involved individuals (Kirton, 2003). There were 29 (63.0%) students with a cognitive style gap of 20 points or higher with the faculty member in Class B. The most adaptiv e student scored 22 points less than this adaptive faculty member in determining total cognitive style. The most innovative student in the class had a 48 point gap in th e direction of innovation with the faculty member. See Table 4-47 for findings of calculated cognitive style gap scores. Table 4-47. Class B Student Mean Scores of Cognitive Style Construct Gaps (n=46) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap 24.43 15.39 -1.59 10.63 15.05 8.45 4.65 6.68 -22 -3 -12 -7 48 34 9 23 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative.

PAGE 174

156 After calculating cognitive style gaps be tween the faculty me mber and individual students, PearsonÂ’s correlation coefficient was used to determine relationships among total and construct scores of st ress, motivation and engagement. Internal correlations of the KAI found that total cognitive style gap had very high correlations with sufficiency of originalit y gap (r=.86, p<.05), and rule/group conformity gap (r=.86, p<.05). A substantial correlation wa s found between sufficiency of originality gap and rule/group conformity gap (r=.63, p<.05 ) providing evidence that in Class B, a strong relationship existed between these two construct gaps. However, efficiency gap only had a moderate correlation with total c ognitive style gap (r=.43, p<.05) a negligible correlation with sufficiency of originalit y (r=.05, p>.05) and ru le/group conformity (r=.20, p>.05). This indicated that efficiency gap was not as closely associated with the two other cognitive style constructs which coincides with findings from cognitive style gap construct mean scores. For example, the mean score for efficiency gap was -1.59 whereas sufficiency of originality gap mean and the rule/group conformity gap mean was 15.39 and 10.63 respectively. Also note the lowe r post-hoc reliability of the efficiency construct. For Class B, there was a negligible correlation between total cognitive style gap and total student stress (r=-.01, p>.05). Add itionally, no significant correlations were found between cognitive style gap co nstructs and stress constructs. Interestingly, the cognitive style gap construct efficiency had a moderate correlation with total motivation (r=-.46, p<.05), intrinsic motivation (r=-.30, p<.05), extrinsic motivation (r=-.36, p< .05), control of learning (r=.35, p<.05) and self-efficacy

PAGE 175

157 (r=-.44, p<.05). These correlations indicated that among students in Class B, a larger efficiency gap was associated with lower levels of motivation in these constructs. Total cognitive style gap was moderately correlated with total student engagement (r=.31, p<.05) and active learning (r=.44, p<.05), indicating that more innovativeness was related to higher levels of student engageme nt and active learning. Likewise sufficiency of originality gap was positively correlated with total engagement (r=.33, p<.05) and active learning (r=.33, p<.05). Also rule/group c onformity gap was correlated with total engagement (r=.36, p<.05), active learning (r=. 48, p<.05) and student -faculty interaction (r=.32, p<.05); all indicating higher levels of engagement with st udent innovativeness. The SSI had internal correlations rangi ng from substantial to very high between total stress and stress constructs: frus trations (r=.76, p<.05), conflicts (r=.50, p<.05), pressures (r=.78, p<.05), changes (r=.69, p<.05) and self imposed (r=.65, p<.05). As expected, these correlations indicate that thes e scales of stress were closely associated. Considering the relationship between st ress and motivation in Class B, selfimposed stress was moderately correlated with total motivation (r=.34, p<.05) and extrinsic motivation (r=.40, p<.05), but substant ially correlated with test anxiety (r=.60, p<.05). These correlations provide evidence that higher levels of self-imposed stress were coupled with higher levels of total motiva tion, extrinsic motivation and test anxiety. Furthermore, test anxiety had moderate to s ubstantial correlations with total stress (r=.56, p<.05), frustrations (r=.34, p<.05), pressure s (r=.46, p<.05) and changes (r=.31, p<.05). These correlations provide evidence of a re lationship between high student stress scores and high test anxiety scores among students enrolled in Class B.

PAGE 176

158 There were significant negative correlati ons found between self-efficacy and total stress (r=-.30, p<.05), frustrations (r=.31, p<.05) and changes (r=-.44, p<.05). This indicated that lower levels of self-efficacy we re related to higher levels of total stress, frustrations and changes. Total motivation measured by the MSLQ had substantial and high correlations with internal constructs. Specifically, tota l motivation was correlated with intrinsic motivation (r=.74, p<.05), extrinsic motiva tion (r=.56, p<.05), ta sk motivation (r=.72, p<.05), control of learning (r=.71, p<.05) and self-efficacy (r=.62, p<.05). These correlations indicated a close relationship between motivation constructs and total motivation. However, the motivation construct test anxiety did not have a significant correlation with total motivation (r=.24, p>.05) i ndicating less of an association with the other motivation constructs. Total motivation was moderately correla ted with total engagement (r=.37, p<.05) as well as academic challenge (r=.43, p< .05). Likewise, intrinsic motivation was moderately correlated with total student engagement (r=.40, p<.05) and academic challenge (r=.35, p<.05) as well as stude nt-faculty interaction (r=.32, p<.05). Additionally, task motivation was moderately correlated with total student engagement (r=.46, p<.05), academic challenge (r=.41, p<.05) and student-faculty interaction (r=.32, p<.05). All of these correlations provide evidence that in Class B, higher levels of student motivation were associated with high er levels of student engagement. Examining the internal correlations of constructs measuring student engagement, very high correlations were found between to tal student engagement and constructs: academic challenge (r=.76, p<.05), active l earning (r=.74, p<.05) and student faculty

PAGE 177

159 interaction (r=.77, p<.05). Th ese correlations indicated a close relationship between scales of student engagement and total stud ent engagement. See Table 4-48 for findings specific to relationships between cognitive st yle gap, stress, motivation and engagement. Demographic questions were asked of st udents regarding age, gender, number of subject area courses and colle ge classification. For Class B, studentsÂ’ age had a positive association with efficiency style gap (r=.30, p<.05) indicatin g that older students were associated with larger efficiency style gap. No other significant co rrelations were found among these demographic variables and cognitive style gap. Age was also correlated with intrinsic motivation (r=-.32, p< .05) indicating that older stud ents were associated with lower intrinsic motivation scores. However, neither gender or colle ge classification was associated with study variables at a moderate level or higher. Considering the number of collegiate courses related to the subject area of Class B, no significant correlations were found.

PAGE 178

160Table 4-48. Class B Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=44) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .86* -3. Gap-efficiency .43* .05 -4. Gap-rule .86* .63* .20 -5. Total stress -.01 .02 .08 -.10 -6. Frustrations .11 .14 .15 -.05 .76* -7. Conflicts .20 .17 .16 .12 .50* .11 -8. Pressures -.18 -.12 -.09 -.19 .78* .38* .36* -9. Changes -.05 -.11 .19 -.09 .69* .41* .43* .56* -10. Self-imposed -.11 -.05 -.10 -.11 .65* .41* .12 .41* .18 -11. Total motivation -.15 .03 -.46*-.05 .06 -.04 -.01 .09 -.24 .34* -12. Intrinsic motivation .03 .13 -.30* .11 -.04 -.04 -.03 .05 -.18 .01 .74* -13. Extrinsic motivation -.19 -.02 -.36*-.16 .17 .13 .03 .03 -.06 .40* .56* .06 -14. Task motivation .10 .11 -.13 .18 -.05 .01 -.10 .01 -.23 .04 .72* .80* .11 15. Control of learning .03 .13 -.35* .15 -.24 -.35* .01 -.13 -.36* .04 .71* .43* .28 16. Self-efficacy -.13 .03 -.44*-.03 -.30*-.31*-.10 -.22 -.44* .04 .62* .42* .17 17. Test anxiety -.31*-.22 -.10 -.35* .56* .34* .18 .46* .31* .60* .24 -.04 .40* 18. Total Engagement .31* .33* -.10 .36* .02 .02 .03 .11 -.17 .03 .37* .40* .07 19. Academic Challenge .11 .22 -.16 .08 .14 .27 -.15 .09 -.12 .19 .43* .35* .33* 20. Active Learning .44* .33* .13 .48* .06 -.05 .19 .18 -.09 -.01 .03 .20 -.26 21. Faculty Interaction .21 .20 -.14 .32* -.16 -.24 .10 -.01 -.17 -.14 .29* .32* -.03 M 23.8 15.2 -1.6410.3 58.1 14.0 6.50 11.4 5.82 20.4 156 17.7 21.8 SD 15.1 8.56 4.74 6.61 13.0 5.30 2.86 4.04 2.73 3.79 25.8 5.37 4.79 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 179

161 Table 4-48 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .47* -16. Self-efficacy .34* .76* -17. Test anxiety -.10 -.27 -.35*-18. Total Engagement .46* .23 .06 .07 -19. Academic Challenge .41* .12 -.06 .32* .74* -20. Active Learning .26 .02 -.11 -.04 .74* .27 -21. Faculty Interaction .32* .36* .29 -.18 .77* .25 .51* -M 29.3 22.7 43.0 21.6 47.1 22.4 12.3 12.5 SD 8.69 5.44 10.9 8.71 7.01 3.70 2.56 3.14 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class C For determining cognitive style gap, 56 responses were deemed acceptable by the researcher. The total cognitive style score was 83 for the faculty member instructing Class C indicating that she was more adaptiv e. The reported total cognitive style gap mean of this class was M=17.86 (SD=14.34). To in terpret, the average student in Class C was 17.86 points more innovative than the faculty member. It is also noted that the mean is just less than 20 points (Kirton, 2003) at which “clearly noticeable differences” (p. 67) may cause psychological stress. There were 26 (46.4%) students with a cognitive style gap of 20 points or higher in Class C. The mo st adaptive student in the class had an 18

PAGE 180

162 point gap with the faculty member while th e most innovative student had a 59 point gap with the faculty member. See Table 4-49 for reported results of cognitive style gap of Class C. Table 4-49. Class C Student Mean Scores of Cognitive Style Construct Gaps (n=56) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap 17.86 4.59 7.61 5.66 14.34 6.59 4.28 7.50 -18 -8 -3 -9 59 22 17 27 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Cognitive style gap calculations allowed the researcher to look for relationships with student stress, motivation and student engagement, using PearsonÂ’s correlation coefficient. Examining total cognitive styl e gap correlations with cognitive style gap constructs found very high and moderate relati onships. Specifically, total cognitive style gap was correlated with sufficiency of originality gap (r=.86, p<.05), efficiency gap (r=.40, p<.05) and rule/group conformity ga p (r=.93, p<.05). Furthermore sufficiency of originality gap was highly correlated with rule/group conformity gap (r=.76, p<.05), but was negligibly correlated w ith efficiency gap (r=-.01, p> .05). These findings indicated that although constructs of cognitive style were associated with total cognitive style, the construct gap efficiency had low or negligible association with the other two constructs. The may partially be explained by the lowe r post-hoc reliability of the efficiency construct. Considering correlations between cognitive style gap and student stress, significant correlations were found only with th e self-imposed stress construct. That is, self-imposed stress had negative moderate corr elations with total c ognitive style gap (r=-

PAGE 181

163 .31, p<.05) and rule/group conformity gap (r= -.31, p<.05). This indicated that lower levels of self-imposed stress were associat ed with more innova tive total cognitive gap and innovative rule/group conformity gap. This finding does not coincide with KirtonÂ’s (2003) adaption-innovation theory as applied in this context. Moderate correlations were found between total motivation and total cognitive style gap (r=-.34, p<.05) and efficiency gap (r=-.30, p< .05) indicating that among students in Class C, larger c ognitive style gaps and efficiency gaps were related to lower total motivation. The same wa s found with extrinsic motivation when comparing this construct to total cognitive style gap (r=-.34, p<.05) and rule/group conformity gap (r=.31, p<.05). Test anxiety was moderately correla ted with total cognitive style gap (r=-.49, p<.05), sufficiency of originality gap (r=-.47, p<.05) and rule/group conformity gap (r=.41, p<.05) signifying an associ ation between low test anxi ety and high innovative gap within these cognitive style constructs. Only one moderate correlation was found be tween cognitive style gap scores and student engagement scores. Sufficiency of orig inality was correlated with active learning (r=.31, p<.05) which provided evidence that having a more innovative sufficiency of originality gap with the faculty member was associated with having a higher level of active learning in Class C. For scales of student stress, correlatio ns were moderate and very high between total stress and constructs of total stress. Specifically in Class C, total stress was correlated with frustrations (r=.81, p<.05), conflicts (r=.70, p<.05), pressures (r=.81, p<.05), changes (r=.54, p<.05) and self-imposed (r=.71, p<.05). These findings indicated that scales of stress were closely associated with total stress.

PAGE 182

164 Total stress was correlated with extr insic motivation (r=.39, p<.05) and test anxiety (r=.52, p<.05) indicating higher levels of student stress was coupled with higher levels of extrinsic motivation and higher levels of test anxiety. Furthermore, test anxiety was moderately or substan tially correlated with frustra tions (r=.41, p<.05), pressures (r=.40, p<.05) and self-imposed stress (r=.54, p<.05). Correlation coefficients in these relationships indicated that hi gher levels of test anxiety were associated with higher levels of these stress constructs. Examining relationships between student stress and student engagement in Class C, a moderate correlation was found between conflicts and student faculty interaction (r=.35, p<.05). This finding indicated that higher levels of conflict stress were related to higher levels of student-facu lty interaction. Furthermore, self-imposed stress was correlated with total st udent engagement (r=.31, p<.05) indi cating higher levels of selfimposed stress were associated with higher levels of st udent engagement. This is interesting in the light of the negative re lationship found between lower self-imposed stress and higher cognitive style gap. Total motivation was substantially to ve ry highly correlated with motivation constructs: intrinsic motivation (r=.62, p<.05) extrinsic motivati on (r=.66, p<.05), task motivation (r=.79, p<.05), control of lear ning (r=.47, p<.05) and test anxiety (r=.57, p<.05). However self-efficacy was not signifi cantly correlated with total motivation. Findings indicated that motivation constructs were related to tota l motivation, except for self-efficacy. Examining the relationships between mo tivation and engagement among students in Class C, total motivation was moderately correlated with academic challenge (r=.43,

PAGE 183

165 p<.05) and substantially correlated with academic challenge (r=.50, p<.05). Likewise, task motivation was moderately correlated with total engagement (r=.41, p<.05) and substantially correlated with academic challenge (r=.56, p<.05). Extrinsic motivation was moderately correlated with total engagement (r=.34, p<.05), active learning (r=.30, p<.05) and student faculty interaction (r=.30, p<.0 5). All of these corr elations indicated a higher level of motivation associated with a higher level of student engagement. For correlations regarded as internal to the NSSE, total student engagement was very highly correlated with constructs: academic challenge (r=.77, p<.05), active learning (r=.80, p<.05) and student-facult y interaction (r=.76, p<.05). Thes e correlations indicated that constructs of engagement were closel y related to total engagement. See Table 4-50 for relationships among cognitive style ga p, stress, motivation and engagement of respondents in Class C. Demographic questions were asked of st udents regarding age, gender, number of related courses taken and co llege classification. For ge nder, a moderate negative correlation was found with total cognitive styl e gap (r=-.38, p<.05). That is, being male was associated with a larger innovative cogni tive style gap in Cla ss C. Student college classification of the student had significantly correlated with total cognitive style gap (r=.42, p<.05). This indicated that higher college rank was associated with higher cognitive style gap. No other demographic variables had statis tically significant relationships with cognitive style gap in this course.

PAGE 184

166Table 4-50. Class C Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=56) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .86* -3. Gap-efficiency .40* -.01 -4. Gap-rule .93* .76* .20 -5. Total stress -.22 -.25 -.07 -.16 -6. Frustrations -.14 -.22 .01 -.08 .81* -7. Conflicts -.08 -.14 -.03 -.02 .70* .55* -8. Pressures -.13 -.19 -.01 -.07 .81* .58* .54* -9. Changes -.03 -.04 -.07 .01 .54* .24 .42* .41* -10. Self-imposed -.31*-.22 -.14 -.31* .71* .41* .23 .43* .21 -11. Total motivation -.34*-.24 -.30*-.28* .26* .14 -.03 .09 .06 .50* -12. Intrinsic motivation .02 .06 -.12 .06 -.06 -.15 -.14 -.16 .06 .14 .62* -13. Extrinsic motivation -.34*-.29*-.17 -.31* .39* .24 .14 .21 .17 .51* .66* .28* -14. Task motivation -.22 -.11 -.22 -.20 -.04 -.10 -.25 -.10 .03 .19 .79* .68* .34* 15. Control of learning .10 .18 -.17 .13 .04 .07 -.02 .02 -.17 .12 .47* .09 .16 16. Self-efficacy .08 .18 -.07 .04 -.27*-.22 -.27* -.36*-.20 -.01 .15 .27* .01 17. Test anxiety -.49*-.47*-.21 -.41* .52* .41* .24 .40* .15 .54* .57* -.05 .33* 18. Total Engagement -.05 .09 -.17 -.08 .22 .01 .16 .23 .14 .31* .43* .30* .34* 19. Academic Challenge -.14 -.01 -.23 -.13 .18 -.02 .04 .29* .11 .24 .50* .23 .21 20. Active Learning .11 .31* -.15 .02 .08 -.06 .01 .03 .01 .27* .25 .25 .30* 21. Faculty Interaction -.06 -.06 -.01 -.06 .27* .09 .35* .12 .20 .21 .21 .22 .30* M 17.9 4.59 7.61 5.66 55.4 13.1 6.00 10.7 4.82 20.8 162 19.3 21.8 SD 14.3 6.59 4.28 7.50 13.5 4.98 2.60 3.43 2.60 4.84 18.8 3.56 4.51 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 185

167 Table 4-50 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .19 -16. Self-efficacy -.02 .29* -17. Test anxiety .29* .13 -.37*-18. Total Engagement .41* -.01 .03 .26* -19. Academic Challenge .56* .16 -.17 .41* .77* -20. Active Learning .22 -.07 .21 .01 .80* .38* -21. Faculty Interaction .11 -.16 .09 .12 .76* .30* .56* -M 30.8 23.2 47.4 20.0 53.4 27.6 13.5 12.3 SD 8.35 3.55 5.32 7.88 8.63 4.35 3.42 3.35 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class D Cognitive style gap was measured by subtra cting faculty membe rÂ’s total cognitive style score and construct scores from studentsÂ’ total cognitiv e style and construct scores in each class. Calculating cognitive style ga p for Class D, the researcher found 73 usable KAI responses. Reported total cognitive st yle gap mean was approximately zero (M=0.37, SD=17.40) indicating that the average stude nt in Class D had lit tle cognitive style gap with the faculty member. This was expected as the total cognitive style score of the faculty member was 94, one point less than the general population mean (Kirton, 1999).

PAGE 186

168 All the same, this course was still worthy of examination as the cognitive style score range included one student who was 39 point s more adaptive and one student 41 points more innovative. Kirton (2003) claims indi viduals become aware of cognitive style differences given a 20-point cognitive style ga p which results in stress. In Class D, 24 (32.9%) student responses from the 73 usable responses had a cogniti ve style gap of 20 points or higher with the faculty member. See Table 4-51 for findings concerning Class D cognitive style gap. Table 4-51. Class D Student Mean Scores of Cognitive Style Construct Gaps (n=73) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -0.37 -6.51 2.06 4.07 17.40 8.37 5.37 7.98 -39 -25 -8 -11 41 7 17 24 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Cognitive style gap scores were used to find associations with student stress, motivation and engagement using PearsonÂ’ s correlation coefficient. Examining correlations between constructs of cognitive style, total cognitive style correlated with sufficiency of originality gap (r=.81, p<.05), efficiency gap (r=.61, p<.05), and rule/group conformity gap (r=.89, p<.05). These correlations indicated that construct measures of cognitive gap were closely associated with total cognitive style gap measured by the KAI. However, efficiency gap was not significantly correlated with sufficiency of originality (r=.16, p>.05) indicating little association. Th is finding may be explained by the degree and direction of cognitive style gaps of constr ucts; the sufficiency of originality cognitive style gap mean was -6.51 whereas the e fficiency gap mean was 2.06 in Class D. Furthermore, the reliability coefficient for the efficiency construct was below par.

PAGE 187

169 Only one moderate correlation was found among cognitive style gap scores and stress scores in Class D. A correlation wa s found between efficiency gap and selfimposed stress (r=-.36, p<.05) indicating an association betw een higher innovative efficiency gap and lower self-imposed stress. This correlation wa s not congruent with previous research conducted by Kirton (2003) as applied in the context of this study. In Class D, only one moderate correl ation was found among student scores of cognitive style and motivation. Specifically, e fficiency cognitive style gap was correlated with extrinsic motivation (r=-.38, p<.05) whic h provided evidence that larger innovative efficiency gap was related to lower leve ls of extrinsic motivation. There were no significant correlations among c ognitive style gap and student engagement in Class D. Considering the internal correlations of th e SSI, total student stress had very high correlations with frustrations (r=.84, p<.05) conflicts (r=.74, p<.05) pressures (r=.84, p<.05), changes (r=.74, p<.05) and self-impos ed (r=.76, p<.05). These correlations indicate that stress constructs were closel y associated with the total stress measure. In Class D, correlations were found among the stress construct self-imposed and the motivation constructs self-efficacy and test anxiety. That is self efficacy was correlated with total stress (r=-.40, p<.05), frustrations (r=-.50, p<.05), co nflicts (r=-.33, p<.05), pressures (r=-.34, p<.05) and changes (r= -.30, p<.05); all indica ting that increase in self-efficacy scores was asso ciated with a decrease of stre ss scores in these students. Test anxiety was moderately to substantially related to stress as this construct correlated with total stress (r=.64, p<.05), frustrati ons (r=.52, p<.05), conf licts (r=.46, p<.05), pressures (r=.59, p<.05), changes (r=.50, p<.05) and self-imposed stress (r=.44, p<.05). All of the correlations between the test anxiety construct and stress scores suggest an

PAGE 188

170 increase in stress was associated with an increase in test anxiety. Furthermore, selfimposed stress was moderately correlated with extrinsic motivation (r=.34, p<.05) indicating that higher levels of extrinsic motivation was coupled with high levels of selfimposed stress. No significant correlations were found among motivation scores and engagement scores for students in Class D. Moderate and very high correlations we re found between total motivation and motivation constructs: intrin sic motivation (r=.78, p<.05), extrinsic motivation (r=.66, p<.05), task motivation (r=.72, p<.05), contro l of learning (r=.77, p<.05) and self-efficacy (r=.70, p<.05). However, test a nxiety did not have a signific ant correlation with total motivation. Considering these correlations, the data suggests a close relationship exists between motivation constructs and the meas ure of total motivation, except for test anxiety. Examining relationships between mo tivation and engagement, moderate correlations were found between total motivation and total engagement (r=.38, p<.05) as well as the engagement construct academic challenge (r=.36, p<.05). These correlations describe an association of higher levels of motivation and higher levels of student engagement and higher levels of academic cha llenge. In Class D, m oderate correlations with task motivation also existed with tota l student engagement (r=.42, p<.05) as well as engagement constructs academic challe nge (r=.33, p<.05) and active learning (r=.38, p<.05). That is, high levels of task motivation in Class D were relate d to high levels of total engagement including the constructs academic challenge and active learning. This study used three constructs to determ ine a total student engagement score. In Class D, total student engagement was corr elated with these c onstructs: academic

PAGE 189

171 challenge (r=.74, p<.05), active learning (r=.74, p<.05) and student-faculty interaction (r=.67, p<.05). This data suggests that constr uct scores of student engagement were closely related to the total measure of student engagement used in this study. See Table 452 for reported associations between Class D cognitive style gap, stress, motivation and engagement. For demographic information concerning Cl ass D, participating students were asked their age, gender, number of cour ses taken related to Class D and college classification. The number of collegiate courses similar to the subject area of Class D was significantly correlated to tota l cognitive style gap (r=-.24, p< .05). Said differently, there was an association between more innovative cognitive style and fewer courses taken similar to the content taught in Class D. No other student demographic variable was statistically significant w ith cognitive style gap. For age, there were no significant correlati ons found with other variables used in this study in Class D. Concerning gender, th ere was a significant lo w correlation (r=-.24, p<.05) with test anxiety. That is in Class D, being male was associated with higher levels of test anxiety. College level of classification had a modera tely negative a ssociation (r=.32, p<.05) with total stress. This finding indi cated an association between higher college classification and lower levels of stress.

PAGE 190

172Table 4-52. Class D Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=67) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .81* -3. Gap-efficiency .61* .16 -4. Gap-rule .89* .56* .49* -5. Total stress -.09 -.03 -.21 -.02 -6. Frustrations .03 -.01 -.05 .10 .84* -7. Conflicts .15 .12 -.03 .22 .74* .69* -8. Pressures -.18 -.15 -.14 -.14 .84* .57* .44* -9. Changes -.08 -.07 -.21 .04 .74* .57* .42* .65* -10. Self-imposed -.21 -.01 -.36*-.19 .76* .42* .42* .61* .40* -11. Total motivation -.05 .11 -.22 -.09 .10 .03 .03 .10 -.03 .24* -12. Intrinsic motivation .07 .19 -.10 -.01 -.06 -.10 -.07 -.02 -.09 .01 .78* -13. Extrinsic motivation -.25*-.09 -.38*-.20 .19 .05 .04 .14 .14 .34* .66* .27* -14. Task motivation -.09 .04 -.13 -.15 -.12 -.12 -.15 -.11 -.19 .02 .72* .72* .24* 15. Control of learning .12 .20 .04 .01 -.12 -.18 -.04 -.11 -.17 .02 .77* .68* .34* 16. Self-efficacy .01 .20 -.18 -.10 -.40*-.50*-.33* -.34*-.30*-.10 .70* .61* .37* 17. Test anxiety .01 -.06 -.03 .09 .64* .52* .46* .59* .50* .44* .17 -.20 .18 18. Total Engagement .09 .20 -.10 .05 .09 -.03 .03 .13 .05 .17 .38* .30* .12 19. Academic Challenge .01 .09 -.12 -.01 .18 .14 .09 .19 .12 .14 .36* .27* .15 20. Active Learning .19 .20 .09 .13 -.08 -.20 -.15 .01 -.06 .06 .28* .29* -.01 21. Faculty Interaction .02 .15 -.18 -.01 .07 -.05 .12 .04 .01 .15 .15 .06 .08 M -0.57-6.571.97 4.03 54.2 12.2 5.40 11.1 4.61 20.9 166 19.1 22.5 SD 16.8 8.56 5.13 7.61 13.8 4.50 2.60 3.71 2.51 4.24 22.0 4.43 4.70 Note. Cases excluded listwise. All constr ucts coded: higher scores equals increased levels. signifies p<.05

PAGE 191

173 Table 4-52 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .55* -16. Self-efficacy .56* .69* -17. Test anxiety -.23 -.16 -.38*-18. Total Engagement .42* .26* .28* .05 -19. Academic Challenge .33* .19 .07 .26* .74* -20. Active Learning .38* .26* .40* -.18 .74* .26* -21. Faculty Interaction .17 .10 .17 -.05 .67* .17 .40* -M 33.7 24.2 46.6 20.3 48.3 24.1 12.0 12.2 SD 6.90 3.61 9.04 7.60 7.78 4.35 3.38 3.11 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class E There were 32 usable responses in determining cognitive style gap of Class E. The faculty member instructing Class E had a cognitive style score of 95 which was equivalent to the general population mean defined by Kirton (1999). The total cognitive style gap mean (M=-5.59, SD=14.95) was more than 14 points lower than the threshold of 20 points at which stress between dissimilar cognitive styles becomes apparent (Kirton, 2003). Only nine (28.1%) of the 32 student respondents had a cognitive style gap of 20 points or higher with the faculty me mber. This course was still worthy of examination as the lowest total cognitive style gap score was 29 points more adaptive

PAGE 192

174 than the faculty member and the highest total cognitive gap score was 21 points more innovative than the faculty member. See Table 4-53 for cognitive style gap findings for Class E. Table 4-53. Class E Student Mean Scores of Cognitive Style Construct Gaps (n=32) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -5.59 -2.00 -0.94 -2.66 14.95 7.87 3.65 7.30 -29 -16 -8 -16 21 15 8 10 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Once calculating the cognitive style gap of Class E, the researcher explored associations with stress, motivation and engagement using the Pearson correlation coefficient. First, cognitive style constructs were examined for relationships with total cognitive style gap. Very high correlations were found between total cognitive style gap and cognitive style constructs sufficiency of originality gap (r=.91, p<.05) and rule/group conformity gap (r=.94, p<.05). However, the corr elation between efficiency gap and total cognitive style gap was not significant (r=.24, p<.05). This finding may partially be explained by examining the differences of gap score means between students and the faculty member instructing Class E. It may al so be partially explained by the low internal reliability of the efficiency construct. There were no significant correlations found among scores of cognitive style gap and scores of student stress. However, c onsidering motivation, moderate correlations were found between sufficiency of originalit y gap and motivation constructs intrinsic motivation (r=.36, p<.05) and self-efficacy (r=. 38, p<.05). These correlations indicated that more innovative sufficiency of originalit y gap was associated w ith higher scores of

PAGE 193

175 motivation and higher scores of self-effi cacy. There were no significant correlations found between scores of cognitive style and sc ores of student engagement in Class E. Regarding the internal correlations of stude nt stress, total stress was substantially to very highly correlated with stress cons tructs: frustrations (r =.78, p<.05), conflicts (r=.58, p<.05), pressures (r=.85, p<.05), change s (r=.56, p<.05) and self-imposed (r=.83, p<.05). These correlations indicate that scales used to measur e stress in this study closely related to total stress scores. Total stress was not significantly corr elated with total motivation (r=.26, p<.05) but was moderately correlated with both extrinsic motivation (r=.48, p<.05) and test anxiety (r=.44, p<.05). This data suggests that higher levels of extrinsic motivation were coupled with high levels of extrinsic motivation and high levels of test anxiety in Class E. Likewise, the pressures constr uct was correlated with extr insic motivation (r=.43, p<.05) and test anxiety (r=.51, p<.05). Considering self -imposed stress, moderate and substantial correlations were found betw een total motivation (r=.44, p< .05), extrinsic motivation (r=.60, p<.05) and task motivation (r=.43, p<.05). All of these correla tions indicated that higher levels of stress were associat ed with higher levels of motivation. Only one significant correlation was found considering st ress and student engagement. The pressures stress construc t was correlated with active learning (r=.50, p<.05) which indicated that higher levels of st udent pressures were a ssociated with higher levels of student active learning. For correlations among scales of student motivation, to tal student motivation was correlated with constructs: intrinsic motivation (r=.70, p< .05), extrinsic motivation (r=.74, p<.05), task motivation (r=.86, p<.05), control of learning (r=.68, p<.05) and test

PAGE 194

176 anxiety (r=.32, p>.05). The correlation between total motivation and test anxiety was smaller, but still moderate. These correl ations provided evidence that motivation constructs were positively asso ciated with total motivation. Interestingly, only one correlation was found between scores of motivation and scores of student engagement. There was a moderate correlati on between extrinsic motivation and active learning (r=.40, p<.05) The finding indicated an association between higher levels of extrinsic motivation and higher levels of act ive learning in this class. Examining internal correlations for stude nt engagement, total student engagement was correlated with constructs: academic challenge (r=.83, p<.05), active learning (r=.59, p<.05) and student-faculty interaction (r=.58, p<.05). These correla tions suggest that student engagement construct sc ores were closely related to the measure of total student engagement. See Table 4-54 for findings specific to correlates of cognitive style gap, stress, motivation and engagement for Class E. The researcher examined correlations among demographic information as students answered questions regarding age, gender, number of courses taken similar to the course and college classification. None of the demographic variables were found to correlate with cognitive style gap in Class E. However, numbe r of college courses related to the subject area of Class E was significantly correlated with student-faculty interaction (r=.48, p<.05). This finding indicat ed that taking more classes in the subject area of Class E was moderately associated with higher levels of st udent-faculty interaction.

PAGE 195

177Table 4-54. Class E Correlations, Means, a nd Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=31) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .91* -3. Gap-efficiency .24 -.08 -4. Gap-rule .94* .82* .07 -5. Total stress -.02 -.03 .13 -.08 -6. Frustrations .08 .02 .16 .05 .78* -7. Conflicts .03 -.01 .10 .03 .58* .40* -8. Pressures -.24 -.20 -.01 -.28 .85* .52* .44* -9. Changes -.09 -.10 -.01 -.07 .56* .39* .69* .38* -10. Self-imposed .05 .06 .15 -.04 .83* .47* .21 .68* .25 -11. Total motivation .19 .35 -.15 .08 .26 .14 -.22 .21 .06 .44* -12. Intrinsic motivation .27 .36* -.11 .22 -.08 -.15 -.28 -.15 .01 .13 .70* -13. Extrinsic motivation .20 .30 .01 .06 .48* .29 .03 .43* .08 .60* .74* .29 -14. Task motivation .15 .26 .01 .02 .22 .15 -.29 .10 .10 .43* .86* .68* .56* 15. Control of learning -.07 .09 -.15 -.18 -.01 -.01 -.18 -.08 -.14 .16 .62* .44* .27 16. Self-efficacy .18 .38* -.33 .11 -.19 -.19 -.44* -.13 -.22 .01 .68* .58* .44* 17. Test anxiety .02 .02 -.09 .07 .44* .33 .20 .51* .30 .30 .32 -.16 .32 18. Total Engagement .14 .15 -.13 .19 .23 .28 .16 .18 .01 .13 .10 .05 .27 19. Academic Challenge .16 .12 -.07 .23 .18 .27 .22 .13 .03 .04 .10 .09 .19 20. Active Learning -.14 -.11 -.05 -.15 .30 .24 -.11 .50* -.10 .26 .20 -.11 .40* 21. Faculty Interaction .19 .24 -.14 .21 .01 .06 .14 -.17 .06 .01 -.08 .08 .02 M -4.94-1.87-0.84-2.2343.3 9.39 4.52 9.16 3.48 16.8 164 18.3 21.1 SD 14.7 7.97 3.67 6.99 12.2 3.95 2.22 3.30 1.12 5.30 21.9 4.59 4.37 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 196

178 Table 4-54 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .49* -16. Self-efficacy .43* .54* -17. Test anxiety .15 -.20 -.12 -18. Total Engagement .20 -.04 -.10 -.02 -19. Academic Challenge .17 .12 -.13 -.07 .83* -20. Active Learning .19 -.14 .08 .32 .59* .33 -21. Faculty Interaction .05 -.14 -.11 -.21 .58* .16 .04 -M 33.5 23.4 49.5 17.8 48.3 24.5 13.0 10.9 SD 7.15 4.16 6.83 6.56 5.75 3.66 2.04 2.66 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class F For cognitive style gap of Class F, 71 responses were considered usable. The faculty member instructing Class F had a total cognitive style score of 103, indicating she was slightly innovative. The total cognitive st yle gap mean was calculated by subtracting the faculty memberÂ’s cognitive style score from each studentÂ’s cognitive style score. The mean difference was -12.83 (SD=16.28) indi cating that the average student was 12.83 points more adaptive than the faculty member. The most adaptive student had a cognitive style gap of 48 points lower than the instru ctor while the most innovative student had a

PAGE 197

179 cognitive style gap of 29 points higher than the faculty member. Note that Kirton (2003) identifies a cognitive style gap of 20 points as the threshold for awareness of cognitive style differences, which may result in stress In Class F, 30 (42.3%) student respondents had a cognitive style gap of 20 points or highe r with the instructing faculty member. See Table 4-55 for findings specific to cognitive style gap scores of Class F. Table 4-55. Class F Student Mean Scores of Cognitive Style Construct Gaps (n=71) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -12.83 -11.80 9.15 -10.18 16.28 8.89 4.01 6.81 -48 -30 1 -26 29 9 17 8 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Calculated cognitive style gaps were used to find relationships with student stress, motivation and engagement. PearsonÂ’s correla tion coefficient was used for this task. Considering internal correlations for cogni tive style gap, total cognitive style gap was correlated with sufficiency of originality gap (r=.89, p<.05), efficiency gap (r=.52, p<.05) and rule/group conformity gap (r=.89, p<.05). Fu rthermore, sufficiency of originality gap was substantially correlated with rule s/group conformity gap (r=.67, p<.05). These correlations indicated that su fficiency of originality gap and rule/group conformity gap were closely related to tota l cognitive gap. Conversely, effi ciency gap did not have a significant correlation with sufficiency of origin ality gap. This data suggests that in Class F, cognitive style construct gaps were relate d to total cognitive style gap, but there was little association between efficiency gap and sufficiency of originality gap. This may partially be explained by examining the degr ee and direction of cognitive style gap mean scores for this class. For example in Class F, the mean score for sufficiency of originality

PAGE 198

180 gap was -11.80 whereas the mean score for efficiency gap was 9.15. Also note that the post-hoc reliability coefficient for the effici ency construct was lower than the acceptable level of .70 (Schmitt, 1996). No moderate correlations were found among scores of cognitive style gap and scores of stress, or scores of engagement However, one moderate correlation was found between sufficiency of originality gap a nd extrinsic motivation (r=-.33, p<.05). This indicated that a more innovative gap was related to less extr insic motivation among students in Class F. For correlations within total stress and st ress constructs subs tantial to very high correlations were found. Specifical ly, total stress was correlated with frustrations (r=.86, p<.05), conflicts (r=.68, p<.05), pressures (r =.81, p<.05), changes (r=.59, p<.05) and selfimposed (r=.65, p<.05). The correlations imply th at these stress constructs were closely related to the measure of total stress. Examining relationships between stress sc ores and motivation scores, significant correlations were found between self-efficacy and total stress (r=-.44, p<.05), frustrations (r=-.50, p<.05), pressures (r=-.46, p<.05) and ch anges (r=-.38). This provided evidence that lower scores of self-efficacy were asso ciated with higher sc ores of these stress constructs. Positive relationships were found between test anxiety and total stress (r=.42, p<.05), frustrations (r=.33, p<.05), pressu res (r=.44, p<.05) and self-imposed (r=.42, p<.05). These correlations indicated that highe r scores of test anxi ety were related to these higher stress constructs. No significant correlations were f ound with regard to stress scores and engagement scores.

PAGE 199

181 Concerning relationships among motiva tion scores, moderate to very high correlations existed within the data. Specifical ly, total motivation correlated with intrinsic motivation (r=.64, p<.05), extrinsic motiva tion (r=.52, p<.05), ta sk motivation (r=.82, p<.05), control of learning (r=.56, p<.05), sel f-efficacy (r=.30, p<.05) and test anxiety (r=.53, p<.05). These correlations indicated that all of the motivation constructs were related to the total measure of motivation. There were three moderate correlations between scales of motivation and scales of engagement. Academic challenge was co rrelated with both total motivation (r=.31, p<.05) and intrinsic motivation (r=.33, p<.05). Th is finding indicates that higher levels of academic challenge were associated with highe r levels of total motivation with emphases on intrinsic motivation. Active learning was nega tively correlated with control of learning (r=-.33, p<.05). Interestingly, this indicated lower leve ls of control of learning associated with higher levels of active learning. For correlations among constructs of stude nt engagement, total engagement was very highly correlated w ith academic challenge (r=.72, p<.05), active learning (r=.73, p<.05) and student-faculty interaction (r=.70, p<.05). This provided evidence that constructs of student engagement were clos ely related to total st udent engagement. See Table 4-56 for findings concerning relations hips among cognitive style gap, stress, motivation and engagement.

PAGE 200

182Table 4-56. Class F Correlations, Means, a nd Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=69) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .89* -3. Gap-efficiency .52* .18 -4. Gap-rule .89* .67* .39* -5. Total stress .06 .11 .05 -.03 -6. Frustrations .01 .01 .06 -.01 .86* -7. Conflicts .18 .25* .01 .10 .68* .54* -8. Pressures -.09 -.04 .02 -.17 .81* .64* .40* -9. Changes .16 .17 .06 .13 .59* .51* .41* .46* -10. Self-imposed .03 .09 .04 -.08 .65* .40* .27* .41* .01 -11. Total motivation -.12 -.12 -.04 -.10 -.04 .04 -.20 .13 -.25* .24* -12. Intrinsic motivation -.01 .10 -.21 -.03 -.01 -.07 .02 .05 -.12 .05 .64* -13. Extrinsic motivation -.29*-.33*-.03 -.23* .14 .29* -.23 .20 -.07 .15 .52* -.15 -14. Task motivation -.15 -.20 -.01 -.08 -.08 .01 -.26* .05 -.30* .05 .82* .55* .39* 15. Control of learning -.01 -.01 -.17 .10 -.17 -.22 -.08 -.15 -.36* .11 .56* .38* .06 16. Self-efficacy .17 .23 -.11 .17 -.44*-.50*-.22 -.46*-.38*-.09 .30* .30* -.12 17. Test anxiety -.06 -.10 .25* -.15 .42* .33* .01 .44* .17 .42* .53* .12 .44* 18. Total Engagement -.01 .03 -.06 -.01 .12 .16 .02 .07 -.01 .13 .09 .07 .10 19. Academic Challenge -.16 -.08 -.16 -.17 .18 .18 .02 .15 -.05 .23 .31* .33* .05 20. Active Learning .16 .12 .04 .19 .09 .20 -.03 .04 .17 -.04 -.15 -.19 .15 21. Faculty Interaction .04 .05 .03 .02 -.04 -.06 .04 -.07 -.13 .04 -.05 -.10 .03 M -13.2-12.09.07 -10.349.1 10.9 5.32 9.32 4.54 19.0 163 17.9 21.6 SD 15.6 8.75 3.95 6.55 12.6 4.01 2.70 3.60 2.53 4.40 18.5 4.63 4.27 Note. Cases excluded listwise. All constr ucts coded: higher scores equals increased levels. signifies p<.05

PAGE 201

183 Table 4-56 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .39* -16. Self-efficacy .19 .45* -17. Test anxiety .25* -.08 -.33*-18. Total Engagement .10 -.14 -.03 .13 -19. Academic Challenge .28* .10 .04 .19 .72* -20. Active Learning -.12 -.32*-.12 .02 .73* .23 -21. Faculty Interaction -.03 -.14 -.01 .05 .70* .16 .46* -M 34.6 23.2 47.8 17.6 47.8 25.6 11.5 10.8 SD 6.98 3.55 5.93 6.99 7.49 4.18 3.05 3.18 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 The researcher examined correlations among demographic information as students answered questions regarding age, ge nder, number of classes taken similar to the class and college classifica tion. No significant correlations were found in Class F regarding these demographic variab les and total cognitive style gap. However, age was associated with ex trinsic motivation, (r=-.30, p<.05). This finding indicated that older st udents were associated with less extrinsic motivation in Class F. Gender was moderately correlated to total student motivation (r=.42, p<.05). That is, females in this class were associ ated with higher levels of total motivation.

PAGE 202

184 Considering studentsÂ’ college classification, a moderate corr elations was found with test anxiety (r=-.35, p<.05). The findi ng indicated that higher leve ls of test anxiety were associated with lower college classification. Considering the number of collegiate courses related to the subject area of Class F, no significant correlations were found. Class G There were 65 responses considered us able by the researcher for determining cognitive style gap. The faculty member inst ructing Class G had a total KAI score of 116, indicating her cognitive style was more innovati ve. The mean cognitive style gap of Class G was -22.58 (SD=15.04). Said differently, th e average student in Class G was 22.58 points more adaptive than the faculty member. The lowest total score cognitive style gap was 67 points more adaptive than the faculty member while the highest total score cognitive style gap was 9 points more innovati ve than the faculty member. Note that Kirton (2003) has determined that a 20 point gap between two individu als is the point at which cognitive style differences become a pparent. In this cla ss, 39 (60.0%) student respondents had a cognitive style gap of 20 poi nts or more with the faculty member. See Table 4-57 for results regarding Class G cognitive style gap. Table 4-57. Class G Student Mean Scores of Cognitive Style Construct Gaps (n=65) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -22.58 -8.82 -2.69 -11.08 15.04 7.67 4.76 7.85 -64 -28 -12 -30 9 7 8 8 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Calculated cognitive gap scores were utilized in finding relationships with student stress, student motivation and student engage ment. PearsonÂ’s correlation coefficient was

PAGE 203

185 used to find correlations among total scores a nd construct scores of cognitive style gap, stress, motivation and engagement. In Class G, total cognitive style gap was internally corr elated with sufficiency of originality gap (r=.83, p<.05), efficiency gap (r=.44, p<.05) and rule/group conformity gap (r=.84, p<.05) indicating a close relationship among construct gap scores and the total measure of cognitive style gap. However, efficiency gap was not significantly correlated with either sufficiency of or iginality gap (r=.12, p> .05) or rule/group conformity gap (r=.13, p>.05). This finding can partially be explai ned by differences among construct gap mean scores found in Cl ass G. Another explanation may be the lower post-hoc reliability coefficient for th e cognitive style efficiency construct. Only one moderate correlation existed among constructs of cognitive style gap and constructs of student stress. Sufficiency of originality gap was correlated with selfimposed stress (r=.33, p<.05) indicating a more innovative sufficiency of originality gap was associated with higher scor es of self-imposed stress. No significant correlations were found be tween scores of cognitive style gap and scores of total motivation. However total cogni tive style gap was co rrelated with active learning (r=.30, p<.05) indicating a more i nnovative gap in this class taught by an innovative faculty member was associated with higher levels of active learning. Considering correlations internal to the measurement of stress, substantial and very high correlations were found between total stress and frustrations (r=.85, p<.05), conflicts (r=.68, p<.05), pressures (r=.85, p<.05), changes (r=.76, p<.05) and selfimposed (r=.71, p<.05). These correlations indicated that constructs of student stress were closely related to the measure of total stress.

PAGE 204

186 Examining correlations between stress and motivation in Class G, self-efficacy was negatively correlated with total stress (r=-.59, p<.05), frustr ations (r=-.63, p<.05), conflicts (r=-.37, p<.05), pressures (r=-.41, p<.05), changes (r=-.61, p<.05) and test anxiety (r=-.30, p<.05). These correlations indi cated that lower self-efficacy scores were associated with high stress scores among all co nstructs. On the othe r hand, test anxiety was moderately to substantia lly correlated with total stre ss (r=.55, p<.05), frustrations (r=.48, p<.05), pressures (r=.44, p<.05), change s (r=.47, p<.05) and self-imposed stress (r=.47, p<.05). The data suggests an association between high levels of test anxiety and high levels of stress among the aforementi oned constructs. Further explaining selfimposed stress, moderate correlations were found with total motivation (r=.38, p<.05) and extrinsic motivation (r=.36, p<.05). Thes e correlations indicat ed an association between higher scores of self-imposed stre ss and higher score of total motivation and extrinsic motivation. Total student stress in Class G was corre lated with total student engagement (r=.32, p<.05). Furthermore, tota l engagement was correlated with frustrations (r=.34, p<.05). Additionally, total stress was correla ted with academic challenge (r=.30, p<.05). These correlations indicated that higher levels of stress were associat ed with higher levels of engagement in the aforementioned constructs. For correlations internal to the measure of motiva tion, total motivation was correlated with intrinsic mo tivation (r=.58, p<.05), extrin sic motivation (r=.69, p<.05), task motivation (r=.67, p<.05), control of learning (r=.51, p<.05), self-efficacy (r=.29, p<.05), and test anxiety (r= .43, p<.05). These correlations provided evidence that

PAGE 205

187 constructs of motivation were related to the measure of total motivation. However, the measure of self-efficacy was less related with total motivation then expected. Considering the relationships among stude nt motivation and student engagement in Class G, total engagement was correlated with total motivation (r=.42, p<.05), intrinsic motivation (r=.42, p<.05), extrinsic motivati on (r=.43, p<.05) and task motivation (r=.36, p<.05). Likewise, the engagement construct academic challenge was correlated with total motivation (r=.43, p<.05), intrinsic motivati on (r=.35, p<.05), extrinsic motivation (r=.45, p<.05) and task motivation (r=.32, p<.05). Si milarly, active learni ng was significantly correlated with total motiv ation (r=.30, p<.05), intrin sic motivation (r=.37, p<.05), extrinsic motivation (r=.31, p<.05) and task motivation (r=.26, p<.05). These correlations indicated that higher levels of intrinsic motivation, extrinsic motivation and task motivation were associated with higher levels of academic challenge and active learning. Examining relationships among constructs of student engagement f ound that total student engagement was correlated with academic challenge (r=.85, p<.05), active learning (r=.71, p<.05) and student faculty interaction (r=.73, p<.05). Th ese very high correlations suggest a close relationship between constructs of student engagement and the measure of student engagement. See Table 4-58 for co rrelations among total cognitive style gap, stress, motivation and engagement. Student respondents were asked demogra phic questions including age, gender, number of classes taken similar to Class G and college classification. None of these demographic variables were significantly a ssociated with total cognitive style gap. Furthermore, no moderate correlations were found among demographic variables.

PAGE 206

188Table 4-58. Class G Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=64) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .83* -3. Gap-efficiency .44* .12 -4. Gap-rule .84* .54* .13 -5. Total stress .17 .18 .12 .08 -6. Frustrations .08 .04 .17 -.01 .85* -7. Conflicts .10 .03 .07 .11 .68* .64* -8. Pressures .08 .07 .07 .05 .85* .63* .49* -9. Changes .15 .13 .01 .16 .76* .67* .46* .66* -10. Self-imposed .22 .33* .12 .04 .71* .40* .23 .52* .31* -11. Total motivation .10 .22 -.15 .07 .17 .02 -.10 .22 -.04 .38* -12. Intrinsic motivation .11 .18 -.12 .11 .10 .01 .01 .11 .03 .18 .58* -13. Extrinsic motivation .04 .13 -.16 .05 .34* .22 .07 .39* .15 .36* .69* .23 -14. Task motivation .13 .11 .06 .12 .09 .01 -.11 .14 -.19 .29* .67* .45* .40* 15. Control of learning .06 .22 -.20 .02 -.25*-.34*-.29* -.21 -.23 .02 .51* .17 .15 16. Self-efficacy -.01 .01 -.11 .03 -.59*-.63*-.37* -.41*-.61*-.30* .29* .02 .02 17. Test anxiety .01 .08 -.01 -.06 .55* .48* .23 .44* .47* .47* .43* .04 .27* 18. Total Engagement .19 .14 .11 .16 .32* .34* .18 .27* .16 .23 .42* .42* .43* 19. Academic Challenge .05 .09 .04 -.02 .30* .36* .13 .24 .14 .22 .43* .35* .45* 20. Active Learning .30* .18 .20 .28* .25* .23 .20 .21 .18 .15 .30* .37* .31* 21. Faculty Interaction .11 .04 .03 .15 .18 .17 .11 .16 .03 .16 .20 .22 .20 M -22.6-8.78-2.73-11.151.3 11.8 5.97 10.7 4.95 17.8 160 19.0 20.5 SD 15.2 7.73 4.79 7.91 14.1 4.16 2.70 3.93 2.57 4.82 17.2 3.81 4.63 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 207

189 Table 4-58 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .02 -16. Self-efficacy .11 .52* -17. Test anxiety .10 .01 -.40*-18. Total Engagement .36* .01 -.08 .14 -19. Academic Challenge .32* .07 -.14 .22 .85* -20. Active Learning .26* -.01 .01 .02 .71* .39* -21. Faculty Interaction .24 -.06 -.03 .03 .73* .46* .27* -M 31.5 22.4 48.4 18.3 49.1 24.1 14.0 10.9 SD 7.07 3.71 7.07 6.81 7.42 3.87 2.99 2.78 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class H For cognitive style gap of Class H, ther e were 50 usable responses. The course was instructed by a faculty member with a total KAI score of 132, indicating a preference to solve problems innovatively. Cognitive styl e gap score was calcu lated by subtracting the faculty memberÂ’s cognitive style from each studentÂ’s cognitive style. The total cognitive style mean for Class H was -36.70 (SD=15.01). That is, the average student in Class H was 36.70 points more adaptive than th e faculty member instructing the course. Note that stress is noticeable between dissimilar cognitive styl es with at least a 20-point gap (Kirton, 2003). In this class, 45 ( 90.0%) of the 50 students with usable KAI

PAGE 208

190 responses had a cognitive style gap of 20 poi nts or more with the faculty member. The most adaptive cognitive style gap score wa s 65 points lower than the faculty member. The most innovative studentÂ’s cognitive style was identified as 3 points more innovative than the faculty member. See Table 4-59 for fi ndings of calculated cognitive style gap for Class H. Table 4-59. Class H Student Mean Scores of Cognitive Style Construct Gaps (n=50) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -36.70 -11.76 -10.36 -14.58 15.01 8.65 3.75 7.04 -65 -31 -19 -31 3 7 -2 1 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Once cognitive style gaps were calculat ed, PearsonÂ’s correlation coefficient was used to find relationships among total and c onstruct scores of stress, motivation and engagement. For correlations examini ng relationships between cognitive style gaps, total cognitive style gap was correlated with su fficiency of originality gap (r=.84, p<.05), efficiency gap (r=.47, p<.05) and rule/group conformity gap (r=.85, p<.05). These correlations indicated that su fficiency of originality gap and rule/group conformity gap were closely related to the to tal measure of cognitive style gap. However, efficiency gap was less associated with total cognitive style gap. The corre lation between sufficiency of originality gap and efficiency gap was not st atistically significant. The lower correlations associated with the efficiency gap may be attributed to differences of size among gap constructs. The efficiency cogniti ve style construct also suffer ed in reliability as the posthoc alpha coefficient was lower than an acceptable level of .70 (Schmitt, 1996).

PAGE 209

191 In Class H, moderate correlations were found between rule/group conformity gap and total stress (r=-.36, p<.05), frustrations (r=-.30, p<.05) and pressures (r=-.49, p<.05). These negative correlations indicated that a more innovative gap was associated with lower levels of stress in this class taught by an innovative faculty member. This finding was congruent with previous rese arch conducted by Kirton (2003). Examining relationships between cognitive style gap and motivation, moderate correlations were found between intrinsic motivation and to tal cognitive style gap (r=.32, p<.05), sufficiency of originality gap (r= .35, p<.05), and rule/group conformity gap (r=.30, p<.05). The data suggests that a more innovative gap in these two constructs was associated with higher levels of intrinsic motivation in Class H. Total cognitive style gap was correlate d with total student engagement (r=.40, p<.05), active learning (r=.46, p<.05) and stude nt-faculty interact ion (r=.41, p<.05). Likewise, sufficiency of orig inality gap was correlated wi th total engagement (r=.33, p<.05), active learning (r=.41, p<.05) and student -faculty interaction (r=.31, p<.05). Also rule/group conformity gap was moderately correlated with tota l engagement (r=.32, p<.05), active learning (r=.40, p<.05) and student -faculty interaction (r=.37, p<.05). This evidence suggests that more innovative sufficiency of originality gap and rule/group conformity gap was associated with higher levels of active lear ning and student-faculty interaction in Class H. For internal correlations of stress sc ales, total stress was correlated with frustrations (r=.64, p<.05), conflicts (r=.71, p<.05), pressures (r=.73, p<.05), changes (r=.67, p<.05) and self-imposed (r=.71, p<.05) These correlations signify a close relationship between stress constructs and the total measure of stress.

PAGE 210

192 In Class H, correlations were found betw een test anxiety an d total stress (r=.52, p<.05), frustrations (r=.31, p<.05), conflicts (r=.30, p<.05), pressu res (r=.36, p<.05) and self-imposed (r=.49, p<.05). The data suggests that higher levels of these stress constructs were related to test anxiety. Negative corre lations were found betw een self-efficacy and total stress (r=-.33, p<.05), frus trations (r=-.46, p<.05) and co nflicts (r=-.31, p<.05). This provided evidence that higher levels of self-efficacy were associated with lower levels of stress constructs frustrations and conflic ts. No significant correlations were found between constructs of stress a nd constructs of engagement. Total motivation in Class H was correl ated with intrinsic motivation (r=.72, p<.05), extrinsic motivation (r=.65, p<.05), task motivation (r=.81, p<.05), control of learning (r=.63, p<.05) and self-efficacy (r=.62, p<.05). These correlations indicated that constructs of motivation were related to the measure of total motivation. However, note that the correlation between total motivation and test anxiety was not significant (r=.22, p>.05) indicating a low association. Considering the relationship between total motivation and total student engagement, a moderate correlation (r=.44, p< .05) indicated higher le vels of motivation associated with higher levels of student engagement. This finding was supported by correlations found between constructs of motivation and engagement. Specifically, moderate and substantial correlations existed between intrinsic motivation and total engagement (r=.50, p<.05), active learning (r=. 49, p<.05) and student -faculty interaction (r=.41, p<.05). Likewise, task motivation was moderately correlated with total student engagement (r=.41, p<.05), academic challenge (r=.31, p<.05) and student-faculty

PAGE 211

193 interaction (r=.32, p<.05). Again, these correl ations provided eviden ce that higher levels of motivation were associated with higher levels of engagement in this class. Considering the internal correlations to scales of student engagement, total engagement was very highly correlated with academic challenge (r=.75, p<.05), active learning (r=.86, p<.05) and student-faculty inte raction (r=.79, p<.05). These correlations provided evidence that these constructs are clos ely related to the tota l measure of student engagement. See Table 4-60 for associations of cognitive style gap, student stress, student motivation, and student engagement for Class H. Student respondents were asked demographi c questions specific to age, gender, number of courses taken similar to the c ontent of Class H and college classification. Answers to these questions were used to find correlations with cogni tive style gap. None of these demographic variables were found to significantly correlate with total cognitive style gap in Class H.

PAGE 212

194Table 4-60. Class H Correlations, Means, and Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=50) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .84* -3. Gap-efficiency .47* .11 -4. Gap-rule .85* .51* .34* -5. Total stress -.23 -.14 .06 -.36*-6. Frustrations -.21 -.14 .05 -.30* .64* -7. Conflicts -.08 -.08 .08 -.13 .71* .44* -8. Pressures -.32*-.15 -.01 -.49* .73* .49* .43* -9. Changes -.03 -.03 .05 -.05 .67* .33* .61* .41* -10. Self-imposed -.12 -.07 .04 -.20 .71* .11 .29* .28* .33* -11. Total motivation .14 .22 -.04 .05 .01 -.16 -.05 -.03 -.21 .24 -12. Intrinsic motivation .32* .35* -.09 .30* -.17 -.21 .03 -.29*-.18 -.01 .72* -13. Extrinsic motivation -.03 .02 .14 -.16 .15 .01 -.10 .18 -.07 .28* .65* .23 -14. Task motivation .20 .27 -.09 .15 -.24 -.19 -.14 -.29*-.42*-.01 .81* .77* .36* 15. Control of learning .09 .12 .03 .03 -.04 -.20 -.09 -.01 -.05 .09 .63* .28* .26 16. Self-efficacy .19 .28* -.18 .16 -.33*-.46*-.31* -.15 -.21 -.11 .62* .36* .31* 17. Test anxiety -.20 -.18 .03 -.22 .52* .31* .30* .36* .13 .49* .22 -.02 .22 18. Total Engagement .40* .33* .23 .32* .10 -.03 .14 -.02 .08 .16 .44* .50* .25 19. Academic Challenge .08 .07 .19 -.01 .22 .05 .24 .10 .11 .22 .32* .29* .27 20. Active Learning .46* .41* .15 .40* .02 .08 .01 -.04 .05 -.01 .35* .49* .18 21. Faculty Interaction .41* .31* .22 .37* .02 -.19 .08 -.11 .04 .16 .37* .41* .16 M -36.7-11.8-10.4-14.649.3 11.4 5.34 9.56 4.10 19.0 158 19.0 21.6 SD 15.0 8.65 3.75 7.04 10.2 3.00 2.21 2.92 1.74 4.84 20.7 4.63 4.07 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 213

195 Table 4-60 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .50* -16. Self-efficacy .56* .68* -17. Test anxiety -.15 -.21 -.41*-18. Total Engagement .41* .32* .27 -.10 -19. Academic Challenge .31* .28* .19 -.10 .75* -20. Active Learning .35* .18 .20 -.08 .86* .50* -21. Faculty Interaction .32* .30* .25 -.05 .79* .32* .57* -M 33.1 22.5 43.5 18.7 51.2 26.3 13.2 11.7 SD 6.81 3.92 8.11 7.34 7.75 3.15 3.22 3.30 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Class I For Class I, there were 60 usable KAI responses for calculating cognitive gap. The faculty member instructing Class I had a total cognitive style score of 134 which was subtracted from each studentÂ’s cognitive style score. In Class I, th e total cognitive style gap mean was -41.10 (SD=14.25) which was obvious ly larger than the 20-point cognitive style gap Kirton (2003) claims as a threshol d for noticeable differences. Furthermore, 55 (91.7%) of the 60 students with usable responses had a cognitive styl e gap of 20 points or higher with this innovative faculty member The most adaptive student was 84 points lower than the faculty memberÂ’s cognitive st yle while the most innovative student in

PAGE 214

196 Class I was 10 points more innovative than the faculty member. See Table 4-61 for findings of calculated cognitive style gap reported in Class I. Table 4-61. Class I Student Mean Scores of Cognitive Style Construct Gaps (n=60) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -41.10 -18.87 -2.37 -19.87 14.25 7.61 4.77 7.42 -84 -35 -13 -36 -10 -2 9 -1 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Calculated cognitive style gap scores were used to find relationships with student stress, student motivation and student engage ment. PearsonÂ’s correlation coefficient was used to complete this task. For correlations internal to the measurement of cognitive style gap, total cognitive style gap was correlated with sufficiency of originality gap (r=.75, p<.05), efficiency gap (r=.46, p<.05) and rule/group conformity gap (r=.84, p<.05). These correlations indicated a positive relationship between construct m easures of cognitive style gap and total cognitive style gap. However, efficiency gap was negatively correlated with sufficiency of originality (r=-.05, p<.05). This correla tion provides evidence of a negligible relationship between the two constructs; howev er this may be explained by examining differences between the sufficiency of origin ality gap mean and the efficiency gap mean. That is, correlations between gap score constructs were affe cted by degree and direction of the gap between the faculty member and the student. Another explanation for lower correlations may be attributed to the lower posthoc reliability of the efficiency construct. One moderate correlation was found be tween efficiency cognitive style gap and conflict stress (r=.35, p<.05) indicating that more innovative students ha d higher levels of

PAGE 215

197 conflict stress in this innovativ e class; a finding that conflicts with KirtonÂ’s (2003) theory as applied in this context. No other cognitive style gap constructs were correlated with stress constructs, motivation constructs or student engagement constructs in Class I. Considering correlations internal to the measurement of student stress, total stress was correlated with the constr ucts: frustrations (r=.74, p<.05 ), conflicts (r=.67, p<.05), pressures (r=.71, p<.05), changes (r=.63, p<.05) and self-imposed (r=.67, p<.05). The data suggests that all of the constructs were at least substantially rela ted to student stress. The motivation construct self-efficacy was negatively correlated with the total student stress (r=-.60, p<.050), as well as st ress constructs: frustr ations (r=-.55, p<.05), conflicts (r=-.35, p<.05), pressures (r=-.52, p<.05) and changes (r=-.43, p<.05). These correlations indicated that lower levels of self-efficacy were associated with higher levels of stress among these stress constructs. On the other hand, test anxiety was positively correlated with total stre ss (r=.63, p<.05) and stress cons tructs: frustrations (r=.58, p<.05), conflicts (r=.33, p<.05), pressures (r= .35, p<.05) and self-imposed (r=.52, p<.05). This provided evidence that a re lationship exists between higher levels of test anxiety and higher levels of the aforementioned stress constructs in Class I. Self-imposed stress was moderately correlated with total motivation (r=.33, p<.05) and the construct extrinsic motivati on (r=.35, p<.05). This fi nding indicated an association between higher levels of self-i mposed stress and higher levels of total motivation with emphasis placed on extrinsic motivation. No moderate correlations were found between constructs of stude nt stress and constructs of student motivation in Class I. Moderate correlations internal to the measurement of total motivation were found to be moderate to substantial. Specifically, total motivation was correlated with intrinsic

PAGE 216

198 motivation (r=.58, p<.05), extrinsic motiva tion (r=.53, p<.05), ta sk motivation (r=.68, p<.05), control of learning (r=.51, p<.05), sel f-efficacy (r=.36, p<.05) and test anxiety (r=.48, p<.05). This data provide d evidence that constructs used to measure motivation were related to the total motivation scale used in Class I. Examining relationships between total engagement and motivation, total student engagement was moderately correlated with total motivation (r=.38, p<.05), intrinsic motivation (r=.33, p<.05) and task motiva tion (r=.41, p<.05). Furthermore, academic challenge was correlated with total motiva tion (r=.40, p<.05), intrin sic motivation (r=.34, p<.05) and task motivation (r=.45, p<.05). The data indicated an association between higher levels of motivation w ith higher levels of engagement among these constructs. Total student engagement had very high correlations with engagement constructs academic challenge (r=.82, p<.05), active l earning (r=.77, p<.05) and student-faculty interaction (r=.74, p<.05). Thes e correlations indicated th at constructs of student engagement were closely related to the total measurement of student engagement in Class I. See Table 4-62 for findings regarding rela tionships among cognitive style gap, stress, motivation and engagement in Class I. Student respondents were asked demographi c questions specific to age, gender number of course relevant classes taken and college cla ssification. Answers to these questions were used to find correlations w ith cognitive style gap. Gender had a moderate negative relationship with intrinsic motivation (r=-.36, p<.05) and control of learning (r=.31, p<.05). That is, females in Class I were a ssociated with lower levels of intrinsic motivation.

PAGE 217

199Table 4-62. Class I Correlations, Means, a nd Standard Deviations among Constructs of Cognitive Style Gap, Stress, Motivation an d Engagement (n=59) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .75* -3. Gap-efficiency .46* -.04 -4. Gap-rule .84* .42* .27* -5. Total stress .06 -.01 .26* -.06 -6. Frustrations .06 .07 .22 -.11 .74* -7. Conflicts .18 .02 .35* .11 .67* .40* -8. Pressures .05 -.10 .13 .12 .71* .42* .31* -9. Changes .11 .15 -.01 .06 .63* .42* .53* .41* -10. Self-imposed -.11 -.09 .16 -.21 .67* .23 .25 .38* .10 -11. Total motivation -.10 .02 .06 -.26* .21 .15 .09 .01 -.02 .33* -12. Intrinsic motivation .04 .09 .01 -.03 .29* .26* .18 .20 .21 .15 .58* -13. Extrinsic motivation -.07 .15 -.04 -.27* .13 .02 -.09 -.01 -.01 .35* .53* .06 -14. Task motivation -.14 -.14 .12 -.21 .06 .05 .12 .01 -.09 .08 .68* .47* .21 15. Control of learning -.10 -.09 .04 -.12 -.21 -.22 -.13 -.24 -.15 -.03 .51* .11 -.02 16. Self-efficacy -.05 .09 -.16 -.08 -.60*-.55*-.35* -.52*-.43*-.24 .36* .03 .17 17. Test anxiety -.01 .01 .12 -.10 .63* .58* .33* .35* .23 .52* .48* .17 .22 18. Total Engagement -.03 .16 -.09 -.16 .29* .27* .05 .27* .26* .15 .38* .33* .25 19. Academic Challenge -.13 .04 -.16 -.19 .30* .25 .11 .32* .19 .17 .40* .34* .24 20. Active Learning -.02 .10 .02 -.15 .08 .13 -.11 .01 .07 .10 .26* .19 .20 21. Faculty Interaction .16 .29* -.03 .02 .26* .24 .06 .24 .38 .05 .17 .21 .12 M -40.7-18.8-2.25-19.750.1 10.8 6.04 9.98 4.66 18.6 156 17.2 21.3 SD 14.1 7.65 4.73 7.33 11.6 4.15 2.69 2.83 2.42 4.78 18.4 4.27 4.47 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 218

200 Table 4-62 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .27* -16. Self-efficacy .17 .50* -17. Test anxiety .05 -.02 -.34*-18. Total Engagement .41* -.01 -.17 .25 -19. Academic Challenge .45* .01 -.13 .23 .82* -20. Active Learning .31* -.01 -.09 .15 .77* .39* -21. Faculty Interaction .12 -.01 -.19 .19 .74* .34* .55* -M 28.6 22.2 49.2 17.7 51.6 25.4 14.0 12.3 SD 7.41 4.06 7.12 7.82 7.61 4.32 2.73 2.64 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Considering the number of collegiate courses re lated to the subject area of Class I, a moderate correlation was found with academic challenge (r=.44, p<.05). This finding signifies that taking more c ourses related to the content of Class I was coupled with higher levels of academic challenge. All Students This study combined all participants in to one group for broader data analysis. There were 511 participants with usable KAI scores to calculate c ognitive style gap. For analysis, students retained their individual cognitive style gap with the faculty member from their respective classes. The mean cogniti ve style gap for all pa rticipants was -6.30

PAGE 219

201 (SD=28.19) indicating that on av erage, studentsÂ’ cognitive style gap in this group of students was 6.30 points more adaptive then their respective faculty member. The most adaptive cognitive style gap was 84 points lower than the studentÂ’s faculty member while the most innovative cognitive style gap was 61 points higher than the faculty member. Note that awareness of cognitive style differences between individuals occurs at a 20point gap (Kirton, 2003). Of the 511 usable KAI respons es from students, 295 (57.7%) students had at least a 20 point cognitive st yle gap with the faculty member. See Table 463 for findings relevant to cognitive style gap for all student respondents. Table 4-63. All Student Class Mean Scores of Cognitive Style Construct Gaps (n=511) Construct Mean SD Min Max Total cognitive style gap Sufficiency of originality gap Efficiency gap Rule/Group conformity gap -6.30 -4.09 1.35 -3.56 28.19 13.24 7.41 12.88 -84 -35 -19 -36 61 34 18 27 Note Cognitive style gap scores were calcula ted by subtracting faculty memberÂ’s KAI score from individual studentÂ’s KAI score. Coded: lower score equals more adaptive, higher score equals more innovative. Using calculated student cogni tive style gaps, the resear cher was able to look for relationships among student stress, student mo tivation and student engagement. PearsonÂ’s correlation coefficient was used to find such relationships. Among all the students, correlations inte rnal to the measurement of cognitive style gap were examined for the purpose of determining degree and direction of relationships. Total cognitive style gap was very highly correlated with sufficiency of originality gap (r=.90, p<.05) and rule/group co nformity gap (r=.94, p<.05). Furthermore, the correlation between sufficiency of origin ality gap and rule/group conformity gap was .80 (p<.05). This indicated a very close associ ation between these tw o constructs and the total measure of cognitive style gap. Efficien cy gap was substantially correlated with

PAGE 220

202 total cognitive style gap (r= .56, p<.05), but had a low correlation with sufficiency of originality gap (r=.23, p<.05) and a moderate co rrelation with rule group conformity gap (r=.40, p<.05). These correlations indicated that efficiency gap was associated with total cognitive style gap, but had le ss association with the other two gap constructs. Partial explanation to why efficiency gap was less re lated to other cognitive style gap constructs may exist in the differences found between cognitive style gap means for all students. That is, efficiency cognitive style gap was th e only construct mean w ith a negative value. Another explanation may exist in the lower pos t-hoc reliability of the efficiency scale. Examining correlations between cognitive style gap and student stress, no moderate correlations were found among total scores and construct scores for these students. There were no moderate correlati ons found between scales of cognitive style gap and scales of student motivation. Furthe rmore, no moderate correlations were found between scales of cognitive style gap a nd student engagement. Although there were relationships between the aforementioned scal es among individual cla sses, the researcher concluded that measures of cognitive styl e gap had little association with stress, motivation and engagement in these students. For correlations between the measurement of stress, total student stress was very highly correlated with frustrati ons (r=.81, p<.05), substantiall y correlated with conflicts (r=.66, p<.05), highly correlate d with pressures (r=.80, p<.05) substantially correlated with changes (r=.65, p<.05) and very highly correlated with self -imposed stress (r=.70, p<.05). These correlations indicated a very close relationship betw een total stress and each of the constructs measuring stress.

PAGE 221

203 Moderate correlations were found be tween total stress and the motivation construct self-efficacy (r=-.40, p<.05). Fu rthermore, self-efficacy was negatively correlated with stress constr ucts: frustrations (r=-.43, p< .05), pressures (r=-.32, p<.05) and changes (r=-.36, p<.05). These correlations indicated that lower levels of selfefficacy were associated with higher levels of these stress constructs. Test anxiety was substantially correlate d with total student stress (r=.53, p< .05) indicating an association between higher levels of stress and highe r levels of test anxiety among all student participants. Test anxiety was also mode rately correlated with stress constructs: frustrations (r=.42, p<.05), pressures (r= .43, p<.05) and self imposed stress (r=.48, p<.05). Again, higher levels of te st anxiety were associated with higher levels of stress among these constructs. There were no moderate correlations f ound among measurements of stress and measurement of student engagement among stude nts surveyed for this study. However, a low correlation was found between total stress and total engagement (r=.18, p<.05) indicating that higher levels of stress were associated with higher levels of student engagement. Concerning the measurement of motivati on, correlations were found between total student motivation and motivation constr ucts: intrinsic motiv ation (r=.68, p<.05), extrinsic motivation (r=.61, p<.05), task mo tivation (r=.75, p<.05), control of learning (r=.58, p<.05), self-efficacy (r=.44, p<.05) a nd test anxiety (r=.37, p<.05). As expected, these correlations suggest that constructs of motivation were related to the total measurement of motivation.

PAGE 222

204 A moderate correlation existed between total motivation and total engagement (r=.30, p<.05) indicating that higher levels of student motivation were associated with higher levels of student engagement am ong these respondents. Additionally, total motivation was correlated with the engageme nt construct academic challenge (r=.33, p<.05) suggesting that higher levels of student motivati on were coupled with higher levels of academic challenge. Moderate co rrelations were also found between task motivation and total engagement (r=.30, p< .05) as well as academic challenge (r=.34, p<.05). The data suggests that among these st udents, higher levels of total motivation were associated with total student engage ment with emphasis placed on the constructs academic challenge and task motivation. For the correlations specific to the m easurement of student engagement, total student engagement was correla ted with internal constructs : academic challenge (r=.77, p<.05), active learning (r=.76, p<.05), and stude nt-faculty interact ion (r=.71, p<.05). These correlations provided evidence that c onstructs measuring motivation were closely related to the total measure of student engagement. See Table 4-62 for findings regarding relationships among cognitive style gap, stre ss, motivation and engagement for all participating students. For demographic questions, student re spondents were asked to specify age, gender, number of similar courses taken and college classification. Responses were used to correlate with cognitive style gap. No moderate correlations existed among demographic variables and the measurement of cognitive style gap, stress, motivation and engagement.

PAGE 223

205Table 4-64. All Student Class Correlations, Means, and Standard Deviations among C onstructs of Cognitive Style Gap, Stress, Motivation and Engagement (n=496) Construct 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Total gap -2. Gap-originality .90* -3. Gap-efficiency .56* .23* -4. Gap-rule .94* .80* .40* -5. Total stress .15* .15* .08 .13* -6. Frustrations .16* .15* .10* .15* .81* -7. Conflicts .10* .10* .07 .08 .66* .49* -8. Pressures .11* .09* .05 .11* .80* .55* .41* -9. Changes .11* .11* .03 .11* .65* .46* .45* .51* -10. Self-imposed .06 .07 .04 .04 .70* .37* .25* .43* .21* -11. Total motivation -.04 .01 -.08 -.03 .11* .03 -.08 .07 -.11* .33* -12. Intrinsic motivation -.01 .05 -.13* .01 -.01 -.04 -.03 -.04 -.04 .10* .68* -13. Extrinsic motivation -.06 -.02 -.08 -.06 .22* .15* -.02 .17* .04 .35* .61* .15* -14. Task motivation -.11*-.12*-.03 -.10*-.06 -.05 -.17* -.07 -.19* .12* .75* .63* .31* 15. Control of learning .08 .09* -.01 .09* -.13*-.18*-.10* -.12*-.23* .08 .58* .30* .19* 16. Self-efficacy -.02 .02 -.03 -.04 -.40* -.43*-.27* -.32*-.36*-.12* .44* .27* .15* 17. Test anxiety .01 .01 -.01 .01 .53* .42* .21* .43* .28* .48* .37* .03 .28* 18. Total Engagement .01 .05 -.04 -.01 .18* .13* .10* .15* .07 .17* .30* .29* .23* 19. Academic Challenge -.04 -.02 .03 -.08 .18* .17* .07 .16* .05 .16* .33* .28* .21* 20. Active Learning -.03 .03 -.10*-.04 .08 .06 .02 .08 .04 .07 .15* .20* .13* 21. Faculty Interaction .11 .13* -.03 .12* .14* .05 .15* .09* .08 .15* .16* .16* .16* M -6.53-4.141.29 -3.6851.8 11.8 5.71 10.5 4.68 19.2 160 18.1 21.5 SD 28.1 13.3 7.39 12.8 12.9 4.36 2.63 3.57 2.38 4.66 20.2 4.47 4.44 Note. Cases excluded listwise. A ll constructs coded: higher scores equa ls increased levels. signifies p<.05

PAGE 224

206 Table 4-64 (continued). Construct 14 15 16 17 18 19 20 21 1. Total gap 2. Gap-originality 3. Gap-efficiency 4. Gap-rule 5. Total stress 6. Frustrations 7. Conflicts 8. Pressures 9. Changes 10. Self-imposed 11. Total motivation 12. Intrinsic motivation 13. Extrinsic motivation 14. Task motivation -15. Control of learning .31* -16. Self-efficacy .24* .54* -17. Test anxiety .06 -.10*-.36*-18. Total Engagement .30* .03 .03 .12* -19. Academic Challenge .34* .08 -.02 .18* .77* -20. Active Learning .17* -.06 .07 .01 .76* .34* -21. Faculty Interaction .12* .02 .04 .05 .71* .25* .44* -M 31.4 23.0 47.2 18.8 49.6 25.1 12.8 11.7 SD 7.70 3.94 7.71 7.47 7.74 4.14 3.13 3.07 Note. Cases excluded listwise. All constructs coded: higher scores equals increased levels. signifies p<.05 Group Comparisons Kirton (2003) theorized and provided evid ence that individuals with more than a 20-point cognitive style gap differ in the manner in which they solve problems. As the two individuals perceive the pr oblem differently, they also di ffer in how they generate solutions, create change and uti lize rules in their problem solv ing efforts. Stress is a result of the two individuals working to gether in solving a problem (Kirton). To test if this data supports KirtonÂ’s theory, a two-tailed indepe ndent sample t-test was conducted to examine stress score differences between st udents with less than a 20-point gap and

PAGE 225

207 students with more than a 20-point gap with their respective instructor. Groups were coded one for students with a 20-point gap or higher and zero for students with less than a 20-point gap. No significant difference was found (t=-.05, p=.96) in total stress scores between students with more than a 20-poi nt gap (M=51.76, n=284) a nd students with less than a 20-point gap (M=51.82, n=212). The data suggests that ther e was no significant difference of stress scores between students with more than a 20-point cognitive style gap and students with less than a 20-point cognitive style gap. The same data analysis was conducted to examine differences among studentsÂ’ level of motivation. A two-tailed independent sample t-test was conducted to examine motivation score differences between stude nts with more than a 20-point gap and students with less than a 20-point gap. A significant difference was found (t=-3.13, p=.00) between motivation scores of students with more than a 20-point gap (M=30.09) and students with less than a 20-point gap (M =31.20). This indicated that students with less than a 20-point gap with their respective instructor had on average scored 1.11 points higher than students with more than a 20-point gap on the MSLQ. Note that in objective 2, courses taught by innovative instructors (Classes A, B & C) and adaptive instructor s (Classes G, H & I) were found to have students with significantly lower levels of motivation than students in course s taught by instructors with a cognitive style score si milar to the general populati on mean (Classes D, E & F). Were motivation scores lower when students had more than a 20-point gap and taught by an adaptive instructor or innovative instructor? A tw o-way ANOVA was conducted to examine the interaction effect of these two variables. With motivation as the dependent variables, a significant difference was f ound (F=3.94, p=.03) for the interaction of

PAGE 226

208 students with a 20-point cogniti ve style gap and the group effect of faculty membersÂ’ cognitive style. A closer examination of group means finds that the lowest motivation scores were found in student s with a 20-point gap taught by adaptive faculty members (M=29.24). This motivation mean score was 2.07 points lower than students enrolled in the same courses with a cognitive style gap less than 20 points (M=31.32). The MSLQ scale has a range of 36 points. Consider ing students enrolled in courses taught by innovative faculty members, students with more than a 20-point gap scored less (M=30.08) than students with less than a 20-point gap (M=30.52), a difference of 0.44 points. In the middle score teaching group, the difference between those with more than a 20-point gap and those with less than a 20point gap was 0.24. The evidence suggests that in these classes, students with more than a 20-point gap enrolled in an adaptive course on average have lower levels of motivation. Another t-test was conducted to examine differences of total student engagement between students with less than a 20-point gap with their instructor and students with more than a 20-point gap with their inst ructor. No significan t difference was found (t=1.14, p=.25) between the group with less th an a 20-point gap (M=49.11) and the group with more than a 20-point gap (M=49.91). This finding provided evidence that among these respondents, students having a 20-point gap with their respective faculty member did not differ in engagement scores measured by the NSSE. Summary of Findings for Objective Three This study calculated dissimilar cognitive style by subtracting the faculty memberÂ’s cognitive style score from each individual studentÂ’s cognitive style score which yielded a cognitive style gap. Kirton ( 2003) stated that indi viduals become aware of cognitive style gap above 20 points, at which stress becomes apparent.

PAGE 227

209 As expected, faculty members who were either highly ad aptive or highly innovative also had more cognitive style gap scores higher than twen ty points with their students. For example, Class E had the least number of students ha ving a cognitive style gap of more than 20 points (n=9, 28.1%), wh ile Class I had the most students having a cognitive style gap of more than 20 points (n=55, 91.7%). Note that every class had students with more than a 20-point cognitive style gap. For all students participating in this study, 295 (57.7%) students had at least a 20 point cognitive style gap scor e with the faculty member. C onsidering all participating students, the most adaptive cognitive style gap was 84 point s lower than the studentÂ’s faculty member while the most innovative c ognitive style gap was 61 points higher than the faculty member. The total cognitive styl e gap score mean (M=-6.30) was slightly adaptive for all students combined. This was reflective of the faculty member cognitive style mean of 98.78 which was slightly innovati ve and the studentsÂ’ cognitive style mean of 93.28 which was slightly adaptive. Note th at number of student s in each class gives different amounts of weight to the mean cognitive style gap when accounting for all students. A two-tailed independent sample t-test wa s conducted to examine if stress scores differed between students with less than a 20point cognitive style gap and students with more than a 20-point cognitive style gap with their respective instructor. No significant difference was found (t=-.05, p=.96) in total st ress scores between students with more than a 20-point cognitive style gap (M=51.76, n=284) and students with less than a 20point cognitive style gap (M=51.82, n=212). Th e data suggests that students in these

PAGE 228

210 classes with a 20-point cognitive style gap with their faculty member do not have higher levels of stress, which does not support previous research co nducted by Kirton (2003). A two-tailed independent sample t-te st was conducted to examine motivation score differences between students with more than a 20-point cognitive style gap and students with less than a 20-point cognitive style gap. A significant difference was found (t=-3.13, p=.00) between motivation scores of students having more than a 20-point cognitive style gap (M=30.09) and students having less than a 20-point cognitive style gap (M=31.20). That is, a 1.11 point difference between the two groups on a 36-point scale. This indicates that in these classes, students with more than a 20-point cognitive style gap with their instructor on average ha ve lower levels of motivation. In further analysis, a two-way ANOVA was conducted to ex amine the interaction effect of students with a 20-point cognitive style gap and students enrolled in a course taught by an adaptive faculty member. With motivation as the dependent variables, a significant difference was found (F=3.94, p=.03). The lowest motivation scores were found in students with a 20-point c ognitive style gap taught by adaptive faculty members (M=29.24). This motivation mean score was 2.07 points lower than students enrolled in the same courses with a cognitive st yle gap less than 20 points (M=31.32). Another t-test was conducted to examine differences of total student engagement between students with less than a 20-point cognitive style ga p with their instructor and students with more than a 20-point cognitiv e style gap with their instructor. No significant difference was found (t=1.14, p=.25) between the group with less than a 20point cognitive style gap (M =49.11) and the group with more than a 20-point cognitive style gap (M=49.91). The data suggests that students in these cl asses with a 20-point

PAGE 229

211 cognitive style gap or more with the facu lty member did not have lower engagement scores. Among all classes, total cognitive styl e gap had very high correlations with sufficiency of originality ga p and rule/group conformity gap indicating a very close relationship. However the efficiency gap was less correlated in these classes and sometimes negatively correlated with the othe r two cognitive style constructs. This may partially be explained by the degree and direction of studentsÂ’ mean cognitive style gap in relation to the faculty member cognitive style gap. The lower post-hoc reliability of the efficiency scale may also have contributed to the efficiency constr uct being less related. For the SSI, total motivation consistently correlated with stress constructs at a substantial or very high level. This indicat ed a very close relationship between stress constructs and the total meas ure of stress. Although stress constructs did not always significantly correlate with other stress constructs, the correlations were always positive. Correlations were examined between total motivation and motivation constructs to determine the degree of association the construct had with the total measure of motivation. The constructs intrinsic motivati on, extrinsic motivation, and task motivation were always substantially or very highly correlated with the total measurement of motivation in these classes. Control of learni ng had correlations with total motivation that ranged from moderate to very high. In all classes except Class C and Class G, selfefficacy was moderately or substantially co rrelated with total mo tivation. In both these classes, this correlation was low. For test a nxiety, correlations were generally moderate in these classes except for Classes B, D and H, in which correlations were low. Still, all

PAGE 230

212 correlations were positive and evidence was provided that motivation constructs were related to the total m easurement of motivation. In these classes, total student engage ment was substantially or very highly correlated with the three constructs of student engagement used from the NSSE. However, the construct active learning had low correlations with academic challenge and student-faculty interaction in Classes B, D, E and F. This indicated that active learning was less associated with the other two engage ment constructs in these four classes. Except for Class A, total student stress had no significant rela tionship with total cognitive style gap. Class A was unique in that both the efficiency gap and the rule/group conformity gap significantly correlated with at least three stress constructs. Further examining other classes found that self-imposed stress had moderate correlations with a cognitive style gap construct in Classes C, D and G. However in these classes, the data suggests that the relationship between student stress and cognitive style gap are minimal. Examining relationships between cogn itive style gap and motivation Classes A and B had moderate correlations between e fficiency gap and specific constructs of motivation. Also, extrinsic motivation was corre lated with at least one cognitive style gap construct in every class except for Classes G, H and I, which were taught by innovative faculty members. Only Classes G and I had no moderate correlations. However, examining all students together as a group, no moderate correlations were found between any measures of motivation and measures of cognitive style gap. The data suggests that there are no consistent relationships between constructs of motivati on and constructs of cognitive style gap, although relationships do exist.

PAGE 231

213 For student engagement, cognitive styl e gaps sufficiency of originality and rule/group conformity had moderate and posit ive relationships with active learning in Classes B and H. However, Classes A, D, E, F and I had no moderate correlations between any student engagement construct and cognitive style gap construct. Likewise considering all students grouped, no mode rate correlation was found between any construct of student engagement and any construct of cognitive style gap. The data suggests that relationships do exist between cognitive style gap and student engagement, but no relationships provided a consistent pattern. In every class the motivation construc t self-efficacy was negatively correlated with constructs of stress, of ten with moderate level correl ations. This evidence supports the literature review written by Pfeiffer (2001). Also, test anxi ety consistently had at least one moderate correlation with a stress constr uct in every class. This finding helps confirm the research conducted by Misra and McKean (2000). Considering correlations between to tal student stress and total student engagement only Classes A and G had modera te correlations between the two measures; both of which were positive relationships. Also in these two classes, the frustrations stress construct was moderately correlate d with total engagement and academic challenge. The data suggests that in thes e two classes moderate relationships exist between student stress and engagement w ith emphasis placed on frustrations and academic challenge. Considering all students, total student engagement had positive and moderate correlations with total student motiva tion (r=.30, p<.05). However, a moderate correlation was not found in Classes A, E and F. Furthermore Classes A and F had

PAGE 232

214 moderate inverse relationships between constr ucts of student engagement and constructs of motivation. This finding contradicts th e depicted relationship by Astin (1984). However, among classes B, D, G, H and I ta sk motivation was moderately correlated with academic challenge and one other st udent engagement construct. Likewise, academic challenge was a prominent construct in finding correlations with motivation. Also, note that Classes B, G and H had many moderate correla tions between constructs of motivation and constructs of engagement when compared to other classes. This provided evidence that the relationships between enga gement and motivation in these classes did not offer consistent patterns of association. Pintrich and Schunk (2002) claim that e xpectancy components of motivation are related to student engagement. The two expect ancy motivational measures used in this study include control of learning and self-e fficacy. Although Classes A, B, D, F and H had a moderate correlation between one expect ancy motivation construct and one student engagement construct, the data did not provide evidence of any consistent pattern for this relationship. That is, these correlations we re negative in Classes A and F. The data provided evidence that the expectancy compone nts of motivation as measured in this study did not provide a consistent pattern to determine the relationship among these variables and student engagement. Objective Four Explain Undergraduate Student Motivation and Student Stress based on Cognitive Style Gap and Selected Demographic Variables. The goal of objective four was to use backward stepwise multiple regression to explain student stress from the cognitive st yle gap between a student and his or her respective teacher. This data analysis proce dure was utilized to determine if dissimilar

PAGE 233

215 cognitive style between a student and the faculty member inst ructing the course contributed to the explanation of student perceived stress. Finding a relationship in the appropriate direction woul d provide evidence to suppor t KirtonÂ’s Adaption-Innovation theory. Demographic variables included as i ndependent variables in data analysis were gender, age, number of similar courses the student had taken and co llege classification. Additionally, backward stepwise multiple regression was used to explain student motivation from cognitive style gap between student and faculty member. This data analysis procedure was utilized to determ ine if dissimilar cognitive style between a student and the faculty member instructing th e course contributed to the explanation of student motivation. Finding a relationship in the appropriate direction would provide evidence that students were using coping be havior (Kirton, 2003) to operate at a cognitive style closer to the faculty me mber instructing the course. Demographic variables included as independent variables in data analysis included gender, age, number of similar courses student had taken and college classification. Gender was the only categorical variable and was dummy code d one for males and two for females. For data analysis cognitive style gap was calculated by subtracting the faculty member cognitive style score from the student cognitive style score. Before employing backward stepwise multiple regression to an alyze data in each class, the researcher examined the possibility of cognitive st yle gap constructs having a curvilinear relationship with total stress and with total motivation. However, no curvilinear relationships contributed more explanation of the dependent va riable stress or motivation. Furthermore, the researcher examined the possibility of interaction effects between

PAGE 234

216 variables. No interaction eff ects were found in the data analysis for determining student stress and student motivation in these classes. Class A In Class A, student stress was explai ned by efficiency cognitive style gap ( =.33) and studentsÂ’ age ( =.31) using a backward stepwise regression analysis. The model had an adjusted R2 of .20 indicating that 20% of the vari ance of studentsÂ’ perceived stress in Class A was contributed by these two variab les (p<.05). For Class A, both independent variables had a positive relationship with student stress, but efficiency cognitive style gap was more important than age in explaining the dependent variable. The data provided evidence that 21 year-old students in Class A with an innovative 5-point efficiency cognitive style gap with the faculty member have an average total stress score of 51.22. However, 21 year-old students with no efficien cy cognitive style gap with this faculty member have an average total stress level score of 47.52. Note that total stress had a possible range of 88 points. The researcher conc luded that in Class A, controlling for age, students with an innovative e fficiency cognitive style gap with this adaptive faculty member have higher levels of total stress. This indicated that as student efficiency cognitive style gap moved from adaptiveness to innovativeness with respect to the faculty member, students have increased stress leve ls. See Table 4-65 for the unstandardized coefficient (B), intercept (Constant ), and standardized coefficient ( ). Table 4-65. Class A Backward Stepwise Multip le Regression Explaining Student Total Stress (n=57) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap Age 17.07 0.74 1.45 11.77 0.27 0.55 .33 .31 1.45 2.76 2.61 .15 .01 .01 8.04 .01 Note. Adjusted R2=.20

PAGE 235

217 In Class A, two independent variable s were best found to explain student motivation. Efficiency cognitive style gap ( =-.37) and number of similar courses ( =.17) contributed to the explanation of variance in student motivation. However, efficiency cognitive style gap was more important in explaining student motivation in Class A. Students in this course having taken 1 to 2 similar courses and with a 5-point innovative efficiency cognitive style gap with the faculty member have an average total motivation score of 28.74. Students with 1 to 2 courses ta ken with similar content to Class A and no efficiency cognitive style gap with the faculty member have an average motivation score of 30.14, a 1.40 point difference. Note the range of motivation in this study was 6 to 48. The data suggests that contro lling for the number of simila r courses taken, students with an innovative efficiency cogni tive style gap with this ad aptive faculty member have lower levels of motivation. This provided evidence that as student efficiency cognitive style gap moved from adaptiveness to innova tiveness, student motiv ation decreased in Class A. The fitted model had an adjusted R2 of .13. That is, 13% of the variance of total motivation in Class A was from efficiency gap and was significant. See Table 4-66 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ) for motivation of Class A. Table 4-66. Class A Backward Stepwise Multip le Regression Explaining Student Total Motivation (n=57) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap Number of similar courses 29.17 -0.28 0.97 1.61 0.09 0.71 -.37 .17 18.18 -2.95 1.35 .00 .01 .18 5.19 .01 Note. Adjusted R2=.13

PAGE 236

218 Class B Backward stepwise multiple regression was used to explain student stress of Class B. No significant model explained student stress given the indepe ndent variables of cognitive style gap, gender, age, number of similar classes taken and college classification. Backward stepwise multiple regression was used to explain student motivation of Class B to find the best fitting model. On e variable, efficiency cognitive style gap, explained 19% of the variance in student motivation ( =-.46, Adjusted R2=.19). The model indicates that students in Class B with an innovative 5-point efficiency style gap with the faculty member have an average total motivation score of 27.00 as opposed to students with no efficiency style gap ha ving a total motivation score of 29.35, a 2.35 point difference. The researcher concluded that in Class B, students with a higher innovative efficiency cognitive style gap with this more adaptive faculty member have less motivation. This finding indicated that as student efficiency cognitive style gap moved from adaptiveness to i nnovativeness students have decr eased levels of motivation. See Table 4-67 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ). Table 4-67. Class B Backward Stepwise Multip le Regression Explaining Student Total Motivation (n=45) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap 29.35 -0.47 0.69 0.14 -.46 42.56 -3.40 .00 .01 11.54 .01 Note. Adjusted R2=.19 Class C For Class C, backward stepwise regre ssion was used to explain total student stress. The best fitting model left three va riables including sufficiency of originality

PAGE 237

219 cognitive style gap ( =-.31), gender ( =-.21) and age ( =.22). Of these three independent variables, sufficiency of originality cognitive style gap was most important in explaining the dependent variable. The model provided evidence that 21 year-o ld female students with an innovative 5-point sufficiency of or iginality cognitive style gap with the faculty member in Class C had an average total st ress score of 3.20 points lower than the same students with no sufficiency of originality c ognitive style gap with this faculty member. The total stress scale had an 88-point range. The data suggests that controlling for age and gender in Class C, students with highe r innovative sufficiency of originality cognitive style gap with this adaptive faculty member exhib ited lower levels of stress. Findings from Class C indicated that as studen t sufficiency of origin ality cognitive style gap moved from adaptiveness to innovativeness students have decreasing levels of total stress while controlling for age and gender. The fitted model had an adjusted R2 of .13 signifying that 13% of the va riance of stress in Class C was from these three variables (p<.05). See Table 4-68 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ). Table 4-68. Class C Backward Stepwise Multip le Regression Explaining Student Total Stress (n=56) Construct B SE Beta t. Sign. Model F Sign. (Constant) Sufficiency of originality gap Gender Age 19.51 -0.64 -6.49 2.07 30.40 0.27 4.31 1.23 -.31 -.21 .22 0.64 -2.35 -1.50 1.68 .52 .02 .14 .10 3.76 .02 Note. Adjusted R2=.13 Also in Class C, total student motivat ion was regressed with the independent variables of efficiency style gap ( =-.18), gender ( =.33) and age ( =.20). Of the three variables, gender was most important in e xplaining student motivation in Class C.

PAGE 238

220 However, the explanation of motivation contributed by effici ency cognitive style gap was the focus of this objective. To interpret the model, students in Class C with an innovative 5-point efficiency cognitive style gap have an average 0.80 points lower to tal motivation score than students with no efficiency cogni tive style gap while controlling for gender and age. Note that the total motivation scal e has a 42-point range. Th e data suggests that students with a higher innovative efficiency c ognitive style gap score with this more adaptive faculty member of Class C have lower levels of total motivation. That is in Class C, as student efficiency c ognitive style gap moved from more adaptive to more innovative, students have lower levels of motivation. The model had an adjusted R2 of .14 indicating that 14% of the variance of motivation in Cl ass C was explained by these three variables (p<.05). See Table 4-69 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ). Table 4-69. Class C Backward Stepwise Multip le Regression Explaining Student Total Motivation (n=56) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap Gender Age 17.18 -0.16 2.88 0.51 8.99 0.12 1.21 0.35 -.18 .33 .20 1.91 -1.38 2.38 1.44 .06 .17 .02 .16 3.88 .01 Note. Adjusted R2=.14 Class D With respect to Class D, backward st epwise multiple regression was used to explain total student stress. The best fitti ng model included the variables efficiency cognitive style gap ( =-.15) and the demographic variab le of college classification ( =.29). For Class D, the more important variable in explaining student stress was college classification. However, the focus of objectiv e 4 was to examine the relationship between

PAGE 239

221 cognitive style gap and student stress. Students enrolled in Class D with an innovative 5point efficiency cognitive styl e gap with the faculty member have an average 1.95 points lower perceived stress score than students with no efficiency cognitive style gap controlling for college classification. The tota l stress scale had a range of 88 points. The data suggests that students with a higher i nnovative efficiency cogni tive style gap have lower perceived stress scores. Th is finding indicated that in Class D, as student efficiency cognitive style gap moved from adaptivene ss to innovativeness stude nts have decreased stress scores. This model had an adjusted R2 of .10. That is, 10% of the variance of stress in Class D was explained by these two va riables (p<.05). See Table 4-70 for the unstandardized coefficient (B), intercept (C onstant), and standard ized coefficient ( ). Table 4-70. Class D Backward Stepwise Multip le Regression Explaining Student Total Stress (n=66) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap College classification 73.70 -0.39 -5.89 7.87 0.32 2.40 -.15 -.29 9.37 -1.23 -2.45 .00 .22 .02 4.44 .02 Note. Adjusted R2=.10 Backward stepwise multiple regression was again used in Class D to explain total student motivation as a result of cognitive style gap and demographic variables. The best fitting model included efficiency cognitive style gap ( =-.21), but this model was not statistically significant (p=.08) and provided little explanatio n of the dependent variable (Adjusted R2=.03). Class E In Class E, backward stepwise multiple regression was utilized to regress total student stress with cognitive style gap betw een the student and the instructing faculty

PAGE 240

222 member as well as demographic variables in cluding gender, age, similar number of courses taken and college clas sification. However, no model wa s statistically significant. For student motivation, sufficiency of originality cognitive style gap ( =.44) and gender ( =.27) contributed to a model to expl ain student motivation in Class E. Sufficiency of originality was a more impor tant variable than gender in explaining student stress. To interpret, students in Class E with an innovative 5-point sufficiency of originality gap have an average total motiva tion score of 1.20 points higher than students with no sufficiency of originality cognitive style gap while controlling for gender. The total motivation measure used in this study had a range of 42 points. The researcher concluded that students in Class E with hi gher innovative sufficiency of originality gap with the faculty member also have higher to tal motivation scores. Said differently, as student sufficiency of originality cognitive style gap moved from more adaptive to more innovative, students have highe r levels of motivation. The model had an adjusted R2 of .13 indicating that 13% of the variance of student total mo tivation was explained by these two variables. See Table 4-71 fo r the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ). Table 4-71. Class E Backward Stepwise Multip le Regression Explaining Student Total Motivation (n=32) Construct B SE Beta t. Sign. Model F Sign. (Constant) Sufficiency of originality gap Gender 26.69 0.24 2.66 3.18 0.10 1.77 .44 .27 8.39 2.49 1.50 .00 .02 .14 3.38 .05 Note. Adjusted R2=.13

PAGE 241

223 Class F Backward stepwise multiple regression was used in Class F to explain the dependent variable of total student stress with cognitive style gap constructs and demographic variables including age, gender, number of similar courses taken and college classification. However, no models us ing these independent variables were found significant to explain student stress. Considering motivation of Class F, backward stepwise multiple regression was used to explain the dependent variable total motivation. Two variables, college classification ( =-.19) and gender ( =.26) contributed to the explanation of student motivation in Class F. None of the cognitive style gap constructs significantly contribute to the model. Gender was found most important in explaining student motivation in Class F. To interpret, college females in Cla ss F scored an averag e 2.20 points higher on the measure of total motivation than males while controlling for college classification. The model had an adjusted R2 of .06 indicating that 6% of the variance of total motivation was attributed to studentsÂ’ college classifi cation and gender. The researcher concluded that cognitive style gap offe red no explanation of student motivation in Class F, but higher motivation scores was credited to ge nder. See Table 4-72 fo r the unstandardized coefficient (B), intercept (Constant ), and standardized coefficient ( ). Table 4-72. Class F Backward Stepwise Multip le Regression Explaining Student Total Motivation (n=71) Construct B SE Beta t. Sign. Model F Sign. (Constant) Gender College classification 30.21 2.20 -0.94 2.41 0.99 0.58 .26 -.19 12.53 2.22 -1.61 .00 .03 .11 3.31 .04 Note. Adjusted R2=.06

PAGE 242

224 Class G In Class G, backward stepwise multiple regression was used to explain studentsÂ’ total stress using variables of cognitive st yle gap and student de mographics including age, gender, number of similar courses ta ken, and college classification. However, no model significantly explained student stre ss with these independent variables. Total student motivation was regressed w ith the same independent variables to find the best fitting model. Again, no signifi cant model had explained student motivation in Class G. Class H In Class H, two independent variables explained the dependent variable student stress using a backward st epwise regression analysis They include rules/group conformity cognitive style gap ( =-.36) and number of similar courses taken by the student ( =.21). In this model, rules/group c onformity cognitive style gap was more important in explaining the variance of stude nt stress. In Class H, students with an adaptive 5-point rule/g roup conformity cognitive style gap have an average 2.55 points higher than students with no rule/group confor mity gap while controlling for number of similar courses taken by the student. The tota l stress scale used in this study had an 88point range. The data suggests that students in Class H with a higher adaptive rule/group conformity gap with this innovative faculty me mber have higher total stress scores. This finding indicated that in Cla ss H, as student rule/group c onformity cognitive style gap moved from more innovative to adaptive, students have hi gher levels of stress. The model had an adjusted R2 of .13 indicating that 13% of th e variance of stress in Class H was from these two variables (p<.05). See Table 4-73 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ) for Class H total stress.

PAGE 243

225 Table 4-73. Class H Backward Stepwise Multip le Regression Explaining Student Total Stress (n=49) Construct B SE Beta t. Sign. Model F Sign. (Constant) Rules/Group conformity gap Number of similar courses 36.86 -0.51 2.02 4.63 0.19 1.31 -.36 .21 7.97 -2.66 1.54 .00 .01 .13 4.68 .01 Note. Adjusted R2=.13 For motivation of Class H, backward stepwise multiple regression was used to find the best fitting model with independent variables including cognitive style gap constructs and studentsÂ’ gende r, age, number of simila r courses taken and college classification. However no models were found significant with these independent variables. Class I Regarding Class I, backward stepwise regression was used to explain the dependent variable, student tota l stress with the best fitti ng model and most explanation by the independent variables. Cognitive styl e gap constructs and student demographic variables were included in the analysis. Two variables, efficiency style gap ( =.29) and age ( =.22) accounted for 10% of the varian ce in explaining tota l student stress (Adjusted R2=.10, p<.05). Of the two independent variables efficiency cognitive style gap was more important in explaini ng the variance of total stress in Class I. To interpret the findings concerning Class I, 21 year-old stude nts with an adaptive 5-point efficiency cognitive style gap with this faculty memb er had an average total stress score of 55.85. This was compared to the same students with no efficiency cognitive style gap having an average stress score of 52.25, a 3.60 point diffe rence. The total stress measure comprised

PAGE 244

226 a range of 88 points. The data suggests that students in Class I having a higher adaptive efficiency cognitive style gap with this innovati ve faculty member have higher total stress scores. That is, as efficiency cognitive style gap moved fr om more innovative to more adaptive, students exhibited higher levels of stress. See Table 4-74 for the unstandardized coefficient (B), intercept (Constant ), and standardized coefficient ( ) for Class I total stress. Table 4-74. Class I Backward Stepwise Mult iple Regression Expl aining Student Total Stress (n=60) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap Age 3.95 0.72 2.30 26.91 0.30 1.30 .29 .22 0.15 2.37 1.77 .88 .02 .08 4.16 .02 Note. Adjusted R2=.10 Considering total student motivation of Class I, the same independent variables were used in backward stepwise regression da ta analysis to find the best fitting model with the most explanation of student mo tivation. Rule/group conformity style gap ( =.28) and gender ( =-.20) were both consider ed significant contributo rs to the model. The adjusted R2=.08 indicating that 8% of the variance of total student motivation was explained by these two indepe ndent variables. The model was statistically significant (p<.05). Of the two independent variables, ru le/group conformity c ognitive style gap was more important in explaining the variance of student motivation in Class I. Students having an adaptive 5-point efficiency cognitive style gap with the faculty member have an average 0.70 points higher to tal motivation score than students having no efficiency cognitive style gap while controlling for ge nder. The total motivation score range was 6 to 48 points. The data suggests that students in Class I with higher adaptive efficiency cognitive style gap with the innovative f aculty member have higher total motivation

PAGE 245

227 scores while controlling for ge nder. That is, as student e fficiency cognitive style gap moved from more innovative to more adaptive students have higher levels of motivation. See Table 4-75 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ) for Class I total stress. Table 4-75. Class I Backward Stepwise Mult iple Regression Expl aining Student Total Motivation (n=60) Construct B SE Beta t. Sign. Model F Sign. (Constant) Rules/Group conformity gap Gender 29.45 -0.14 -1.52 1.94 0.06 0.93 -.28 -.20 15.20 -2.22 -1.64 .00 .03 .11 3.60 .03 Note. Adjusted R2=.08 All Students All students participating in this study were grouped together to explain student stress based on cognitive style gap and student demographic va riables. Considering all of the students, the best fitting model to expl ain total stress included the independent variables sufficiency of originality cognitive style gap ( =.15), college classification ( =.11) and number of similar courses taken ( =.09). The most importa nt variable of these three was sufficiency of originality cognitive style gap. Controlling for college classification and number of similar course s taken, students having an innovative 5-point sufficiency of originality c ognitive style gap have an av erage 0.75 point higher stress scores than the same students with no sufficiency of originality cognitive style gap. The total stress range as measured in this study was 22 to 110. The data suggests that considering these students and controlling fo r college classifica tion and number of similar courses taken, students having an innova tive sufficiency of originality cognitive style gap with their faculty member have higher stress scores. That is, as student

PAGE 246

228 sufficiency of originality cognitive styl e gap moved from more adaptive to more innovative, students have higher levels of stress. See Table 4-76 for the unstandardized coefficient (B), intercept (Constant ), and standardized coefficient ( ) for student stress. Table 4-76. All Students Backward Stepwise Multiple Regression Explaining Student Total Stress (n=496) Construct B SE Beta t. Sign. Model F Sign. (Constant) Sufficiency of originality gap College classification Number of similar courses 55.51 0.15 -1.93 1.21 2.90 0.04 0.84 0.60 .15 -.11 .09 19.12 3.51 -2.30 2.02 .00 .00 .02 .04 5.74 .01 Note. Adjusted R2=.03 Considering total motivation of all stude nts participating in this study, backward stepwise multiple regression was used to fi nd the best fitting model with the most explanation using the independent variables cognitive style gap and student demographics. Three independent variables were found to be significant (p<.05) in explaining student motivation. They incl ude efficiency cognitive style gap ( =-.08), gender ( =.12) and number of similar courses taken by the student ( =.10). Considering all the student participants and controlling for gender and number of similar courses, students with an innovative 5-point efficien cy cognitive style gap have an average 0.20 point lower total motivation score than the same students with no efficiency cognitive style gap. The total motivation scale used in this study had a 42-poi nt range. The data suggests that students participating in this study having an innovative efficiency cognitive style gap have lower total motivation scor es controlling for gender and number of relevant courses. This finding indicated th at among these participants, as student efficiency cognitive style gap moved from more adaptive to more innovative, students

PAGE 247

229 have lower levels of motivation. However, the adjusted R2 was .03 for the model indicating that only 3% of th e variance of motivation wa s attributed to the three independent variables. See Table 4-77 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ) for explanation of motivation for all students. Table 4-77. All Students Backward Stepwise Multiple Regression Explaining Student Total Motivation (n=502) Construct B SE Beta t. Sign. Model F Sign. (Constant) Efficiency gap Gender Number of similar courses 27.92 -0.04 0.97 0.41 0.81 0.02 0.37 0.18 -.08 .12 .10 34.38 -1.80 2.61 2.29 .00 .07 .01 .02 5.20 .01 Note. Adjusted R2=.03 Summary of Findings for Objective Four This study used backward stepwise multip le regression to examine if cognitive style gap contributed to th e explanation of student stress and student motivation. Considering the explanation of student st ress in adaptive courses, Class A provided evidence that students in these classes with a more innovative cogni tive style construct had increased levels of total stress. The same was true for Class D. Note that Class D was taught by a middle score faculty member, but scored 94 for tota l cognitive style designating him adaptive by one point. For inno vative courses, Classes H and I provided evidence that students in these classes with a more adaptive cognitive style construct had increased levels of stress. The data suggest s that KirtonÂ’s (2003) A-I theory was upheld in Classes A, D, H and I as a cognitive styl e construct gap did expl ain some variance of stress in these four courses. Class C was taught by a more adaptive facu lty member, but the data suggests that students having an innovative sufficiency of originality cognitive style gap have lower

PAGE 248

230 stress scores. This finding conflicts with Kirt onÂ’s A-I theory as appl ied in this context. Classes B, E, F and G found no significant e xplanation of student stressed based on a cognitive style gap between the student and th e faculty member. The data suggests that in these four courses, cognitive style gap was not a significant contributor to the explanation of student stress. Considering all students participating in this study, students having an innovative sufficiency of originality cogni tive style gap have higher le vels of stress. This makes sense as the cognitive style for all st udents on average was slightly adaptive. For demographic variables associated with explaining stress, the most frequent was age. In Classes A, C and I, older stude nts had higher levels of stress. However, considering all participat ing students as a group, stude nts having a lower college classification had higher levels of stress while controlling for number of similar courses taken. Models describing classes that found a cognitive styl e gap contributing to the variance of total stress had Adjusted R2 values ranging between .10 and .20, indicating that 10% to 20% of the vari ance was explained. Considering all students, only 3% of the variance was explained. There was evidence th at in these courses, cognitive style gap between the undergraduate studen t and the faculty member of fered some explanation to student total stress scores. For student motivation, Classes A, B, and C each found that students with an innovative efficiency cognitive style gap had lower total motivation scores. The amount of variance of student motiva tion explained in these mode ls ranged from 13% to 19%.

PAGE 249

231 Note that in the three classes taught by adap tive faculty members an innovative efficiency cognitive style gap was detrimental to student motivation. Class I was taught by an innovative faculty member and evidence from the data suggests that students with la rger adaptive rules/group confor mity gap have higher levels of motivation. Findings in Class I differed fr om findings in Classes A, B and C in that motivation increased as students bridged their sufficiency of originality cognitive style gap. The finding in Class I may suggest ev idence of coping behavior (Kirton, 2003). Interestingly, students in Class E with an innovative sufficiency of originality cognitive style gap have increased levels of motivation. Note that the faculty member instructing Class E scored 95 points for to tal cognitive style placing her in the middle score teaching group. Classes D, F, G and H found no model that included a cognitive style gap construct to significan tly explain student motivation. For demographic variables contributing to the explanation of motivation, gender was the most prominent variable. In Classes C, E, and F, females had higher motivation scores. Note that these three courses were taught by women. Conversely, in Class I males had higher levels of motivation in a course taught by a young woman. Note that the other two classes taught by women (Classes A & G) di d not have a gender effect in explaining motivation indicating that f aculty member gender was not necessarily a factor in increasing female student motivation. Objective Five Explain undergraduate student engagement based on cognitive style gap, student stress, student motivation and selected demographic variables. Objective five was achieved by using backward stepwise multiple regression to explain student engagement from the independent variables of cognitive style gap

PAGE 250

232 constructs, student stress, student motiv ation and student demographics. Student demographic variables used in this study incl uded age, gender, number of similar courses taken and college classifica tion. Gender was the only cat egorical variable and was dummy coded one for males and two for fema les. This data analysis procedure was utilized to determine if dissimilar learni ng style measured by cognitive style gap can explain student engagement. Classes were coded by letter according to the faculty member’s cognitive style score along the c ontinuum of adaptivene ss to innovativeness. Therefore, the most adaptive faculty member was assigned the letter “A” and the letter “I” was assigned to the most innovative faculty member. Class A For Class A, backward stepwise regression was used to explain student engagement. The best fitting model with the mo st explanation of the dependent variable left five variables including sufficiency of originality cognitive style gap ( =-.19), total motivation ( =.25), total stress ( =.39), number of similar courses ( =.26) and college classification ( =.30). In this model, the most impor tant independent variable was total stress, however the focus of the objective fo r this study was to examine the relationship between cognitive style gap and student engagement. Students in Class A with an innovative 5-point sufficiency of originality gap scored an average 0.80 points lower on the measure of student engagement than students with no sufficiency of originality cognitive style gap while controlling for motiv ation, stress, number of similar courses and college classification. The total engagement measure had a range of 24 to 96, a scale of 72 points. Also, note that this faculty member was more adap tive. The data suggests that in Class A, controlling for motivation, st ress, number of similar courses and college classification, students having a more innovative sufficiency of originality gap with the

PAGE 251

233 faculty member have lower total engageme nt scores. The model had an adjusted R2 of .34 signifying that 34% of the variance of student engagement in the class was from these five variables (p<.05). See Table 4-78 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ). Table 4-78. Class A Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=56) Construct B SE Beta t. Sign. Model F Sign. (Constant) Sufficiency of originality gap Total Motivation Total Stress Number of similar courses College classification 5.64 -0.16 0.53 0.27 3.07 2.88 9.45 0.10 0.25 0.08 1.40 1.10 -.19 .25 .39 .26 .30 0.60 -1.57 2.16 3.39 2.20 2.61 .55 .12 .04 .00 .03 .01 6.62 .00 Note. Adjusted R2=.34 Class B Considering Class B, backward stepwise regression was used to explain student engagement with independent variables of cognitive style gap, total stress, total motivation and student demographic variables. The best fitting model with the most explanation of the dependent variable left two variables including rules/group conformity cognitive style gap ( =.33) and total motivation ( =.38). The most important independent variable in this model was total motivati on, however the focus of this study was to examine the relationship between dissimilar cognitive style and student engagement. For students in Class B with an innovative 5-point rules/group conformity gap have an average 1.80 points higher engagement score than students with no rules/group conformity gap controlling for student motivatio n. Note that the scale used to measure of total engagement used in this study comprise d 72 points. The data suggests that students

PAGE 252

234 in Class B with more innovative rules/gr oup conformity gap have a higher total engagement score when controlling for mo tivation. The model had an adjusted R2 of .20. This signifies that 20% of the variance of st udent engagement in the class was from these two variables (p<.05). See Table 4-79 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ) determining student engagement of Class B. Table 4-79. Class B Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=45) Construct B SE Beta t. Sign. Model F Sign. (Constant) Rules/Group conformity gap Total motivation 26.69 0.36 0.55 6.23 0.15 0.20 .33 .38 4.29 2.46 2.79 .00 .01 .01 6.59 .01 Note. Adjusted R2=.20 Class C In order to explain student engagement of Class C, backward stepwise multiple regression was employed to find the best fitting model utilizing the independent constructs of cognitive style gap and variables total stress, total motivation and student demographics. Data analysis found that suffi ciency of originality cognitive style gap ( =.20), total motivation ( =.45) and age ( =.22) best contributed to total engagement of Class C. The most important independent va riable in this model was total motivation. However, the focus of this objective was to examine the relations hip between cognitive style gap and student engage ment. Students in Class C w ith an innovative 5-point sufficiency of originality c ognitive style gap have an av erage 1.25-point higher score on the measure of student engagement used in this study while controlling for studentsÂ’ total motivation and age. The scale of this m easure comprised 72 points. The researcher concluded that students in Class C with more innovative cognitive style gap with the

PAGE 253

235 faculty member have higher total engagement scores while controlli ng for studentsÂ’ total motivation and age. The model was statistical ly significant (p<.05) and the adjusted R2 was .24. This indicated that age and total mo tivation contributed to explain 24% of the variance of student engagement in Class C. See Table 4-80 for the unstandardized coefficient (B), intercept (Constant ), and standardized coefficient ( ) determining student engagement of Class C. Table 4-80. Class C Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=56) Construct B SE Beta t. Sign. Model F Sign. (Constant) Sufficiency of originality gap Total motivation Age -11.88 0.25 0.21 1.33 17.61 0.16 0.06 0.71 .20 .45 .22 -0.68 1.60 3.74 1.89 .50 .12 .00 .07 6.66 .01 Note. Adjusted R2=.24 Class D To explain student engagement of Class D, backward stepwise multiple regression was used to determine the independe nt variables that best fit into a model. Data analysis found that only total motivation ( =.38) best explained student engagement of Class D. To interpret, st udents in Class D with a tota l motivation score of 32.00 have an average total engagement score of 48.71. Th is contrasts with students with a total motivation score of 33.00 points and have an average total engagement score of 49.44. The measurement of total engagement had a range of 24 to 96. The data suggests that students who have more motivation in Class D have more total engagement. However, there was no indication th at cognitive style gap signi ficantly explained student engagement in Class D. The model was significant (p<.05) and the adjusted R2 was .13, indicating that motivation contributed to 13% of the variance in explaining student

PAGE 254

236 engagement of Class D. See Table 4-81 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ) determining student engagement of Class D. Table 4-81. Class D Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=66) Construct B SE Beta t. Sign. Model F Sign. (Constant) Total motivation 25.35 0.73 7.02 0.22 .38 3.61 3.34 .00 11.13 .00 Note. Adjusted R2=.13 Class E Backward stepwise multiple regre ssion was employed to explain student engagement in Class E given the independent variables of cognitive style gap constructs, total stress, total motivation, and selected student demographic variables. After data analysis, the best fitting model was made usi ng three variables: number of similar courses taken by the student ( =.40), total stress ( =.27) and age ( =.45). The most important independent variable in this model was student sÂ’ age. Controlling for studentsÂ’ number of similar courses taken and tota l stress, 21 year-old student s have an average 0.85 points higher score than 20 year-old students in Cla ss E. The engagement scale used in this study had a range of 24 to 96. The data suggests that students who are older have a higher engagement score while controlling for studentsÂ’ stress and number of similar courses the student had taken. However there was no indication that cognitive style gap significantly explained studentsÂ’ engagement in Class E. The adjusted R2 of the model was .26 signifying that 26% of the variance was expl ained by the three variables. The model was statistically significant (p<.05) See Table 4-82 for the unsta ndardized coefficient (B), intercept (Constant), and st andardized coefficient ( ) determining student engagement of Class E.

PAGE 255

237 Table 4-82. Class E Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=31) Construct B SE Beta t. Sign. Model F Sign. (Constant) Number of similar courses Total stress Age 17.81 2.80 0.13 0.85 8.68 1.12 0.08 0.30 .40 .27 .45 2.05 2.49 1.72 2.79 .05 .02 .10 .01 4.51 .01 Note. Adjusted R2=.26 Class F Regarding Class F, student engagement was explained using backward stepwise multiple regression. Independent variables included students cognitive style gap constructs, total stress, total motivation and student demographic variables. However, no model was found significant in explaining stud ent engagement given these independent variables. Class G Student engagement of students in Cl ass G was explained through the use of backward stepwise multiple regression to identify the independent variables that serve as the best fitting model. These independent variables included cognitive style gap constructs, total stress, total motivation and se lected student demographic variables. After data analysis, total stress ( =.25) and total motivation ( =.38) were identified to best explain student engagement of Class G. The more important of the two independent variables was total motivation for the explan ation of student engagement in Class G. Students in Class G with a motivation sc ore of 32 points have an average total engagement score of 25.92 points controlling for student stress. This compares to students with a motivation score of 33 points ha ve an average total engagement score of

PAGE 256

238 26.73 points while controlling for student stress. Again, the measure of total engagement used in this study had a range of 24 to 96 point s. The data suggests that students in Class G having higher scores on total motivation have higher scores on total engagement while controlling for student stress. However, ther e was no indication that cognitive style gap significantly contributed to the explanation of variance of stud ent engagement in Class G. The adjusted R2 for the model was .22 signifying th at 22% of the variance of the dependent variable was attribut ed to studentsÂ’ total stress and total motivation. The model was significant (p<.05). See Table 4-83 for th e unstandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ) determining student engagement of Class G. Table 4-83. Class G Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=63) Construct B SE Beta t. Sign. Model F Sign. (Constant) Total Motivation Total stress 17.53 0.81 0.13 7.58 .24 .06 .38 .25 2.31 3.31 2.24 .02 .01 .03 9.58 .00 Note. Adjusted R2=.22 Class H For Class H, student engagement was explained by using backward stepwise multiple regression to identify independent va riables that best contribute to a fitting model. Four variables were identified that contributed to the explanation of Class H student engagement. Those independent vari ables include rule/group conformity style gap ( =.40) number of similar courses previously taken by the student ( =-.26), total motivation ( =.47) and total stress ( =.30). In this model, the independent variable offering the most explanation to the variance of student en gagement was total motivation. However, the focus of the objective was to examine the relations hip between cognitive style gap and student engageme nt. In Class H, students ha ving an adaptive 5-point

PAGE 257

239 rules/group conformity cognitive style gap ha ve an average 2.20 point decrease in their student engagement score than students w ith no rules/group conformity cognitive style gap with the faculty member, controlling fo r number of similar courses taken, total motivation and total stress. The total engage ment score range was 24 to 96 points. Also note that the faculty member instructi ng Class H was highly innovative. The data suggests that students in Cla ss H with a more adaptive rules/group conformity cognitive style gap with the faculty member have lower student engagement scores when controlling for number of similar courses th e student had taken, student motivation and student stress. The model was signifi cant (p<.05) and had an adjusted R2 of .33. That is, 33% of the variance of student engagement in Class H was explained by the four independent variables listed above. See Table 484 for the unstandardized coefficient (B), intercept (Constant), and st andardized coefficient ( ) determining student engagement of Class H. Table 4-84. Class H Backward Stepwise Multip le Regression Explaining Student Total Engagement (n=49) Construct B SE Beta t. Sign. Model F Sign. (Constant) Rules/Group conformity gap Number of similar courses Total motivation Total stress 22.89 0.44 -1.92 0.93 0.23 8.45 0.14 0.93 0.24 0.10 .40 -.26 .47 .30 2.71 3.17 -2.07 3.91 2.34 .01 .01 .04 .00 .02 7.03 .00 Note. Adjusted R2=.33 Class I Student engagement of Class I was e xplained by employing backward stepwise multiple regression and finding the best fitti ng model with the most explanation of the dependent variable. Two independent variables were identified as a best fit for explaining

PAGE 258

240 student engagement in Class I. They include total motivation ( =.37), and number of similar courses take n by the student ( =.27). The most important independent variable included in this model was total motivation. St udents in Class I having a total motivation score of 31.00 points and had taken 1 to 2 sim ilar courses have an average engagement score of 50.49. This compares to the same st udents in Class I with a total motivation score of 32.00 points having an average engagement score of 51.27. The total engagement score range was 24 to 96. The data suggests that student s in Class I with higher motivation scores have higher total engagement scores while controlling for number of similar courses the students have taken. However, cognitive style gap did not offer significant explanation to student e ngagement in Class I. The adjusted R2 was .19 meaning 19% of the variance of student enga gement in Class I can be explained by the two independent variables. The model was st atistically significan t (p<.05). See Table 485 for the unstandardized coefficient (B), intercept (Constant), and standardized coefficient ( ) determining student engagement of Class I. Table 4-85. Class I Backward Stepwise Mult iple Regression Expl aining Student Total Engagement (n=59) Construct B SE Beta t. Sign. Model F Sign. (Constant) Total motivation Number of similar courses 24.00 0.78 2.31 7.59 0.25 1.00 .37 .27 3.16 3.16 2.31 .01 .01 .03 7.83 .01 Note. Adjusted R2=.19 All Students All of the students who participated in th is study were grouped together to explain student engagement based on the independent variables of cognitive style gap constructs, total stress, total motivation and student demographics. Demographic information

PAGE 259

241 consisted of studentsÂ’ age, gender, numb er of similar courses taken and college classification. Student engagement in this group was explained by the variables gender ( =-.09), total motivation ( =.30), total stress ( =.14) and student age ( =.09) using a backward stepwise regression analysis. Th e most important independent variable contributing to the explanation of variance of student engagement was studentsÂ’ total motivation. Students in this study who have a motivation score increase of one point have an average 0.59-point increase on the measure of student engagement while controlling for studentsÂ’ gender, age and total stress. Th e total engagement measure consisted of a 72 point scale. The data suggests that student s in this study with higher motivation scores have higher total engagement scores while c ontrolling for studentsÂ’ gender, age and total stress. However, cognitive style gap between students and their faculty member had no significant explanation to the variance of student engagement. The model had an adjusted R2 of .13 meaning that 13% of the variance of student engagement among students in this study was explained by the four independent student variables above. The model was significant (p<.05). See Table 4-86 for the uns tandardized coefficient (B), intercept (Constant), and standa rdized coefficient ( ). Table 4-86. All Student Backward Stepwise Multiple Regression Explaining Student Total Engagement (n=493) Construct B SE Beta t. Sign. Model F Sign. (Constant) Gender Total motivation Total stress Age 21.58 -1.48 0.59 0.09 0.37 5.00 0.70 0.08 0.03 0.18 -.09 .30 .14 .09 4.32 -2.11 7.08 3.31 2.03 .00 .04 .00 .01 .01 18.69 .00 Note. Adjusted R2=.13

PAGE 260

242 Summary of Findings for Objective Five This study used backward stepwise mu ltiple regression to examine if cognitive style gap contributed to the e xplanation of student engagement The nine classes in this study were coded by letter according to the facu lty member’s cognitive style score along the continuum of adaptiveness to innovativen ess. The most adaptive faculty member’s course was coded with the letter “A” and th e most innovative faculty member’s course was coded with the letter “I”. For student engagement, only four of th e nine classes (A, B, C and H) had a cognitive style gap construct si gnificantly explain student en gagement. In Class A, this relationship was negative indi cating that students having a higher innovative gap have lower engagement. In Class H, this relations hip was positive indicating that students with a more adaptive cognitive style gap have lowe r engagement scores. Both of the findings in these two courses provide evidence that a cognitive st yle construct gap inhibited student engagement. In Classes B and C the opposite was found as a positive relationship indicated that students with an innovative cognitive styl e construct gap have higher levels of engagement. Classes B and C were taught by more adaptive faculty members. The highest level of variance in stud ent engagement explained among these models was 34%; some of which included four independent variables. Also note that in objective four, the efficiency cognitive style construct gap was predominate in explaining stress and motivation, but was not found to si gnificantly explain st udent engagement in any of the nine courses. Examining the relationships other variables, student motivation was significant in explaining student engagement in seven of th e nine courses. In these seven models the

PAGE 261

243 relationship was positive indicating that students having higher levels of motivation have higher levels of engagement. Stress added to the explanation of st udent engagement in four of the nine courses; all of which indicated th at students with hi gher levels of stress have higher levels of engagement. For demographic variables contribution to student engagement, the number of courses previously taken that was similar to the present course significantly explained student motivation in four cour ses. In Classes A, E and I, evidence suggested that more courses taken that related in content to the present course was related to more student engagement. Conversely, the opposite was found in Class H; students having taken more classes have less student engagement. Although number of courses previously taken was a significant variable in expl aining engagement in four cour ses, it was not a significant explanatory variable when compiling all stude nts together. For all students, gender and age were the two demographic variables sign ificant in explaining student engagement. That is, among these respondents females were less engaged when controlling for age, and older students were more enga ged while controlling for gender. Summary This chapter presented the findings speci fic to the objectives of the study with details given to each specific class, as well as groups of students determined by the faculty memberÂ’s cognitive style score and fi nally all students who participated in the study. The objectives of the study were to 1) describe selected faculty and students according to their selected demographic vari ables; 2) determine the cognitive style, student stress, student motivation, and student engagement of underg raduate students; 3) determine the cognitive style gap between faculty and students and to explore its relationship with student stre ss, student motivation, student engagement and specific

PAGE 262

244 demographic variables of undergraduate student s; 4) explain undergraduate student stress and student motivation based on cognitive style gap and selected demographic variables; and 5) explain undergraduate student engage ment based on cognitive style gap, student stress, student motivation and selected demographic variables. Chapter 5 will summarize the findings presented in this chapter and provide conclusions and implications based on th ese results. Recommendations will also be provided.

PAGE 263

245 CHAPTER 5 SUMMARY AND DISCUSSION This chapter summarizes the study and provides conclusions, implications, and recommendations derived from the results of this study. The first section of this chapter presents a summary of the problem stat ement, purpose and objectives, methodology and findings. The second section presents conc lusions and discussion specific to each objective. The end of the chapter provides recommendations as a re sult of this study. Summary of Problem Statement Trends in undergraduate education are moving from teacher-centered classrooms to student-centered classrooms for the impr ovement of student engagement (Acharya, 2002). However, twenty percen t of freshman and seniors in colleges and universities across the nation were still identified as disengaged (Kuh, 2003). Even though instructional techniques of faculty members have improved, student engagement in college classrooms still seems to be at a low level. In the search for variables to explain student engagement, researchers have examin ed the cognitive level of instructional discourse (Whittington, 1998), but few have examined the cognitive style of the instructor. That is, does a dissimilar learni ng style between student and faculty member contribute to explaining w hy some students are not e ngaged in the classroom? KirtonÂ’s (2003) measure of learning styl e is independent of learning capability and motivation towards learning; this theore tical view is different than many other measures of learning style (Coffield, Mose ley, Hall & Ecclestone 2004). The separation of learning style from lear ning level is warranted to allow for the distinguished

PAGE 264

246 relationship and accurate measure of other classroom variables such as engagement, motivation and stress. Learning style and learning level are both employed by the student to complete the learning process (Kirton, 2003) However, most learning style theories assert that students learn best in their own learning style (Coffield et al., 2004). Students receiving instructional discour se from a faculty member with a dissimilar learning style may cause stress (Kirton, 2003) which may inhibit student engagement (Pintrich & Schunk, 2002). Separating learning style from l earning level allows the researcher to examine each independently and determine the significance each has on the dependent variable. That is, if dissimilar learning style is not f ound to be a large contributor to student engagement, perhaps learning level is a superseding variable. Faculty members are increasingly faced w ith more accountability for increasing student achievement. As student engagement is closely related to academic achievement (Kuh, 2001), faculty members need to utilize in structional strategies and curricula to increase student engagement a nd foster student learning. Resear ch is needed to determine if cognitive style contributes to the explanat ion of student engagement. Does dissimilar learning style between studentsÂ’ preferre d cognitive style and facultiesÂ’ preferred cognitive style impair student engagement? Are students able to overcome cognitive style gap as the faculty member engages the learner? Purpose and Objectives The purpose of this study was to determ ine if significant relationships exist between cognitive style gap, student stress, student motivation, student engagement and selected demographic variables of undergraduate students at th e University of Florida in the College of Agriculture and Life Sciences The specific objectives of this study were to: 1) describe selected faculty and student s according to their selected demographic

PAGE 265

247 variables; 2) determine the cognitive style, student stress, student motivation and student engagement of undergraduate students; 3) determine the cognitiv e style gap between faculty and students and to expl ore its relationship with stude nt stress, student motivation, student engagement, and specifi c demographic variables of undergraduate students; 4) explain undergraduate student stress and st udent motivation based on cognitive style gap and selected demographic variables; and 5) explain undergraduate student engagement based on cognitive style gap, student stress, student motivation and selected demographic variables. Methodology The researcher used an ex post facto design one (Ary, Jacobs & Razavieh, 2002) to accomplish the objectives of the study. This design allows for the control and measure of the independent variable(s) to test hypot heses concerning varia tion in the dependent variable(s). Objectives one and two were accomplished using descriptive statistics (frequencies and measures of central tende ncy). Objective three was accomplished by finding the difference of cognitive style ga p score between faculty members and their respective students. PearsonÂ’s correlation coeffi cient was then used to find relationships among variables identified in th is study. Objectives four and five were accomplished through backward stepwise multiple regression to determine the effects of cognitive style gap on student stress, student motivation a nd student engagement. Student demographic variables for the study include age, gender, academic major, fulltime status, college classification and number of classes similar to the course examined in this study. Students were also asked how many problem sets assi gned during a week had taken more than one hour to complete to provide evidence that pr oblem solving assignments and projects were used in the classes examined in the study.

PAGE 266

248 To select undergraduate cl asses to participate in th is study, sixty-four faculty members in the College of Agriculture and Li fe Sciences at the University of Florida were contacted. Faculty members were iden tified by contacting indi viduals serving as undergraduate coordinator of each department in the college. Although many times the identified faculty members believed their c ourse did not match the criteria of the study, they often provided contact information of another faculty member within the same department whose course was better suited. In the end, 21 faculty members were willing to participate in the study. From this group, only 15 faculty members were identified as suitable for the study. Nine faculty members were chosen based on their cognitive style score. The selection of faculty members pr ovided a range along the adaptive-innovative continuum as three faculty members had similar adaptive cognitive style scores, three faculty members had similar innovative cogniti ve style scores and three faculty members had similar cognitive style scores located in the middle of the adaptive-innovative continuum. Each of the nine faculty members’ course s was coded with a letter to maintain confidentiality in data pres entation. In order of adaptiven ess to innovativeness faculty members were coded beginning with the letter “A” and ending with the letter “I”. The grouping of classes by faculty member cognitiv e style resulted in an adaptive teacher group (Classes A, B and C), a middle score teacher group (Classes D, E and F) and an innovative teacher group (Classes G, H and I). Four instruments were used to collect da ta in order to accomplish the objectives of this study. The Kirton Adaption-Innovation Inventory (KAI) was used to determine the cognitive style of students and faculty member s. It was also used in calculating the

PAGE 267

249 cognitive style gap that existed between st udents and faculty members. Cognitive style gap was calculated by subtracting the faculty member’s cognitive style score from each student’s cognitive style score. Constructs of the KAI include sufficiency of originality, efficiency and rules/group conformity. The Student-life Stress Inventory (SSI) was used to measure student’s classroom stress and included factors of frustration, conflicts, pressures, changes and self-imposed stre ss. The Motivation Strategies for Learning Questionnaire (MSLQ) determined student moti vation with scales of intrinsic motivation, extrinsic motivation, task value, control of le arning beliefs, self-effica cy for learning, and test anxiety. Finally, the National Survey of Student Engagement (NSSE) (2005) was used to determine student engagement with constructs of academic challenge, active and collaborative learning, and student-faculty interaction. The res earcher also used demographic items found on the NSSE. Chap ter 3 discusses how instruments were altered to meet the objectives of this study. All four in struments were used with permission. Conclusions and Discussion The findings and conclusions of this study are summarized in order of their respective objective. Nine undergraduate cour ses were used for this study and were presented in order of the faculty members’ cognitive style score using the letters “A” through “I”. The faculty member with the mo st adaptive cognitive style score was coded with the letter “A” while the faculty member with the most innovative cognitive style was coded with the letter “I”. See Table 4-1 for the faculty members’ sp ecific cognitive style scores. Each class and a group us ing all student participants we re subject to the same data analysis procedures. The reader should use cau tion in applying these conclusions to other populations as classes in th is study were intact groups.

PAGE 268

250 Objective One The first objective described faculty and students according to their demographic variables respective of each class. First, f aculty members were asked their age, gender and years of teaching experience. For students, demographic variables were taken from the National Survey for Student Engageme nt (NSSE) which included studentÂ’s age, gender, major, college classification, fulltime status, and number of courses taken similar to the subject area of the class. Fu rthermore, students were asked to state the number of problem sets given during a typical week that had taken more than an hour to complete. This question was asked only to provide evidence that problem solving assignments and projects existed in the cl asses used in the study. Conclusions for objective one will be discussed after a brief presentation of the findings. Class A The instructor of Class A was a 25 year-o ld female graduate assistant. She had three years of teaching experience at the colleg e level, but no prior teaching experience in any other setting. There were 100 students enrolled in Cl ass A, of which 70 responded. Of the respondents, 55.7% (n=39) were female. For Class A, the mode age was 20 years old (32.9%, n=23). Almost half of Class A we re juniors (48.6%, n=34), while 30% were seniors (n=21), and 18.6% were sophomores (n=13). Regarding the number of courses the students had taken in th e subject area respective of Class A. Respondents reported 17.4% (n=12) had taken no courses, 68.1% (n=4 7) had taken one to two courses, 13.0% (n=9) had taken three to four courses, and 1.4% (n=1) had taken more than four courses in the subject area.

PAGE 269

251 Class B The faculty member providing instruction for Class B was a male at 54 years of age. He had taught at the University of Fl orida for 29 years and had no other teaching experiences. There were 100 students enrolled in Class B, of which 72 responded. The majority of the respondents in Class B were female (81.2%, n=56). The mode age was 21 years old (37.7%, n=26). The class was only composed of junior s (50.7%, n=35) and seniors (49.3%, n=34). Students were asked how many courses were taken in the subject area respective of Class B. Respondents re ported 23.2% (n=16) had taken no courses, 37.7% (n=26) had taken one to two courses, 13. 0% (n=9) had taken three to four courses, and 26.1% (n=18) had taken more than four c ourses in the subject ar ea. There were three students that did not answer this specific question. Class C The faculty member instructing Class C was a 36 year-old female. She had taught 12 years at the college level, but had no teaching prior teaching experience. Two sections of Class C were combined to form a group total of 90 students, of which 71 responded. The majority of the studen ts in Class C were male (71.8%, n=51). The mode age of Class C was 22 years old ( 35.2%, n=25). Class C comprised mostly of seniors (91.5%, n=65) with the remainder of the class being j uniors (8.5%, n=6). Respondents reported 2.8% (n=2) had taken no courses, 23.9% (n=17) had taken one to two courses, 26.8% (n=19) had taken three to four courses, and 46.5% (n=33) had taken more than four course s in the subject area.

PAGE 270

252 Class D The faculty member for Class D was a 62 y ear-old male with 38 years of teaching experience at the college level. He had no prior teaching experience before becoming a university faculty member. Class D enrolled 116 students of which 108 participated in this study. The majority (81.3%, n=87) in the class were fe male. The mode age of Class D was 21 years old (36.4%, n=39). The college classification of students in Class D comprised 60.4% (n=64) juniors, 32.1% (n=34) seniors, 6.6% (n=7) sophomores, and one freshman. Respondents reported 5.7% (n=6) had taken no courses, 32.1% (n=34) had taken one to two courses, 28.3% (n=30) had taken three to four courses, and 34.0% (n=36) had taken more than four courses in th e subject area. There were two students that did not answer this specific question. Class E The faculty member of Class E was a 54 year-old female. She had 23 years of teaching experience at the college level. Ho wever, she had no prior teaching experience at any other level. There were three sections of Class E which combined to form a group of 105 students, of which 48 responded. Considering ge nder, female students held the majority (77.1%, n=37). Student age in Class E was bimodal with 14 students (29.2%) being 21 years of age and 22 years of age. Class E was made up of mostly seniors (64.6%, n=31) and juniors (33.3%, n=16). Students were as ked how many courses were taken in the subject area respective of Class E. Regarding nu mber of classes taken relevant to Class E, respondents reported 8.3% (n=4) had taken no courses, 54.2% (n=26) had taken one to

PAGE 271

253 two courses, 14.6% (n=7) had taken three to four courses, and 22.9% (n=11) had taken more than four courses in the subject area. Class F The faculty member who taught Class F wa s a 46 year-old female who had taught 11 years at the University of Florida. Previous to that, she had worked as a dietician for 7 years and conducted some educational programs. There were 150 students enrolled in Cl ass F of which 115 participated. The majority of the course was female (69.6%, n= 80). Class F students had a mode age of 21 years old (43.5%, n=50). Of the respondents, juniors made up the majority (58.3%, n=67) of Class F with seniors constituting 33.0% (n =38). Also enrolled in Class F were six sophomores (5.2%) and four freshmen ( 3.5%). Respondents reported 1.7% (n=2) had taken no courses, 29.6% (n=34) had taken one to two courses, 44.3% (n=51) had taken three to four courses, and 24.3% (n=28) had taken more than four courses in the subject area relevant to Class F. Class G The faculty member who taught Class G wa s a 35 year-old female. She had taught five years in her current position at the Univer sity of Florida, eight years tutoring student athletes, and two years as a cooperative extension agent. Class G enrollment consisted of 110 student s of which 85 agreed to participate. Most of the respondents in Class G were female (81.2%, n=69). Eleven respondents did not answer the question regard ing demographics. Class G students reported a mode age of 21 years old (34.1%, n=29). Class G res pondents were 57.6% (n =49) juniors, 30.6% (n=26) seniors, and 11.8% (n=10) sophomore s. Students were asked how many courses were taken in the subject area respective of Class G. Respondents reported 2.4% (n=2)

PAGE 272

254 had taken no courses, 23.8% (n=20) had ta ken one to two courses, 41.7% (n=35) had taken three to four courses, and 32.1% (n=27) had taken more than four courses in the subject area. There were 12 students that did not answ er this specific question. Class H The faculty member who taught Class H was a 42 year old male. He had five years of collegiate teaching experience at the Un iversity of Florida. This faculty member also reported teaching experiences with the Peace Corps and an agricultural technical college. Class H consisted of 122 students of which 70 responded. Males held a slim majority of the respondents at 55.1% (n=38) For ages of respondents in Class H the mode age was 21 years old (31.9%, n=22). Cl ass H respondents were primarily juniors (53.6%, n=37) and seniors (36.2%, n=25). Sophom ores made up 7.2% (n=5) of the class and freshman accounted for 2.9% (n=2). St udents were asked how many courses were taken in the subject area respective of Cl ass H. Respondents reported 15.9% (n=11) had taken no courses, 34.8% (n=24) had taken one to two courses, 27.5% (n=19) had taken three to four courses, and 21.7% (n=15) had taken more than four courses in the subject area. There was one student that did not answer this specific question. Class I The instructor who taught Class I was a 26 year-old female graduate assistant with one year of collegiate teaching experience. Furthermore, she had less than 6 months of experience leading non-formal educa tional programs at a community level. There were 100 students enrolled in Cla ss I with 77 participating in this study. Females constituted 62.3% (n=48) of the student respondents. Concerning the ages of Class I, respondents were bimodal with 34.2% (n=26) each at 20 and 21 years of age.

PAGE 273

255 Class I respondents reported their college clas sification as 48.1% (n=37) juniors, 26.0% (n=20) seniors, 20.8% (n=16) sophomores, and 5.2% (n=4) freshman. Students were asked how many courses were taken in the subj ect area similar to Cla ss I. In response, it was found that 34.2% (n=26) had taken no cour ses, 46.1% (n=35) had taken one to two courses, 13.2% (n=10) had taken three to f our courses and 6.6% (n=5) had taken more than four courses in the subject area. All Participants For all faculty members, teaching experience ranged from a graduate student employed as a lecturer to tenured faculty me mbers. Years of experience ranged from 1 to 38 years while ages of these faculty members ra nged from 25 to 62 years. Six of the nine faculty members were female. Faculty member s represented six academic departments in the College of Agriculture and Life Scie nces at the University of Florida. There were 993 undergraduate students enrolle d in the nine courses examined in this study, of which 716 participated. Of th e student respondents, 65.7% (n=467) were female. The mode age of student participan ts was 21 years (33.6%, n=239). The majority of the students were juniors (48.6%, n=345) and seniors (41.4%, n= 294). Considering the number of courses students had taken in the same subject area to the course of which instruments were administered, 11.5% (n=81) had taken no courses, 37.2% (n=263) had taken one to two courses, 26.7% (n=189) ha d taken three to four classes and 24.6% (n=174) had taken more than four courses. Discussion for objective one Analysis of the faculty members in the st udy found that six of the nine agreeing to participate were female. There was a variet y of years of teaching experience among the faculty participants ranging from the faculty member for Class D who had taught 38

PAGE 274

256 years to the Class I faculty member who ha d about six months of teaching experience. Furthermore, faculty participantsÂ’ in this study ranged from a graduate student to full professor. Demographic variab les of faculty members were provided only to describe the selected faculty members instructing the cla sses and not used in further analysis. There was no indication found in the literature th at faculty member demographics greatly contribute to student stress, motivation or engagement. Furthermore, research has found that teachers give instructional discourse in their own preferred learning style despite other individual differences and situa tions (Dunn & Dunn, 1979; Farthing & Stubbs, 2003; Witkin, 1973). It was determined that faculty member demographic information was not needed for further data analysis. This study used student demographic variab les to describe each class used in the study and to determine if demographic variable s contributed to the e xplanation of student stress, student motivation and student engage ment (Gadzella & Guth rie, 1993; Gadzella & Fullwood, 1992; Pike, 2004). These comparisons will be made in the following objectives. Of the student respondents, 65.7% (n=467) were female. As expected, some classes had more females than males or more males than females depending on the academic department in which the course wa s based. For example, females accounted for almost 70% of the participating students of Class F; a class re quired for nutritional science and dietetics majors of which the majority are female Still, seven of the nine classes were mostly female. Most of the students were classified as full-time students. The majority of participating students were also between th e ages of 18 and 23 years of age, giving

PAGE 275

257 evidence that classes were mostly filled with traditional college students. Furthermore, the majority of students in the classes of study were either j uniors (48.6%, n=345) or seniors (41.4%, n=294). This i ndicated that courses used in this study were of higher level. Regarding the number of courses the st udent had taken with similar content varied across classes. However, considering all pa rticipants, 50.7% (n=363) of the students had taken at least three classes that were similar to the course used in the study. This not only helps confirm that classes were of higher le vel, but also that these students had some content familiarity with the course. Objective Two Objective two determined the cognitive styl e, student stress, student motivation and student engagement of undergra duate students. Four separate instruments were used to determine student cognitive style (KAI), student level of stress (SSI), student level of motivation (MSLQ), and the level of student e ngagement (NSSE) specific to the class to which the instruments were administered. Conclusions for objective two will be discussed after a brief presentation of the findings. Class A For Class A (N=100, n=70), the total cogniti ve style mean of student respondents was 4.40 points more adaptive (M=90.60, SD=18.01, n=58) than the general population norm determined by Kirton (1999). Class AÂ’s total stress mean for student respondents was 52.44 (SD=10.57, n=68), 17.65 points lower th an reported norms (Gadzella & Baloglu, 2001). For motivation, a mean of 29.31 (SD=3.59, n=69). This was 1.15 points lower than the norm reported by Pintric h, Smith, Garcia and McKeachie (1991). Regarding student engagement, the summed me an of the three construct scores was 48.43

PAGE 276

258 (SD=7.70, n=70). This value was 15.31 points lo wer than the reported national mean for college seniors (Kuh, Hayek, Carini, Ouimet, Gonyea & Kennedy, 2001). Class B For Class B (N=100, n=72), studentsÂ’ tota l cognitive style mean score (M=92.43, SD=15.05, n=46) was 2.57 points more adaptiv e than the general population (Kirton, 1999). The total stress mean for student in Class B was 55.85 (SD=12.95, n=68). This score was 14.24 points lower, or one standard deviation lower than the norm (Gadzella & Baloglu, 2001). Of total student motivati on of Class B, the mean was 30.65 (SD=12.95, n=68), only 0.19 point higher than test sample from the work of Pintrich et al. (1991). Class B total student engagement me an was 17.73 points lower (M=46.01, SD=7.47, n=70) than the national mean for college seniors (Kuh et al., 2001). Class C For Class C (N=90, n=71), the total cogni tive style mean was slightly more innovative (M=100.86, SD=14.34), or 5.86 point s higher than the general population mean defined by Kirton (1999). Total stress mean of participants in this class was 15.18 points lower (M=54.91, SD=12.59, n=70) than the reported test sample mean (Gadzella & Baloglu, 2001). Total motivation in this co urse was had a mean 0.40 points higher (M=30.86, SD=3.80, n=71) than the standardi zed norm (Pintrich et al., 1991). Total student engagement score mean of Cl ass C (M=54.73, SD=9.15, n=71) was 9.01 points lower than the national mean for college seniors (Kuh et al., 2001). Class D For Class D (N=116, n=108), the total cogni tive style mean for Class D student respondents was 93.63 (SD=17.40, n=73). This indi cated students in the class were slightly more adaptive as the mean was 1.37 points more adaptive than the standard mean

PAGE 277

259 of typical populations defined by Kirton (1999) Total stress score mean for the class (M=52.80, SD=13.71, n=101) was 17.29 points lower than the standardized norm mean of 70.09 points (Gadzella & Baloglu, 2001). Ho wever, Class D total student motivation (M=32.42, SD=3.77, n=102) was 1.96 points higher than the norm mean reported by the instrument authors (Pintrich et al., 1991). C oncerning student engagement of Class D, the total calculated student e ngagement mean was 48.33 (SD=7.89, n=108). That value was 15.41 points lower than the national reported m ean for college seniors (Kuh et al., 2001). Class E For Class E (N=105, n=48), the total cogni tive style mean was slightly more adaptive (M=89.41, SD=14.95, n=32), which wa s 5.59 points lower than the general population mean of 95.00 (Kirton, 1999). The total stress mean of Class E was 42.89 (SD=11.28, n=47) which was more than two stan dard deviations lower than the test sample used by Gadzella and Baloglu (2001) Considering motivation for Class E, the total score mean (M=31.20, SD=3.79, n=48) wa s 0.74 points higher than test sample mean (Pintrich et al., 1991). The total student engagement mean score was 47.88 (SD=5.50, n=48) which was 15.86 points lower than the national reported college senior total engagement mean (Kuh et al., 2001). Note that the faculty member instructing Class E did not encourage student pa rticipation by providing extracredit, as the other eight faculty members did. The lack of student par ticipation in Class E makes this class more subject to non-response error. However, no data were collected from non-responders. Class F For Class F (N=150, n=115), responding students were slightly adaptive (M=90.17, SD=16.28 n=71), which was 4.83 point s lower than the general population mean (Kirton, 1999). The total stress scor e mean of respondents in Class F was 49.29

PAGE 278

260 (SD=13.20, n=112). This compares to being 20.8 0 points lower than the test sample reported by Gadzella and Bal oglu (2001). However, the total motivation mean (M=31.25, SD=3.71, n=115) for Class F was 0.79 points highe r to the reported norm mean of 30.46 (Pintrich et al., 1991). Class F respondentsÂ’ total student engagement mean of 47.87 (SD=7.81, n=115) was 15.87 points lower than the national mean for college seniors (Kuh et al., 2001). Class G For Class G (N=110, n=85), the total cognitive style mean was 93.42 (SD=15.04, n=65) indicating a slightly more adaptive group; the mean score was 1.58 points lower than the general population mean determined by Kirton (1999). The student total stress mean of Class G was 51.54 (SD=13.72, n=84), 18.55 points lower than the results published by Gadzella and Baloglu (2001). Th e Class G total student motivation mean score was 30.01 (SD 4.27, n=85). That is 0.45 po ints lower than th e total motivation norm provided by Pintrich et al. (1991) indicating a slightly lower level of motivation in this class. Considering total student engage ment for Class G respondents, the mean was 50.05 (SD=7.80, n=85), 13.69 points lower than the national average of 63.74 for college seniors (Kuh et al., 2001). Class H For Class H (N=122, n=70), the total cogni tive style mean of Class H was 95.30 (SD=15.01, n=50), only 0.30 points higher than th e general population cognitive style of 95 (Kirton, 1999). Student total stress m ean (M=50.73, SD=11.54, n=70) in Class H was approximately 19.36 points lower than the reported normalized mean (Gadzella & Baloglu, 2001) indicating lower levels of stress. Considerin g student motivation in Class H, the total motivation mean score was 31.05 (SD=4.60, n=70) which was 0.59 points

PAGE 279

261 higher than the normalized mean reported by Pintrich et al. (1991). The total student engagement mean (M=52.28, SD=8.59, n=69) was 11.46 points lower than the national mean for college seniors (Kuh et al., 2001). Class I For Class I (N=100, n=70), the student tota l cognitive style mean of the Class I was 92.90 (SD=14.25, n=60). This indicated a sl ightly adaptive student group with 2.10 points lower than the general population repor ted by Kirton (1999). Th e total stress mean for Class I was 49.82 (SD=11.96, n=77), which wa s 20.27 points lower than the reported norm stress mean (Gadzella & Baloglu, 2001). For motivation, the student total motivation mean of Class I was 29.86 (SD= 3.74, n=77). That value was 0.60 points lower than the total motivation mean reported by Pintri ch et al. (1991). Stude nt participants had a total engagement mean of 51.58 (SD=7.72, n=76), 12.16 points lower than the national mean for college seniors (Kuh et al, 2001). All students All participating st udents (N=993, n=716) of the nine classes were grouped together to determine total mean scores for cognitive style, stress, motivation and engagement. The cognitive style mean sc ore of the total group was 93.28 (SD=15.95, n=511), which was 1.72 points lower than the general population defined by Kirton (1999). This indicated that overa ll, student participants were slightly adapti ve. The total stress mean for all students was 51.35 (SD=12.91, n=697), 18.74 points lower than the reported norm (Gadzella & Baloglu, 2001). For total motivation, the mean was 30.81 (SD=4.04, n=706), 0.35 points higher than the no rm reported by Pintrich et al. (1991). Considering all participati ng students the total enga gement mean was 49.58 (SD=8.20,

PAGE 280

262 n=712) which was 14.16 points lower than the national mean for college seniors (Kuh et al., 2001). Comparison between classes A one-way ANOVA was conducted to examin e significant differences of student stress levels between students taught by ad aptive faculty members, middle score faculty members and innovative faculty members. A significant difference was found between the three groups perceived stress scores (F=8.92, p=.00) and a Bonf erroni post-hoc test was conducted to further examine between group differences. The adaptive teacher group had student stress scores (M= 54.51) that were significantly higher than the student stress scores of the middle score teacher group (M= 49.49, p=.00) and significantly higher than student stress scores of the innovative teacher group (M=50.72, p=.01). This finding coincides with the results reported by Pu ccio et al. (1993) who found that courses perceived to be more adaptive ha d increasing levels of stress. A one-way ANOVA was conducted to examin e significant differences of student motivation scores between courses taught by adaptive teacher group, courses taught by the middle score teacher group and course s taught by the innovative teacher group. A significant difference was found between the three groups (F=10.41, p=.00). A Bonferroni post-hoc test was used to find group differences. Student s in the innovative teacher group had total motivation scores (M =30.27) that were significantly lower than student motivation scores in the middle score teacher group (M=31.69, p=.00). Student respondents in the adaptive teacher group also had significantly lower motivation scores (M=30.28, p=.00) than found in the middle score teacher group. These findings indicated that these students taught by adaptive a nd innovative faculty on average have lower

PAGE 281

263 levels of motivation th an students taught by faculty members with a cognitive style close to the general population m ean of 95 (Kirton, 2003). To determine if student engagement va ried among adaptive faculty members and innovative faculty members, a one-way AN OVA was conducted to examine significant differences between an adaptive teacher group, a middle score teacher group and an innovative teacher group. Using total student engagement as the dependent variable, a significant difference was found between groups (F=9.61, p=.00). A Bonferroni post-hoc test was conducted to further examine gr oup differences. Students in the innovative teacher group had total engagement scores (M=51.22) significantly higher than engagement scores of students enrolled in the middle score teacher group (M=48.05, p=.00). However, students in the innovativ e teacher group did not have significantly higher engagement scores than students in the adaptive teacher group (M=49.75, p=.17). These findings indicated that students in these classes taught by innovative faculty members tended to have higher levels of stude nt engagement, but not significantly higher than adaptive faculty members. Discussion for objective two Cognitive style was determined by the KAI. Among the classes, all cognitive style score means were less than ten points away from 95 which indicates that these students on average were not highly adaptive or highly innovative. That is, these classes on average had the same cognitive style as the general population mean (Kirton, 2003). Overall, studentsÂ’ cognitive style mean score was 93.28 (SD= 15.95). Fluctuation of cognitive mean scores between classes may be attributed to the number of female students in the class. Kirton ( 1999) states that the total c ognitive style mean score for females is 90, which is five points lower than the general population mean. In this study,

PAGE 282

264 the overall cognitive style mean score for females was 91.49 (n=339), while the cognitive style mean score for males was 96.81 (n=170) Therefore, classes with a larger percentage of females tended to ha ve more adaptive mean scores. The level of student perceive d stress was determined by the SSI. All of the classes had student mean stress levels between 42.89 (Class E) and 55.85 (Class B). Overall, stress scores were low compared to the findings of Gadzella and Baloglu (2001). The lower stress scores may be explained by examining the time of year the SSI was administered. Shields (2001) found that st udent stress decreases during the second semester of the academic year. Data for this study were collected in the second semester of the academic year for this study, however Gadzella and Baloglu (2001) did not report the semester to which the SSI was administer ed. Nevertheless, student stress scores were still low. Another small contributor to lo wer stress scores was the removal of one frustrations stress construct item prior to data collection which would have affected the frustrations construct and the total stress score for each student. A one-way ANOVA was conducted to examin e significant differences of student stress levels between students taught by ad aptive faculty members, middle score faculty members and innovative faculty members. A significant difference was found with a Bonferroni post-hoc test identi fying that students enrolled in classes taught by adaptive faculty members on average had significantly high er levels of stress than students taught by other faculty members. Although this findi ng may not be practica lly significant, it confirms the results found by previous rese arch (Puccio, Talbot, & Joniak, 1993) and indicated a relationship between student st ress and courses taught by adaptive faculty members. Why does this relationship exis t with adaptive faculty members and not

PAGE 283

265 innovative faculty members as they both tend to have students with larger cognitive style gap? Puccio et al. recommended that research ers examine faculty membersÂ’ instructional discourse to identify adaptive and innovativ e components that may trigger stress. The researcher echoes this recommendation but w ould also include examining how adaptive and innovative students structure new know ledge (Kuhn, 1970). For example, KirtonÂ’s (2003) A-I theory suggests that the adaptiv e faculty member would have a narrow and detailed focus on course material, uses ru les to solve problems and does not discuss irrelevant topics. An innova tive student may find working with this adaptive faculty member stressful because he has to cope to stay focused, use rules and discuss only relevant topics in the process of learni ng. An adaptive student on the other hand, may learn with his own narrow focus, his own rule s and his own ties to relevancy which is different than the faculty memberÂ’s. Therefor e the adaptive student taught by an adaptive faculty member may have increased levels of stress as the student is still forced to structure knowledge differently than he pref ers (Kuhn, 1970). More research is needed to determine if this explanation is upheld in the undergraduate classroom. The MSLQ was used to determine level of student motivation in each course. Throughout the nine courses total motiva tion means ranged from 29.31 (Class A) to 32.42 (Class D) with no construct mean sc ore in any class exceeding one standard deviation from the norm (P intrich et al., 1991). A one-way ANOVA was conducted to examin e significant differences of student motivation scores between courses taught by the adaptive teacher group, middle score teacher group and innovative teacher group. A significant difference was found between the three groups and a Bonf erroni post-hoc test was conducted to further examine

PAGE 284

266 motivation score differences. However, the value differences may have little practical significance in the classroom. Nevertheless, the findings indi cated that these students on average were less motivated in classes ta ught by more adaptive faculty members and more innovative faculty members than when taught by faculty memb ers with a cognitive style score approximately at 95; the general population mean (Kirton, 2003). Kirton readily cites increased stress as a result of a large cognitive style gap, but refers to motivation as the affective component of coping behavior necessary to bridge the cognitive style gap when it is required to work with someone who has a different cognitive style. The evidence suggests that in these nine classes, student motivation is inhibited when taught by an adaptive or innovative faculty member which makes coping behavior seem less likely. Educational practi tioners that are highly adaptive or highly innovative should be made aware that they ha ve a different problem solving style than many of their students which may inhibit stude nt motivation. More re search is warranted to determine effective coping behaviors that faculty members can implement to increase motivation in students with a different cognitive style. Three student engagement constructs were utilized from the NSSE to determine student engagement of part icipating students. These c onstructs included academic challenge, active learning and student-faculty interaction which were summed to provide a total student engagement score. All construc t mean scores of student engagement were lower than the national averages for senior college students (Kuh et al., 2001). Although this study used freshman, sophomores and seni ors, there was no signi ficant difference in engagement scores in comparing these groups.

PAGE 285

267 A one-way ANOVA was conducted to examin e significant differences of student engagement between the adaptive teac her group, middle score teacher group and innovative teacher group. A significant di fference was found between groups (F=9.61, p=.00). A Bonferroni post-hoc test was conducted to furt her examine group differences. Students in the innovative teacher group had total engagement scores (M=51.22) significantly higher than engagement scores of students enrolled in the middle score teacher group (M=48.05, p=.00), but not signifi cantly higher than the adaptive teacher group (M=49.75). This finding indicated that students in these classes taught by innovative faculty members on average have high er engagement scores than other faculty members and significantly higher than mi ddle score faculty members. Does the extraverted tendency of innovative faculty members (Kirton, 1999) facilitate higher levels of student-faculty in teraction and student engageme nt? Does the reliable and predictable adaptive faculty member have deta iled rules (Kirton) that provide structure for student engagement? Not much can be said until more research is conducted to better determine the relationship between facu lty member cognitive style and student engagement. Again, caution is given as these fi ndings may not be prac tically significant. Objective Three The goal of objective three was to calc ulate a cognitive style gap found between faculty members and students for the purpose of exploring its rela tionship with student stress, student motivation, student engageme nt and specific demographic variables of undergraduate students. To id entify a cognitive style gap between a student and their class teacher, the KAI was used to determine the cognitive style of the student and the faculty member. Cognitive style gap was calcu lated by subtracting the faculty memberÂ’s cognitive style score from each studentÂ’s cogni tive style score. Correlations were used to

PAGE 286

268 determine the strength and direction of the relationship that cognitive style gap has with studentsÂ’ stress, motivation and engagement in each class. Chapter 4 provides a complete listing of all correlations found in this study. Correlations presented in this chapter include only cognitive style gap associat ions with student stress, motivation and engagement; the focus of this study. Objective three conclusions will be discussed after a brief presentation of the findings. Class A For Class A, the mean cognitive style gap score was 26.60 (SD=18.00, n=58). Total cognitive style gap was found to have a si gnificant moderate relationship with total stress (r=.34, p<.05). Efficiency cognitive styl e gap had moderate co rrelations with stress constructs: frustrations (r=.37, p<.05), c onflicts (r=.35, p<.05), and pressure (r=.38, p<.05). Furthermore, rules group/conformity cognitive style gap had moderate correlations with stress constructs: fru strations (r=.34, p<.05), and changes (r=.33, p<.05). Total cognitive style gap was found to ha ve a significant low relationship with total motivation (r=-.29, p<.05). Efficiency c ognitive style gap scores had moderate correlations with total motivation (r=-.33, p< .05), extrinsic motivation (r=-.36, p<.05) and self-efficacy (r=-.44, p<.05). Considering the rule/group conformity cognitive style gap, moderate correlations were found with total motivation (r=.31, p<.05) and task motivation (r=-.33, p<.05). Total cognitive style gap had a negligib le correlation with total engagement (r=.04, p>.05) signifying little associ ation between the two variables. For demographic correlations in Cla ss A, student age was found positively associated with total cognitive style gap (r=.31, p<.05). Student gender was found

PAGE 287

269 moderately related with tota l cognitive style gap (r=-.32, p< .05). Considering the number of collegiate courses taken re lated to the subject area of Class A, a correlation was found with total engagement (r=.39, p<.05). Also, college classification had a moderate association (r=.30, p<.05) with total student engagement. Class B The cognitive style gap mean between the faculty member and students in Class B was 24.43 (SD=15.05, n=46). No significant corr elations were found between cognitive style gap constructs and stress constructs. Cognitive style gap construct efficiency had a moderate correlation (r=-.46, p<.05) with total motivation, intrinsic motiv ation (r=-.30, p<.05), extrinsic motivation (r=-.36, p<.05), control of learning (r=.35, p<.05) and self-efficacy (r=-.44, p<.05). Total cognitive style gap was moderately correlated with total student engagement (r=.31, p<.05) and active learning (r=.44, p<.05). Li kewise sufficiency of originality gap was positively correlated with total engage ment (r=.33, p<.05) and active learning (r=.33, p<.05). Also rule/group conformity gap was correlated with total engagement (r=.36, p<.05), active learning (r=.48, p<.05) and stude nt-faculty interact ion (r=.32, p<.05). Demographic questions were asked of st udents regarding age, gender, number of subject area courses and colle ge classification. In Class B, studentsÂ’ age was found to have a positive association with efficien cy style gap (r=.30, p<.05). Age was correlated with intrinsic motivation (r=-.32, p<.05). Class C The total cognitive style gap mean of Class C was 17.86 (SD=14.34, n=56). Considering correlations between cognitive style gap and student stress in this class, selfimposed stress had negative moderate correla tions with total cognitive style gap (r=-.31,

PAGE 288

270 p<.05) and rule/group conformity gap (r= -.31, p<.05). This finding conflicts with previous research condu cted by Kirton (2003). Moderate correlations were found between total motivation and total cognitive style gap (r=-.34, p<.05) and efficiency gap (r=-.30, p<.05) The same was found with extrinsic motivation when comparing this c onstruct to total cognitive style gap (r=-.34, p<.05) and rule/group conformity gap (r=.31, p<.05). Test anxiet y was moderately correlated with total cognitive style gap (r= -.49, p<.05), sufficiency of originality gap (r=-.47, p<.05) and rule/group conformity gap (r=-.41, p<.05). For correlations between cogni tive style gap and student engagement, sufficiency of originality gap was correlate d with active learning (r=.31, p<.05). Examining correlations among demogra phics, gender had a moderate negative correlation with total cognitive style gap (r=.38, p<.05). Student college classification of the student was found to significantly corr elate with total cognitive style gap (r=.42, p<.05). Class D For Class D, the total cognitive style gap mean was -0.37 (SD=17.40, n=73). For correlations between cognitive style gap a nd stress, a correlation was found between efficiency gap and self-imposed stress (r=.36, p<.05). This correlati on was not consistent with previous research conducted by Kirton (2003). Only one moderate correlation was found am ong student scores of cognitive style gap and motivation; efficiency cognitive style gap was correlated with extrinsic motivation (r=-.38, p<.05). There were no sign ificant correlations among cognitive style gap and student engagement in Class D.

PAGE 289

271 For demographic information concerni ng Class D, the number of collegiate courses similar to the subject area of Cla ss D was significantly correlated to total cognitive style gap (r=-.24, p<.05). Concer ning gender, there wa s a significant low correlation (r=-.24, p<.05) with te st anxiety. College level of classification was found to have a moderately negative associa tion (r=-.32, p<.05) with total stress. Class E For Class E, the total cognitive style gap score mean was -5.59 (SD=14.95, n=31). There were no significant correlations f ound among scores of cognitive style gap and scores of student stress in the class. Cons idering motivation, moderate correlations were found between sufficiency of originality gap and motivation constructs intrinsic motivation (r=.36, p<.05) and self-efficacy (r =.38, p<.05). There were no significant correlations between scores of cognitive style and scores of student engagement in Class E. The researcher examined correlations among demographic information for Class E students. The only moderate correlati on concerning demographics was between number of college courses re lated to the subject area of Class E and student-faculty interaction (r=.48, p<.05). Class F For cognitive style gap between the student s and their faculty member instructing Class F, the mean was -12.83 (SD=16.28, n=71). No moderate correlations were found among scores of cognitive style gap and scores of stress, or scores of engagement. However, one moderate correlation was found be tween sufficiency of originality gap and extrinsic motivation (r=-.33, p<.05).

PAGE 290

272 For demographic information, no significan t correlations were found in Class F regarding these demographic variables and to tal cognitive style gap. However, age was found to be associated with extrinsic motivation, (r=-.30, p< .05). Gender was moderately correlated to total student motivation (r=.42, p<.05). Cons idering studentsÂ’ college classification, a moderate correlations was found with test anxiety (r=-.35, p<.05). Class G The calculated total cognitive style gap mean for Class G was -22.58 (SD=15.04, n=65). Only one moderate correlation existe d among constructs of cognitive style gap and constructs of student stress. Sufficiency of originality gap was correlated with selfimposed stress (r=.33, p<.05). No significant correlations were found be tween scores of cognitive style gap and scores of total motivation. However total cogni tive style gap was co rrelated with active learning (r=.30, p<.05). Student respondents were asked demogra phic questions including age, gender, number of classes taken similar to Class G and college classification. None of these demographic variables were significantly a ssociated with total cognitive style gap. Furthermore, no moderate correlations were found among demographic variables. Class H StudentÂ’s cognitive style gap mean score with the faculty member was -36.70 (SD=15.01, n=50) for Class H. Moderate corr elations were found between rule/group conformity gap and total stress (r=-.36, p<.05) frustrations (r=-.30, p<.05) and pressures (r=-.49, p<.05). This finding was congruent wi th previous research conducted by Kirton (2003).

PAGE 291

273 Examining relationships between cognitive style gap and motivation, moderate correlations were found between intrinsic motivation and to tal cognitive style gap (r=.32, p<.05), sufficiency of originality gap (r= .35, p<.05), and rule/group conformity gap (r=.30, p<.05). Total cognitive style gap was correlate d with total student engagement (r=.40, p<.05), active learning (r=.46, p<.05) and stude nt-faculty interact ion (r=.41, p<.05). Likewise, sufficiency of orig inality gap was correlated wi th total engagement (r=.33, p<.05), active learning (r=.41, p<.05) and student -faculty interaction (r=.31, p<.05). Also rule/group conformity gap was moderately correlated with tota l engagement (r=.32, p<.05), active learning (r=.40, p<.05) and stude nt-faculty interact ion (r=.37, p<.05). Student respondents were asked demographi c questions specific to age, gender, number of courses taken similar to the c ontent of Class H and college classification. Answers to these questions were used to find correlations with cogni tive style gap. None of these demographic variables were found to significantly correlate with total cognitive style gap in Class H. Class I The total cognitive style gap mean for Class I was -41.10 (SD=14.25, n=60). One moderate correlation was found between efficien cy cognitive style gap and conflict stress (r=.35, p<.05); a finding that c onflicts with research presen ted by Kirton (2003). No other cognitive style gap constructs were corre lated with stress constructs, motivation constructs or student engagement constructs in Class I. For demographic information, gender was found to have a moderate negative relationship with intrinsic motivation (r=.36, p<.05) and control of learning (r=-.31,

PAGE 292

274 p<.05). Considering the number of collegiate co urses related to the subject area of Class I, a moderate correlation was found with academic challenge (r=.44, p<.05). All Students The mean cognitive style gap for all participants was -6.30 (SD=28.19, n=511). Examining correlations between cognitive style gap and student stress, no moderate correlations were found among total scores and construct scores for these students. There were no moderate correlations found between sc ales of cognitive styl e gap and scales of student motivation. Furthermore, no moderate correlations were found between scales of cognitive style gap and stude nt engagement. Although there were relationships between the aforementioned scales among individual cl asses, the researcher concluded that accounting for all participants, measures of c ognitive style gap had li ttle association with stress, motivation and engagement in these students. This finding will be discussed further in the discussion for objective 3. For a ll students, no moderate correlations existed among demographic variables and the measur ement of cognitive style gap, stress, motivation and engagement. Group Comparisons Kirton (2003) theorized and provided evid ence that individuals with more than a 20-point cognitive style gap differ in the manner in which they solve problems. As the two individuals perceive the pr oblem differently, they also di ffer in how they generate solutions, create change and uti lize rules in their problem solv ing efforts. Stress is a result of the two individuals working together in solving a problem (Kirton). A two-tailed independent sample t-test was conducted to examine stress score differences between students with less than a 20-point gap and st udents with more than a 20-point gap with their respective instructor. No significant difference was found (t=-.05, p=.96) in total

PAGE 293

275 stress scores between students with mo re than a 20-point gap (M=51.76, n=284) and students with less than a 20-point gap (M=51.82, n=212). The same data analysis was conducted to examine differences among studentsÂ’ level of motivation. A two-tailed independent sample t-test was conducted to examine motivation score differences between stude nts with more than a 20-point gap and students with less than a 20-point gap. A significant difference was found (t=-3.13, p=.00) between motivation scores of students with more than a 20-point gap (M=30.09) and students with less than a 20-point gap (M=31.20). Realizing that objective 2 findings included the discovery that courses taught by innovative instructors and adap tive instructors have student s with significantly lower levels of motivation than students in course s taught by instructors with a cognitive style score similar to the general population mean. A two-way ANOVA was conducted to examine the interaction effect between c ognitive style of the faculty member and cognitive style gap higher than 20 points on motivation. A significant difference was found (F=3.94, p=.03) for the intera ction of students with more than a 20-point cognitive style gap and faculty membersÂ’ cognitive style. The lowest motivation scores were found in students with a 20-point gap taught by adaptive faculty members (M=29.24), which compared to a mean score of 31.32 for student s enrolled in the sa me courses with a cognitive style gap less than 20 points. A t-test was conducted to examine diffe rences of total student engagement between students with less than a 20-point gap with their instructor and students with more than a 20-point gap with their inst ructor. No significan t difference was found.

PAGE 294

276 Discussion for objective three As expected, faculty members who were either highly ad aptive or highly innovative had more students with cognitive st yle gap scores higher than twenty points. For example, Class E had the least number of students having a c ognitive style gap of more than 20 points (n=9, 28.1%), while Class I had the most students having a cognitive style gap of more than 20 points (n=55, 91.7%). Naturally, cognitive style gap mean scores were the most adaptive in Class I (M =-41.10) and the most innovative in Class A (M=26.60). Note that every class had students with more than a 20-point cognitive style gap. A two-tailed independent sample t-test wa s conducted to examine if stress scores differed between students with less than a 20point cognitive style gap and students with more than a 20-point cognitive style gap with their respective instructor. No significant difference was found (t=-.05, p=.96) between th e two groups. The data suggests that students in these classes with a 20-point gap with their faculty member do not have higher levels of stress, which does not corrobo rate previous research conducted by Kirton (2003). This finding is surprising, but there ma y be two explanations. Kirton claims that coping outside of oneÂ’s preferred cognitive style is stressful and can only occur for a limited intensity and duration. However, colleg e students may walk into a classroom with full expectations to cope outside of their preferred style as required to get their desired grade. Also note that these st udents were attending one of the top universities in Florida, if not the nation, suggesting higher admission st andards and students w ith higher levels of intelligence. Kirton (1994) suggests that higher intelligence may attribute to higher levels of coping behavior. That is, these students ha ve learned to cope a nd are proficient in coping outside of their preferred style. Future research should exam ine if students with a

PAGE 295

277 cognitive style gap larger than 20 points ha s more stress than students with a cognitive style gap less than 20 points in student popula tions with lower levels of intelligence. The second explanation to why students w ith a cognitive style gap larger than 20 points did not have higher levels of stress ma y be that they did not work enough with the faculty member to recognize a need to operate outside of their preferred cognitive style. Again, coping behavior can only occur for a limited intensity and duration before the situation is too psychologically stressful (K irton). A three credit hour class presented over a time span of one semester simply may not allow the intensity or duration for students to have higher levels of stress. Second, faculty members may be interacting with students, but not participating with students in the pr oblem solving assignments. Said differently, faculty members assign problem sets, but do not provide feedback through the problem solving process used to complete the assignm ents. If students are working by themselves in completing problem solving assignments, cognitive style gap with the faculty member may be irrelevant. If students are worki ng in groups or teams, it would be more appropriate to assess the cogniti ve style gap between students in the team, not the faculty member. More research is needed to determ ine if students with a cognitive style gap larger than 20 points contributes to student stress in different academic atmospheres in which students and faculty members work toge ther in solving problems and in shorter time periods. Examples of thes e situations include smalle r class sizes, independent studies, service learning proj ects, undergraduate research programs and three hour courses offered in interim semesters. A two-tailed independent sample t-te st was conducted to examine motivation score differences between students with more than a 20-point gap and students with less

PAGE 296

278 than a 20-point gap. A significant diffe rence was found (t=-3.13, p=.00) between motivation scores of students having more than a 20-point gap a nd students having less than a 20-point gap. In further analysis a two-way ANOVA was conducted to examine the interaction effect of students with a 20-point gap and students enrolled in a course taught by an adaptive faculty member. With motivation as the dependent variables, a significant interaction effect was found (F=3.94, p=.03). The lowest motivation scores on average were found in students with a 20-point gap taught by adaptive faculty members. Note that there may be little practical signifi cance as group score differences vary little compared to the size of the motivation scal e. Still, Kirton (2003) mentions that motivation is required to employ coping behavi or, and that the benefits of achieving a goal must offset the stress resulting from c ognitive style gap. Kirton has not specifically claimed that a reduction of mo tivation is related to cognitive style gap, but the finding makes sense. Students may be less motivated to learn from a faculty member that does not communicate in their preferred cognitive style and be even less motivated to learn from an adaptive faculty member that differs in cognitive style. The data does not suggest that a reduction of student motiv ation is a precursor to an in crease of student stress, but this may be possible. This would correspond wi th the findings of Pu ccio et al. (1993) in that students in more adaptive courses not only are associated with hi gher levels of stress, but also lower levels of motivation. Further research is required to determine if the relationship between cognitive style gap and motivation. Considering the importance of motivation in increasing student engagement (Astin, 1984; Paris & Turner, 1994) this research has great implications for incr easing student engagement in undergraduate

PAGE 297

279 classrooms in colleges of agriculture. This re search would bring great er understanding to how students are affected by coping behavior. Another t-test was conducted to examine differences of total student engagement between students with less than a 20-point gap with their instructor and students with more than a 20-point gap with their inst ructor. No significan t difference was found (t=1.14, p=.25) between the two groups. This finding provides evidence that just as all students have the ability to solve all problem s (Kirton, 2003), all st udents can be engaged in learning despite the fact that they are mo re adaptive or innovative than their faculty member. Educational practitioners should be aware that having a cognitive style gap larger than 20 points with a student has no relationship with student engagement. Examining the correlations between cognitive style gap, student stress, motivation, engagement and demographic variab les helped identify re lationships between variables for further analysis in objective 4 and 5. Each class in this study had similar findings with some classes and different fi ndings with other classes. In the end, the researcher found no consistent pattern in relati onships among individual classes or groups of classes. That is, every class was uni que which was evident in objective 4 and 5. Considering all students, cognitive styl e gap had no moderate correlations with student stress, student motiva tion or student engagement, i ndicating little association between a dissimilar cognitive style and stress, motivation and engagement. This finding may be deceiving. Using the variables cognitive style gap and stress as an example, notice that a positive correlation indicated an innovative cognitive style gap was associated with higher levels of stress in a course taught by an adaptive faculty member. On the other hand, a negative correlation indi cated an adaptive cognitive style gap was

PAGE 298

280 associated with higher levels of stress in a course taught by an innovative faculty member. Both positions are th eoretically correct (Kirton, 2003) but require a different direction of relationship. Therefore concer ning cognitive style gap, positive and negative correlations from individual classes may ha ve canceled out when combined as a total group. Objective Four The purpose of objective four was to e xplain undergraduate student motivation and student stress based on constructs of cognitive style gap and selected demographic variables which included stude nt age, gender, number of similar courses taken, and college classification. Backward stepwise mu ltiple regression was used in each class to explain the variance of the de pendent variable, stress. Sign ificant findings would provide evidence that the cognitive style gap between faculty member and st udent contributed to student stress in the nine classes examined in the study. Then, backward stepwise multiple regression was used in each class to explain the variance of the dependent variable motivation. Significan t findings would provide eviden ce that the cognitive style gap between faculty member and student cont ributed to explaining levels of student motivation in the classes studied. Finally, all students were grouped together for similar analyses. Objective four conclusions will be discussed after a brief presentation of the findings. Class A In Class A, student stress was explai ned by efficiency cognitive style gap ( =.33) and studentsÂ’ age ( =.31) using a backward stepwise regression analysis. The model had an adjusted R2 of .20 indicating that 20% of the vari ance of studentsÂ’ perceived stress in Class A was contributed by th ese two variables (p<.05).

PAGE 299

281 For motivation in Class A, two independent variables were best found to explain student motivation. Cognitive style gap of efficiency ( =-.37) and number of similar courses ( =.17) contributed to explaining 13% of the variance in student motivation. The fitted model had an adjusted R2 of .13. Class B Backward stepwise multiple regression was used to explain student stress of Class B. No significant model was found to expl ain student stress given the independent variables of cognitive style gap, gender, age, number of similar classes taken and college classification. Backward stepwise multiple regression was used to explain student motivation of Class B to find the best fitting model. One variable, efficiency c ognitive style gap, was found to explain 19% of the va riance in student motivation ( =-.46, Adjusted R2=.19). Class C For Class C, backward stepwise regre ssion was used to explain total student stress. The best fitting model left three va riables including sufficiency of originality cognitive style gap ( =-.31), gender ( =-.21) and age ( =.22). The fitted model had an adjusted R2 of .13 signifying that 13% of the varian ce of stress in Class C was from these three variables (p<.05). Also in Class C, total student motivat ion was regressed with the independent variables of efficiency style gap ( =-.18), gender ( =.33) and age ( =.20). The fitted model had an adjusted R2 of .14 signifying that 14% of th e variance of stress in Class C was from these three variables (p<.05).

PAGE 300

282 Class D With respect to Class D, backward st epwise multiple regression was used to explain total student stress. The best fitti ng model included the variables efficiency cognitive style gap ( =-.15) and the student demogr aphic variable of college classification ( =-.29). This model had an adjusted R2 of .10. That is, 10% of the variance of stress in Class D was explaine d by these two variables (p<.05). Backward stepwise multiple regression was again used in Class D to explain total student motivation as a result of cognitive style gap and demographic variables. No model provided a significant expl anation of student motivation. Class E In Class E, backward stepwise multiple regression was utilized to regress total student stress with cognitive style gap betw een the student and th e faculty member as well as demographic variables including gende r, age, similar number of courses taken and college classification. However, no m odel was found statisti cally significant. For student motivation, sufficiency of originality cognitive style gap ( =.44) and gender ( =.27) contributed to a model to explai n student motivation in Class E. The model had an adjusted R2 of .13 indicating that 13% of the variance of student total motivation was explained by these two variables. Class F Backward stepwise multiple regression was used in Class F to explain the dependent variable of total student stress with cognitive style gap constructs and demographic variables including age, gender, number of similar courses taken and college classification. However, no models us ing these independent variables were found significant in ability to explain student stress.

PAGE 301

283 Two variables, college classification ( =.-.19) and gender ( =.26) contributed to the explanation of student motivation in Class F. The model had an adjusted R2 of .06 indicating that 6% of the vari ance of total motivation was attributed to studentsÂ’ college classification and gender. Class G In Class G, backward stepwise multiple regression was used to explain studentsÂ’ total stress using variables of cognitive st yle gap and student de mographics including age, gender, number of similar courses ta ken, and college classification. However, no model was found to significantly explain student stress with these i ndependent variables. Total student motivation was regressed w ith the same independent variables to find the best fitting model. Again, no signi ficant model was found to explain student motivation in Class G. Class H In Class H, two independent variables explained the dependent variable student stress using a backward st epwise regression analysis They include rules/group conformity cognitive style gap ( =-.36) and number of similar courses taken by the student ( =.21). The model had an adjusted R2 of .13 indicating that 13% of the variance of stress in Class H was from these two variables (p<.05). For motivation of Class H, backward stepwise multiple regression was used to find the best fitting model with independent variables including cognitive style gap constructs and studentsÂ’ gende r, age, number of simila r courses taken and college classification. However no models were found significant with these independent variables.

PAGE 302

284 Class I Regarding Class I, Two variables, efficiency style gap ( =.29) and age ( =.22) accounted for 10% of the variance in expl aining total student stress (Adjusted R2=.10, p<.05). Considering total student motivation of Cl ass I, the same independent variables were used in backward stepwise regression da ta analysis to find the best fitting model with the most explanation of student mo tivation. Rule/group conformity style gap ( =.28) and gender ( =-.20) were both consider ed significant contributo rs to the model. The adjusted R2=.08 indicating that 8% of the variance of total student motivation was explained by these two i ndependent variables. All Students All students participating in this study were grouped together to explain student stress based on cognitive style gap and student demographic va riables. Considering all of the students, the best fitting model to expl ain total stress included the independent variables sufficiency of originality cognitive style gap ( =.15), college classification ( =.11) and number of similar courses taken ( =.09). However, the adjusted R2 was only .03 Considering total motivation of all stude nts participating in this study, backward stepwise multiple regression was used to fi nd the best fitting model with the most explanation of the dependent variable using the independent variables cognitive style gap and student demographics. Th ree independent variables we re found to be significant (p<.05) in explaining student motivation. They include efficiency cognitive style gap ( =.08), gender ( =.12) and number of similar courses taken by the student ( =.10). However, the adjusted R2 was .03 for the model indicating that only 3% of the variance of motivation was attributed to the three independent variables.

PAGE 303

285 Discussion for objective four This study used backward stepwise multip le regression to examine if cognitive style gap contributed to th e explanation of student stress and student motivation. Considering the explanation of student st ress in adaptive courses, Class A provided evidence that students in these classes with a more innovative cogni tive style construct had increased levels of total stress. The sa me evidence was found in Class D. Note that Class D was taught by a middle score faculty member, but scored 94 for total cognitive style designating him adaptive by one point. For innovative courses, Classes H and I provided evidence that students in these cl asses with a more adaptive cognitive style construct had increased levels of stress. The data suggests that Kirt onÂ’s (2003) A-I theory was upheld in Classes A, D, H and I as a c ognitive style construct gap did explain some variance of stress in these four courses. Class C was taught by a more adaptive faculty member, but the data suggested that students having an innovative sufficiency of originality cognitive style gap in this class have lower stress scores. This finding c onflicts with KirtonÂ’s A-I theory, but why? If faculty members teach in their preferred c ognitive style, as suggested in the literature (Dunn & Dunn, 1979; Farthing & Stubbs, 2003; Witkin, 1973), this adaptive faculty member may provide structure for rewarding in novativeness in her class. Note that this class was on average the most innovativ e class (M=100.86) in the study. The mean cognitive style score of an intact group was termed cognitive style climate by Kirton (2003). A student may also have a cognitive style gap with the groupÂ’s cognitive style climate which also gives rise to coping beha vior and stress if the cognitive style climate is different than the studentÂ’s preferred cognitive style. Perh aps in Class C, students were not coping to work with the adaptive faculty member, but we re coping to work with the

PAGE 304

286 innovative students. It seems logical that the faculty member may heavily influence cognitive climate in a course, as they pr ovide instruction thro ugh their own cognitive style. However, relinquishing control of the class to stude nts provides opportunities for cognitive style climate of the students to intervene. More research is needed to first identify perceived cognitive style climat e of a course and second determine how cognitive style climate influences stude nt stress, motivation and engagement. Classes B, E, F and G found no significant explanation of student stressed based on a cognitive style gap between the student and the faculty member. There were no patterns found in demographic variables, or co nstruct scores of stre ss and cognitive style gap to indicate that these courses were diffe rent than classes that did have significant relationships between cognitive style gap and student stress. Furthermore, Classes E and F had faculty members with cogni tive style scores similar to the cognitive climate of the class, so one could not argue that cognitive styl e climate was a factor in these two classes. Considering that in objective 3, findings indi cated that students with at least a 20 point cognitive style gap did not have higher levels of stress than students with less than a 20 point gap, it is not surprising th at some classes did not find th at increasing cognitive style gap contributed to higher stress levels. For student motivation, Classes A, B, and C each found that students with an innovative efficiency cognitive style gap had lower total motivation scores in these classes taught by an adaptive faculty memb er. The amount of variance of student motivation explained in these models ranged from 13% to 19%. This finding coincides with results of objective 3 indicating that students taught by adaptive faculty members and have more than a 20 point cognitive style gap also have lower le vels of motivation.

PAGE 305

287 Also note that the efficiency construct style gap was the contributing construct in all three classes. This indicates that the narrow fo cus to which these adaptive faculty members approach problems is principal to decreasing student motivation. Class I was taught by an innovative faculty member and evidence from the data suggests that students with larger adaptiv e rules/group conformity gaps have higher levels of motivation. A possible explanation may be that these more adaptive students were exhibiting coping behavi or (Kirton, 2003). This however does not coincide with previous findings in objective 3 which indicated that students with a cognitive style gap larger than 20 points on average had lower leve ls of motivation. If true, Class I marks the only class in this study which students exhibited coping behavior. There was no indication that this class wa s different than the other ei ght classes with regard to demographic variables or other group categoriz ations. More research is needed to fully examine the relationship between cognitive style gap and student motivation and how student coping behavior may be fostered. In Classes D, F, G and H, a cognitive style construct did no t contribute to a significant model in explaining student leve l of motivation. Furthermore, findings in Class E suggest that students with an innovati ve sufficiency of orig inality cognitive style gap have increased levels of motivation. Howe ver, the faculty member instructing Class E had a cognitive style score equal to th e general population mean (Kirton, 2003). To state it again, more research is needed to fully examine the relationship between cognitive style gap and student motivation and how st udent coping behavior may be fostered. Objective Five Objective five was achieved by using independent variables of cognitive style gap constructs, total stress, total motivation, a nd student demographic variables to explain

PAGE 306

288 student engagement in each of the nine clas ses. Demographic variables included studentsÂ’ age, gender, number of similar courses taken and college classification. Backward stepwise multiple regression was employed for analysis of the data to provide evidence of support for the theoretical framework of this study. Said differently, data analysis will examine if student engagement in the classroom can be explained by cognitive style construct gaps, stress an d motivation. Finally, all students who participated in the study were grouped together to further explain stud ent engagement. Objective five conclusions will be discussed after a brief presentation of the findings. Class A For Class A, backward stepwise regression was used to explain student engagement with independent variables of c ognitive style gap cons tructs, total stress, total motivation and student demographic. The best fitting model with the most explanation of student engagement used five independent variable s including sufficiency of originality cognitive style gap ( =-.19), total motivation ( =.25), total stress ( =.39), number of similar courses ( =.26) and college classification ( =.30). The model had an adjusted R2 of .34 signifying that 34% of the varian ce of student engagement in the class was from these five variables (p<.05). Class B Considering Class B, backward stepwise regression was used to explain student engagement with independent variables of c ognitive style gap cons tructs, total stress, total motivation and student demographic. The best fitting model with the most explanation of the dependent variable left two variables including rules/group conformity cognitive style gap ( =.33) and total motivation ( =.38). The model had an adjusted R2 of

PAGE 307

289 .20. This signifies that 20% of the variance of student engagement in the class was from these two variables (p<.05). Class C In order to explain student engagement of Class C, backward stepwise multiple regression was employed to find the best fitting model utilizing the independent constructs of cognitive style gap and variables total stress, total motivation and student demographics. Data analysis found that suffi ciency of originality cognitive style gap ( =.20), total motivation ( =.45) and age ( =.22) best contributed to total engagement of Class C. The model was statistically significant (p<.05) and the adjusted R2 was .24. Class D To explain student engagement of Class D, backward stepwise multiple regression was used to determine the independe nt variables that best fit into a model. Data analysis found that only total motivation ( =.38) best explained student engagement of Class D. The model was signifi cant (p<.05) and the adjusted R2 was .13, indicating that the independent variable contributed 13% of the variance in explaining student engagement of Class D. Class E Backward stepwise multiple regre ssion was employed to explain student engagement in Class E given the independent variables of cognitive style gap constructs, total stress, total motivation, and selected student demographic variables. After data analysis, the best fitting model was made usi ng three variables: number of similar courses taken by the student ( =.40), total stress ( =.27) and age ( =.45). The adjusted R2 of the model was .26 signifying that 26% of the vari ance was explained by the three variables.

PAGE 308

290 Class F Regarding Class F, student engagement was explained using backward stepwise multiple regression. Independent variables included students cognitive style gap constructs, total stress, total motivation and student demographic variables. However, no model was found significant in explaining stud ent engagement given these independent variables. Class G Student engagement of students in Cl ass G was explained through the use of backward stepwise multiple regression to identify the independent variables that serve as the best fitting model. These independent variables included cognitive style gap constructs, total stress, total motivation and se lected student demographic variables. After data analysis, total stress ( =.25) and total motivation ( =.38) were identified to best explain student engagement of Class G. The adjusted R2 for the model was .22 signifying that 22% of the variance of the dependent variab le was attributed to studentsÂ’ total stress and total motivation. The model was significant (p<.05). Class H For Class H, student engagement was explained by using backward stepwise multiple regression to identify independent va riables that best contribute to a fitting model. Four variables were identified that contributed to the explanation of Class H student engagement. Those independent vari ables include rule/group conformity style gap ( =.40) number of similar courses previously taken by the student ( =-.26), total motivation ( =.47) and total stress ( =.30). The model was signifi cant (p<.05) and had an adjusted R2 of .33. That is, 33% of the variance of student engagement in Class H was explained by the four independent variables previously listed.

PAGE 309

291 Class I Student engagement of Class I was e xplained by employing backward stepwise multiple regression and finding the best fitti ng model with the most explanation of the dependent variable. Two independent variables were identified as a best fit for explaining student engagement of Class I. They include total motivation ( =.37), and number of similar courses take n by the student ( =.27). The adjusted R2 was .19 meaning 19% of the variance of student engagement in Class I can be explained by the five independent variables. All Students All of the students who participated in th is study were grouped together to explain student engagement based on the independent variables of cognitive style gap constructs, total stress, total motivation and student demographics. Demographic information consisted of studentsÂ’ age, gender, numb er of similar courses taken and college classification. Student engagement in this group was explained by the variables gender ( =-.09), total motivation ( =.30), total stress ( =.14) and student age ( =.09) using a backward stepwise regression analysis. The model had an adjusted R2 of .13 meaning that 13% of the variance of student engagement am ong students in this study can be explained by the four independent student variables a bove. The model was found to be significant (p<.05). Discussion for objective five For student engagement, only four of th e nine classes (A, B, C and H) had a cognitive style gap construct si gnificantly explain student en gagement. In Class A and H, the data suggests that a larger cognitive st yle gap that opposes the cognitive style of the faculty member was detrimental to student engagement scores. In Classes B and C the

PAGE 310

292 opposite was found as positive relationships indi cated that students with an innovative cognitive style construct gap have higher leve ls of engagement in these adaptively taught classes. These findings are inconclusive in determining the relationship between student engagement and cognitive st yle gap; this mimics the results of Smith, Sekar and Townsend (2002) in that every study that supports the contention that information presented in oneÂ’s preferred learning style in creased learning, there is a study indicating that learning from a dissimilar learning style does not affect learning. Should there be a relationship between cognitive style gap a nd student engagement? Given the rising attention given to increasing student engage ment in undergraduate classrooms (Acharya, 2002; Lunde, Baker, Buelow & Hayes, 1995; Study Group on the Conditions of Excellence in American Higher Education, 1984), this question deserves to be answered. Further research is necessary to determine if dissimilar cognitive styles contribute to explaining student engagement. The highest level of variance in stud ent engagement explained among these models was 34%; some of which included four independent variables. If the researcher were to suggest what major variables account for the remaining 66% of the variance in student engagement, the most critical would be instructional methods used by the faculty member (Astin, 1993; Chickering and Gameson, 1987; Kuh, 2003; Whittington, 1998). It may be that how the faculty member em ploys an instructional method emphasizes components of their cognitive style more so th an other instructiona l methods (Puccio et al., 1993). The researcher calls for qualitat ive research to identify cognitive style components within a faculty memberÂ’s inst ructional discourse and how this affects

PAGE 311

293 student engagement. However, th ere may be no implications from this research as Jensen (2000) has found learning style to be irrelevant in the complex process of learning. Concluding Discussion The most important finding in this study was that in these classe s, cognitive style gap was detrimental to motivation in course s taught by adaptive faculty members; even more detrimental if studentÂ’s cognitive style gap with the faculty member was larger than 20 points. Realizing the necessity of student motivation to foster student engagement (Astin, 1984), this relationship must be fu rther researched. Anothe r important finding was that in these undergraduate courses, students with a 20-point cognitive style gap or higher did not have significantly higher levels of stress than students with less than a 20-point gap with their faculty member. If this relations hip is confirmed with further research, this finding limits the application of KirtonÂ’s (2003) adaption-innovation theory in the undergraduate classroom. Maybe the undergradu ate classroom environment is unique in that students are more conforming to the expectations of the faculty member; more so than the group dynamics of managing change for which the KAI was developed. If the undergraduate classroom is a different envir onment, the threshold for cognitive style gap to cause stress in undergraduate students may be higher than 20 points. Also, the undergraduate classroom environment may not offer enough individual interaction between the faculty member and each student to constitute working together to solve a problem. This study provides evidence that dissimilar cognitive styles between students and their faculty member may contribute to stude nt stress, student motivation and student engagement. However, results were inconclu sive to why some classes found cognitive style gap contributing to these dependent variables and in other classes cognitive style

PAGE 312

294 gap was not a significant cont ributor. Said again, for every class that found cognitive style gap to be significant in explaining stress, motivation and engagement, there was a class that did not find cognitive style gap significant (Smith, Sekar & Townsend, 2002). The answer to why findings were inconsis tent among classes was beyond the scope of this study. In spite of that, this study did have limitations that may have been a factor in the inconsistent findings. The KAI experienced a high mortality rate with 29% of the self-reported cognitive style scores of students unusable for analysis. That is, 29% of the students participating in this study provided suspect scores which likely did not give a true measure of studentsÂ’ preferred cognitive style. The mortality rate of the KAI among college students was anticipated as Kirton (1999) estimated a 20% mortality rate for this type of population. Provided that students who are pressured or have higher levels of stress tend to have suspect scores (Kirton), there may be valuab le information lost among participants who inappropriately completed the KAI. If research concerning cognitive style is to continue in undergraduate classroom, the mortality issue of the KAI must be addressed. This study had a 9% higher mortality rate than Kirton (1999) estimated; this may be explained by examining both the instrument administrator and the population. There is evidence that individuals inexperienced in administering the KAI also have responses with higher mortality and lo wer internal consistency among the instrument items (Kirton). This may be the case as the rese archer had only been certified as a KAI practitioner for approximately one year befo re data collection for this study and had limited experience.

PAGE 313

295 The population of students used in this study also may have contributed to the higher mortality rate. The University of Flor ida is a leading research institution in the United States with high admission standards for undergraduate students. Accordingly, students attending the University of Florida have higher levels of academic skills and are expected to exhibit those skills. These hi gh expectations may have pressured students into answering most KAI items as easy or very easy to complete a task which may not be accurate. In doing this, a student does not repo rt their preferred cognitive style, but a mixture of both cognitive style and cognitive level (Kirton). This is also known as the halo effect, which is characterized by a respondent answering items in a way that promotes the respondent as better than his or her realistic self (A ry, Jacobs & Razavieh, 2002). All KAI scores that were found to be suspect of not accurately determining cognitive style were removed from the data (Kirton, 1999). The efficiency construct of the KAI ha d a post-hoc CronbachÂ’s alpha of .66 in determining its reliability. Yet, the efficien cy construct was found as the most prominent cognitive style construct gap in explaining st ress and motivation. Th e literature has found the KAI as a reliable and valid instrument (Kirton, 1999), but use of the KAI may have limitations with college students considering th e high mortality rate and low reliability of the efficiency construct. The researcher ca lls for more development of the KAI for the purpose of studying undergraduate students. Th is development should include qualitative research to fully understand the relationship of variables in the undergraduate classroom. Another limitation of this study was low pa rticipant response in some classes. Non-response error limited the findings of this study as not every student in these classes completed all four instruments. Although 72% of all participants re sponded overall, there

PAGE 314

296 was no way to determine the reasons for non -response. For example, students may not have responded because of higher stress le vels provoked by the faculty member teaching the course. Particularly, non-response error ma y be higher in Class E given that only 46% of the students responded. Furthermore, cons idering the mortality rate of the KAI only 30% of the Class E population was examined in this study. There was no data collected concerning non-responders; thus non-response error in Class E was a limitation to this study. Whittington (1998) found that faculty memb ers in a college of agriculture predominately teach at lower level thinking sk ills. The researcher took great effort in selecting courses that re quired the use of problem solving assignments which encompassed higher level thinking skills. Howe ver, this study found that 60.7% (n=432) had no assignments assigned duri ng a typical week that had taken more than one hour to complete. This finding indicated that problem solving assignments did exist in these courses, but not to the ex tent anticipated by the res earcher. If problem solving assignments that utilize higher level thinking skills are rare in college of agriculture undergraduate courses (Whittington) there may be limitations to KirtonÂ’s A-I theory in these classrooms. That is, if problem solv ing learning (Gagne, 1965) was not present in the classroom, differences in problem solv ing style may not be a salient cognitive difference between faculty member and student This may explain why some classes did not find cognitive style gap to be a factor of stress, motivation and engagement. Another limitation may be that participants were from larger class sizes. Class size may be a variable to consider in the a pplication of A-I theory to the undergraduate classroom as smaller classes tend to increase student engagement with the faculty

PAGE 315

297 member (McKeachie, Chism, Menges, Svinic ke & Weinstein, 1994) and A-I theory requires the interaction between student and faculty member in the problem solving process (Kirton, 2003). Simply said, there ma y not have been enough interaction between faculty members and students in some of these classes for the realization of cognitive style gap. The student participants examined in th is study were representative of college students in the College of Agricultural and Life Sciences at the University of Florida, however these college students may not be ty pical of the general population identified by Kirton (2003). The University of Florida has high admission standards that not all students are able to meet. The cognitive level (intelligence) of these students are higher, allowing them to learn how to cope (operate in another cognitive st yle) as the situation requires. It is possible that these students were able to recognize the faculty memberÂ’s expectations and meet those expectations with ease, more so than other individuals. Said differently, student engagement in these cl asses could be better explained by cognitive level instead of cogn itive style (Kirton, 2003). Instrument measurement error may have also contributed to the limitations of this study. In data analysis, only total scores of stress, motivation and engagement were used, each with acceptable post-hoc reliability (Sch mitt, 1996). However, the measure of stress used in this study could be questioned. This study used a perceived stress instrument which was limited in ability to measure stre ss specifically to academics. Measures of perceived stress have been found to inadverten tly measure chronic stre ss, or other sources of stress (Cohen, Kamarck & Mermelsein, 1983) The researcher recommends that more research be conducted to accurately m easure undergraduate academic stress.

PAGE 316

298 To close, learning style theories can be traced to the observational work conducted by Jung (1960) who recognized indi vidual difference in learning. To this day, most psychologists agree that individuals each have a preferred style of learning. However, in the field of learning styles, th ere are an abundance of instruments with many questions of validity and reliability (Coffield et al., 2004). The next line of research to study learning styles should us e qualitative methods to bri ng greater understanding to how dissimilar cognitive styles interact in the context of learning. This research has potential to provide practitione rs with sound instructional pr actices that foster student engagement and provide opportunitie s for all students to learn. Recommendations The following recommendations were ma de based on the conclusions of this study which examined nine intact classes. Caution should be used in applying c onclusions and recommendations beyond the populations of this study as cl asses were not chosen randomly. Greater understanding is needed to why these students have more stress in adaptive courses. Examination of how adaptive facu lty members use instructional discourse may facilitate this research. Also, exam ining how students learn from adaptive faculty members versus innovative faculty members may show promise. Research indicating a relationship betw een adaptive faculty members and higher levels of student stress were limited to intact groups; both in this study and previous studies. More research is needed to confirm this relationship. Based on the findings of this study, adap tive faculty members in colleges of agriculture should be made aw are that their cognitive styl e may incite higher stress levels in their undergraduate students and should be provided in -service to learn coping behaviors that allow the faculty member to operate in a more innovative behavior. High adaptive and high innovative faculty members should be made aware that their cognitive style naturall y gives rise to large cognitiv e style gaps with students. These large cognitive style gaps may be detrimental to student motivation. Inservice should be provided to instruct faculty members the proper coping behaviors to operate at cognitive styles closer to that of students.

PAGE 317

299 Much has been accomplished to create more student centered classrooms. However, students in these nine agricultura l classes were still comparably low in levels of student engagement. Recognizi ng the need to facil itate higher order thinking skills and problem solving inst ructional methods in the undergraduate classroom, more in-service is needed to he lp faculty members become proficient in facilitating higher level thinking. More research is needed to determin e how adaptive and innovative faculty members in the college of agriculture us e instructional discourse differently to facilitate higher levels of student engagement. This may provide greater understanding to improving student engagement. In this study (n=511), student participants with a cognitive style gap higher than 20 points did not have higher levels of stre ss than students with less than a 20 point gap with their respective faculty member Future research should examine if student populations with lower levels of student intelligence have the same finding. More research is needed to determine if students with a cognitive style gap larger than 20 points contributes to student stre ss in different academic atmospheres in which students and faculty members work together in solving problems and in shorter time periods. Examples of these situations include sm aller class sizes, independent studies, servi ce learning, undergraduate research programs and three hour courses offered in interim semesters. This study found that these students on av erage had lower motivation scores in courses taught by adaptive faculty memb ers and when cognitive style gap was higher than 20 points with the instructing faculty member. There was also an interaction effect between the two variables upon examin ing these nine courses. Further research is warran ted to understand the rela tionship of student coping behavior and student motivation. Emphasis sh ould be placed on KirtonÂ’s efficiency construct. Based on these findings, faculty members s hould be in-serviced with instructional techniques that help stud ents increase motivation despite the faculty memberÂ’s cognitive style. More research is needed to first identify perceived c ognitive style climate of a course and second determine how cognitive style climate influences student stress, motivation and engagement. Based on this study, educational practitione rs should be made aware that having a dissimilar cognitive style greater than 20 points does not affect student engagement. More research is needed to fully exam ine the relationship between cognitive style gap and student motivation and how stude nt coping behavior may be fostered.

PAGE 318

300 Further research is necessary to determine if dissimilar cognitive styles contribute to explaining student enga gement. Given the rising atte ntion given to increasing student engagement in undergraduate clas srooms, this question deserves to be answered. The researcher calls for qualitative resear ch to identify cognitive style components within a faculty memberÂ’s instructional discourse and how a cognitive style gap affects student engagement. Qualitative research is needed to fu lly understand the relationships between variables that may inhibit or facilitate student engageme nt with regard to adaptioninnovation theory. The mortality of the KAI was high when administered to undergraduate students and must be addressed to facilitate the measurement of problem solving cognitive style to better study adaption-innovation th eory in the contex t of learning and improving student engagement. Summary Chapter 5 summarizes the entire stud y. More details concerning the problem statement, limitations, assumptions, significance of the study and defi nitions can be found in chapter 1. Chapter 2 presents theoretical and empirical literature pertinent to the study. The methodology including details of the inst rumentation, population, and data collection procedures can be found in chapter 3. Data from each class were fully presented and organized by objective in chapter 4. This chap ter closes with discus sion of findings from the data and provides implic ations and recommendations.

PAGE 319

301 APPENDIX A INTERNAL REVIEW BOARD SUBMISSION FORM 1. TITLE OF PROTOCOL: Effects of Cognitive Style on Undergraduate Student Engagement, Stress, and Motivation. 2. PRINCIPAL INVESTIGATOR(s): Curt Friedel PO Box 110540 310 Rolfs Hall, Ph: 352-392-1663 cfriedel@ufl.edu Fax: 352-392-9585 3. SUPERVISOR (IF PI IS STUDENT): Dr. Rick Rudd PO Box 110540 218 Rolfs Hall, Ph: 352-392-0502 x230, rrudd@ufl.edu Fax: 352-392-9585 4. DATES OF PROP OSED PROTOCOL: From February 1, 2006 To July 30, 2006 5. SOURCE OF FUNDING FOR THE PROTOCOL: No outside funding will be used for this re search. All funding is internal from the Department of Agricultural Education and Communication. 6. SCIENTIFIC PURPOSE OF THE INVESTIGATION: To determine the effects of dissimilar cognitive styles on student engagement, motivation, and stress of undergra duate students enrolled at th e University of Florida in respect to demographic information. UFIRB #2006-U-78

PAGE 320

302 7. DESCRIBE THE RESE ARCH METHODOLOGY IN NON-TECHNICAL LANGUAGE. Faculty in the College of Agriculture and Life Sciences at the University of Florida will be administered the Kirton Adaption-Innova tion Inventory (KAI) (measure of cognitive style). Faculty members will be determined as a participant based on their KAI score. Once a sample of faculty members is dete rmined, students in th e faculty membersÂ’ respected courses will be recruite d as participants for this study. Students at the University of Florida will be administered the Kirton Adaption-Innovation Inventory (KAI) (measure of cognitive style) National Survey of Student Engagement (NSSE) (measure of student engagement ), Motivated Strategies for Learning Questionnaire (MSLQ) (measure of motivati on and use of learning strategies), and Student-Life Stress Inventory (SSI) (measure of stress). The instruments will be packaged together prior to the study to maintain anonymity and confidentiality of the responses while still allowing to look for re lationships between specific student responses between instruments. All instruments are attached. 8. POTENTIAL BENEFITS AND ANTICIPATED RISK. There are no direct benefits to the participants themselves. There may be benefits for developing future curriculum and courses to promote the recognition and use of cognitive style effects in the classroom. There ar e no known risks associated with this study. 9. DESCRIBE HOW PARTICIPANT(S) WILL BE RECRUITED, THE NUMBER AND AGE OF THE PARTICIPANTS, A ND PROPOSED COMPENSATION (if any): Faculty members will be recruited as instru ctors based on their KAI scores. A maximum of 20 faculty members will be recruited and all participation will be voluntary. Age of faculty participants is 30 to 65 years of age. No compensation will be provided to faculty members. Students will be recruited from faculty memb ersÂ’ respective classes at the College of Agriculture and Life Sciences at the Univer sity of Florida. A maximum of 900 students will be recruited and all partic ipation will be voluntary. Age of student participants is 18 years to 60 years of age, with the majority of students from 18 to 22 years of age. No compensation will be provided to students. However, some students may or may not receive extra credit points for participation in this study. The allocation and amount of extra credit points will be determ ined by the course instructor.

PAGE 321

303 10. DESCRIBE THE INFORMED CONSENT PROCESS. INCLUDE A COPY OF THE INFORMED CONSENT DOCUM ENT (if applicable). Please use attachments sparingly. See attached form. __________________________ Principal Investigator's Signature _________________________ Supervisor's Signature I approve this protocol for submission to the UFIRB: ______________________ _________ Dept. Chair/Center Director Date

PAGE 322

304 APPENDIX B FACULTY INFORMED CONSENT Informed Consent (Faculty) Protocol Title : Effects of Cognitive Style on Undergraduate Student Engagement, Stress and Motivation Purpose of the research study : To determine the effects of dissimilar cognitive styles on student engagement, motivation, and stress of undergraduate students enrolled at the University of Florida in respect to demographic information. What will you be asked to do in the study : Please complete the Kirton Adaption-Innovation Inventory (KAI). Time required : Approximately 15 minutes. Risks : There are no known risks as sociated with this study. Benefits/Compensation : There is no compensation, monetary or otherwise, for participation in this study. There are no direct benefits fo r participating in this study. Confidentiality : Your identity will be kept confidential to the extent provided by the law. Your responses to the instruments will be anonymous. Voluntary participation : Your participation is completely vol untary. There is no penalty for not participating. You can stop at any time without penalty and you do not have to answer any question you do not want to answer. Whom to contact if you have questions about the study : Curt Friedel Department of Agricultural Education and Communication PO Box 110540 310 Rolfs Hall, Ph: 352-392-1663 Fax: 352-392-9585 cfriedel@ufl.edu Dr. Rick Rudd Department of Agricultural Education and Communication PO Box 110540 218 Rolfs Hall, Ph: 352-392-0502 x230 Fax: 352-392-9585 rrudd@ufl.edu UFIRB #2006-U-78

PAGE 323

305 Whom to contact about your rights in the study : UF IRB Office PO Box 112250 University of Florida Gainesville, FL 32611-2250 Phone: 352-392-0433 Agreement : I have read the procedure described above. I voluntarily agree to participate in the procedure and have received a copy of this description. Participant: ____________________________________ Date: _____________ Principal Investigator: ____________________________ Date: _____________

PAGE 324

306 APPENDIX C STUDENT INFORMED CONSENT Informed Consent (Student) Protocol Title : Effects of Cognitive Style on Undergraduate Student Engagement, Stress and Motivation Purpose of the research study : To determine the effects of dissimilar cognitive styles on student engagement, motivation, and stress of undergraduate students enrolled at the University of Florida in respect to demographic information. What will you be asked to do in the study : Please complete the Kirton Adaption-Innovation Inventory (KAI), National Survey of Student Engagement (NSSE), Motivated Strategies for Learning Questionnaire (MSLQ), and Student-Life Stress Inventory (SSI). Time required : Approximately 45 minutes. Risks : There are no known risks as sociated with this study. Benefits/Compensation : You may or may not receive compensation in the form of extra credit points in this course for participating in this study. Th e allocation and amount of ex tra credit points will be determined by the course instructor. Other than allo cation of extra credit points, there is no compensation, monetary or otherwise for participation in this study. There are no direct benefits for participating in this study. Confidentiality : Your identity will be kept confidential to the extent provided by the law. Your responses to the instruments will be anonymous. Voluntary participation : Your participation is completely vol untary. There is no penalty for not participating. You can stop at any time without penalty and you do not have to answer any question you do not want to answer. Whom to contact if you have questions about the study : Curt Friedel Department of Agricultural Education and Communication PO Box 110540 310 Rolfs Hall, Ph: 352-392-1663 Fax: 352-392-9585 cfriedel@ufl.edu UFIRB #2006-U-78

PAGE 325

307 Dr. Rick Rudd Department of Agricultural Education and Communication PO Box 110540 218 Rolfs Hall, Ph: 352-392-0502 x230 Fax: 352-392-9585 rrudd@ufl.edu Whom to contact about your rights in the study : UF IRB Office PO Box 112250 University of Florida Gainesville, FL 32611-2250 Phone: 352-392-0433 Agreement : I have read the procedure described above. I voluntarily agree to participate in the procedure and have received a copy of this description. Participant: ____________________________________ Date: _____________ Principal Investigator: ____________________________ Date: _____________

PAGE 326

308 APPENDIX D EXAMPLE ITEMS OF KIRTON’S ADAPTION-INNOVATION INVENTORY Directions: Mark an “X” to signify how easy or difficult do you find it to present yourself, consistently, over a long period as: Easy Hard 1. A person who likes to solve problems inductively . . . . . . . . . . . . . . . . . 2. A person who likes to solve problems deductively . . . . . . . . . . . . . . . . . The Kirton’s Adaption-Innovation Inventory (K AI) is a copyrighted questionnaire and used with permission. For information regarding the KAI, please contact: KAI Distribution Centre 55 Heronsgate Rd Chorleywood Hertfordshire WD3 5BA UK Telephone: 01923 286999 (From USA: 01144-192-328-6999) Fax: 0870 0527901 (From USA: 01144-870-052-7901) E-mail: dist@kaicentre.com

PAGE 327

309 APPENDIX E EXAMPLE ITEMS OF STUDENT-LIFE STRESS INVENTORY Directions: Respond to each statement using this 5-point scale which indicates the level of your experiences in this course. 1 = Never, 2 = Seldom, 3 = Occasionally, 4 = Often, 5 = Most of the time. N S Oc Of M As a student in this course: 1. I have experienced stress in reaching my goal. 1 2 3 4 5 2. I have found that problems in setting goals. 1 2 3 4 5 3. I have been deficient in resources 1 2 3 4 5 The Student-life Stress Inventory (SSI) is a copyrighted questionna ire and used with permission. For information regarding the SSI, please contact: Dr. Bernadette Gadzella Department of Psychology and Special Education Texas A & M University – Commerce Commerce, TX 75429 Phone: (903) 886-5588 E-mail: Bernadette_Gadzella@tamu-commerce.edu

PAGE 328

310 APPENDIX F EXAMPLE ITEMS OF MOTIVATED STRATEGIES FOR LEARNING QUESTIONNAIRE Directions: In answering the following questions, thin k about your motivation for this course you are currently taking. Using the scale below, please answer the following questions. Please remember to answer accurately as there are no wrong or right answers. 1. In a class like this, I pref er challenging coursework. Not at all true of me 1 2 3 4 5 6 7 Very true of me 2. Studying aptly will help me succeed in learning course material. Not at all true of me 1 2 3 4 5 6 7 Very true of me 3. When taking an exam, I worry about doing poorly. Not at all true of me 1 2 3 4 5 6 7 Very true of me The Motivated Strategies for Learning Qu estionnaire (MSLQ) is a copyrighted questionnaire and used with permission. Fo r information regarding the MSLQ, please contact: Marie Bien The University of Michigan 610 E. University Ave.; Room 1323 Ann Arbor, MI 48109-1259 E-mail: mabien@umich.edu

PAGE 329

311 APPENDIX G EXAMPLE ITEMS OF NATIONAL SU RVEY OF STUDENT ENGAGEMENT Directions: In your experience in this c ourse, how often have you done each of the following? Mark your answers in the boxes below. Very often Often Sometimes Never Part 1 a. Participated by asking questions b. Worked with other students in class c. Discussed assignment grades with teacher or other students The National Survey of Student Engagement (NSSE) is a copyrighted questionnaire and used with permission. For informati on regarding the NSSE, please contact: National Survey of Student Engagement Center for Postsecondary Research Indiana University Bloomington 1900 East Tenth Street Eigenmann Hall Suite 419 Bloomington, IN 47406-7512 Phone: (812) 856-5824 Fax: (812) 856-5150 E-mail: nsse@indiana.edu

PAGE 330

312 LIST OF REFERENCES Abouserie, R. (1994). Sources and levels of stress in relation to locus of control and self-esteem in university students. Educational Psychology 14, 323330. Acharya, C. (2002, September). StudentsÂ’ le arning styles and their implications for teachers. CDTL Brief 5(6), 1-3. Ajzen, I. (2002). Percieved behavioral cont rol, self-efficacy, locus of control, and the theory of planned behavior. Journal of Applied Social Psychology 32, 1-20. Allinson, C., & Hayes, J. (1996). The cognitive style index. Journal of Management Studies 33, 119-135. Apter, M. J. (2001). Motivational styles in everyday life: A guide to reversal theory Washington DC: American Psychological Association. Ary, D., Jacobs, L. C., & Razavieh, A. (2002). Introduction to research in education (6th Ed.). Belmont, CA: Wadsworth/Thomson Learning. Aspinwall, L., & Taylor, S. (1992). Mode ling cognitive adaptation: A longitudinal investigation of the impact of indi vidual differences and coping on college adjustment and performance. Journal of Personality and Social Psychology 63(6), 989-1003. Astin, A. W. (1984). Student involvement: A developmental theory for higher education. Journal of College Student Development, 25, 297-308. Astin, A. W. (1993). What matters in college? Four critical years revisited San Francisco: Jossey-Bass. Astin, A. W. (1996). Involvement in learni ng revisited: Lessons we have learned. Journal of College Student Development 37 (2), 123-134. Atkinson, J. W. (1957). Motivational dete rminants of risk-taking behavior. Psychological Review 64, 359-372. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavior change. Psychological Review 84, 191-215.

PAGE 331

313 Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1994). Self-efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of human behavior (Vol. 4, pp. 71-81) New York: Academic Press. Barnhart, K. R. (2002). The application of Kirton adaption-innovation theory to the flight instruction envi ronment of higher education. Unpublished dissertation. Indiana State Un iversity, Terre Haute, IN. Battle, E. (1965). Motivational determinan ts of academic task persistence. Journal of Personality and Social Psychology 2, 209-218. Boekaerts, M. (1999). Self-regulate d learning: Where we are today. International Journal of Educational Research 31, 445-457. Boyd, B. L., & Murphrey, T. P. (2004) Evaluating the scope of learning instruments used in studies publishe d in the journal of agricultural education. Journal of Southern Agricultural Education Research 54(1), 124-133. Breen, R., & Lindsay, R. (2002). Differe nt disciplines require different motivations for student success. Research in Higher Education 43(6), 693-725. Cassidy, S. (2004). Learning styles: An overview of theories, models, and measures. Educational Psychology 24(4), 419-444. Chickering, A. W., & Gameson, Z. F. ( 1987). Seven principles for good practice in undergraduate education, AAHE Bulletin 39(7), 3-7. Choi, N. (2005). Self-efficacy and self-concep t as predictors of college studentsÂ’ academic performance. Psychology in the Schools, 42(2), 197-205. Clark-Thayer, S. (1987). The relationship of the knowledge of student-perceived learning style preferences, and study ha bits and attitudes, to achievement of college freshmen in a small urba n university. (Doctoral dissertation, Boston University, 1987). Dissertation Abstracts International 48, 872A. Clark-Thayer, S. (1988). Designing studyskills programs based on individual learning styles. Learning-Style Network Newsletter 9(3), 4. New York: St. JohnÂ’s University and the National Association of Secondary School Principals. Claxton, S., & Ralston, Y. (1978). Learning styles: Their impact on teaching and administration. ASHE-ERIC/Higher Education Report, No. 10, Association for the study of Highe r Education: Washington D. C.

PAGE 332

314 Coffield, F., Moseley, D., Hall, E., & Ecclestone K. (2004). Learning styles and pedagogy in post-16 learning: A systematic and critical review London: Learning and Skills Research Centre. Cohen, S., Kamarck, T., & Mermelstei n, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior 24(4), 385-396. Collins, M. A., & Amabile, T. M. (1999). Motivation and creativity, In Sternberg, R. J. (Ed.). Handbook of creativity. (pp. 297-312). New York: Cambridge University Press. Curry, L. (1987). Integrating concepts of cogniti ve or learning style: A review with attention to psychometric standards. Ottawa, ON: Canadian College of Health Service Executives. Curry, L. (1990). A critique of research on learning styles. Educational Leadership 48(2), 50-56. Cuseo, J. (2004). The empirical case against large class size: Adverse effects on the teaching, learning, and rete ntion of first-year students. Marymount College Retrieved November 19, 2005 at www.brevard.edu/listserv/rem arks/cuseoclasssize.pdf. Deci, E. L. (1975). Intrinsic motivation. New York: Plenum. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum. de Ciantis, S. M., & Kirton, M. J. (1 996). A psychometric re-examination of kolbÂ’s experiential learni ng cycle construct: A seperation of level, style and process. Journal of Educational and Psychological Measurement 56(5), 809-820. Delbecq, A. L., & Mills, P. K. (1985) Managerial practices that enhance innovation. Organizational Dynamics 14, 24-34. DeTure, M. (2004). Cognitive style and self-efficacy: Predicting student success in online distance education. American Journal of Distance Education 18(1), 21-38. Dewey, J. (1938). Experience and education New York: Kappa Delta Pi. DiBartola, L. M., Miller, M. K., & Turl ey, C. L. (2001). Do learning style and learning environment affect learning outcome? Journal of Allied Health 30(2), 112-115. Dunn, R. (1984). Learning style: State of the science. Theory into Practice 23(1), 10-19.

PAGE 333

315 Dunn, R. S., & Dunn, K. J. (1979). Lear ning styles/teaching styles: should they...can they...be matched? Educational Leadership, 238-244. Dunn, R., Dunn, K., & Price, G. E. (1979). Productivity environmental preference survey Lawrence, KS: Price Systems, Inc. Dyer, J. E. & Osborne, E. (1996). Effects of teaching approach on achievement of agricultural educati on students with varying learning styles. Journal of Agricultural Education 37(3), 43-51. Eccles, J. (1983). Expectancies, values and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motives (pp. 75-146). San Francisco: Freeman. Eccles, J., & Wigfield, A. (1995). In the mind of the actor: The structure of adolescentsÂ’ achievement task valu es and expectancy-related beliefs. Personality and Social Psychology Bulletin 21, 215-225. Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology 53, 109-132. Ehrenberg, R. G., Brewer, D. J., Gamoran, A., & Willms, D. J. (2001). Class size and student achievement. Psychological Science in the Public Interest, 2(1), 1-30. Entwistle, N. (1978). Identifying di stinctive approaches to studying. Higher Education 8, 365-380. Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C., (1993). The role of deliberate practice in acquisit ion of expert performance. Psychological Review 100(3), 363-406. Farthing, D. W., & Stubbs, G. (2003). Learning to learn: A report on a longitudinal study of the learning styles of computing undergraduates. NUI Galway, UK: LTSN Centre for In formation and Computer Sciences. 168-173. Feather, N. (1988). Values, valences, and course enrollment: Testing the role of personal values within an expectancy-value framework. Journal of Educational Psychology 80, 381-391. Flavell, J. (1963). The developmental psychology of Jean Piaget. New York: Van Nostrand Rienhold Co. Fox, J. & Barthol omae, S. (1999). Student learning style and educational outcomes: Evidence from a family financial management course. Financial Services Review 8, 235-251.

PAGE 334

316 Ford, N., & Chen, S. Y. (2001). Matching/ mismatching revisited: an empirical study of learning and teaching styles. British Journal of Educational Technology 32(1), 5-22. Fraser, B. J. (1998). Classroom environm ent instruments: Development, validity and applications. Learning Environments Research 1, 7-33. Gagne, R. M. (1965). The conditions of learning. New York: Holt, Rinehart and Winston, Inc. Gadzella, B. M., (1994). Student-life stre ss inventory: Identification of and reactions to stressors. Psychological Reports 74, 395-402. Gadzella, B. M., & Baloglu, M. (2001). Conf irmatory factor analysis and internal consistency of the student -life stress inventory. Journal of Instructional Psychology 28, 84-94. Gadzella, B. M., & Fullwood, H. L. (1992) Differences among university student age groups on their perceptions of stress. Proceedings of the 95th annual meeting of the Texas Academy of Science. Midwestern State University, Wichita Falls, TX, 176-180. Gadzella, B. M., & Guthrie, D. (1993 ). Analysis of a stress inventory. Proceedings of the 96th annual meeting of the Texas Academy of Science University of North Texas, Denten, TX, 301-308. Gadzella, B. M., & Masten, W. G. (2005). An analysis of th e categories in the student-life stress inventory. American Journal of Psychological Research 1(1), 1-10. Gadzella, B. M., Masten, W. G., & Stacks, J. (1998). StudentsÂ’ stress and their learning strategies, test anxiety, and attributions. College Student Journal, 32, 416-422. Garton, B. L. (1993). The relationship between agri cultural teachersÂ’ learning style and problem-solving ab ility and the extent of use of the problemsolving approach to teaching. Unpublished thesis. The Ohio State University, Columbus, OH. Garton, B. L. (1999). Learning styles, teaching performance, and student achievement: A relational study. Journal of Agricultural Education 40(3), 11-20. Gellin, A. (2003). The effect of undergra duate student involvement on critical thinking: A meta-analysis of the li terature, 1991-2000. Journal of College Student Development, 44, 746-762.

PAGE 335

317 Germain, M., & Scandura, T. A. (2005). Grade inflation and student individual differences as systematic bias in faculty evaluations Journal of Instructional Psychology 32(1), 58-67. Goldsmith, R. E. (1994). Creative style and personality theory. In M. J. Kirton, M. (Rev. Ed.). Adaptors and innovators: Styl es of creativity and problemsolving. (pp. 34-50). New York: Routledge. Gregorc, A. F. (1979). Learning/teaching styles: Their nature and effects. In Student learning styles: diagnos ing and prescribing programs. Reston, VA: National Association of Secondary Principles, 19-26. Guilford, J. P. (1975). Creativity: A quarter century of progress. In I. A.Taylor & J. W. Getzels (Eds.), Perspectives in creativity 37-59. Chicago: Aldine. Hargreaves, D. (2005). About learning: Report of the learning working group London: Demos. Hembree, R. (1988). Correlates, causes, e ffects, and treatment of test anxiety. Review of Educational Research 58(1), 47-77. Hennessey, B. A., & Amabile, T. M. ( 1993). The conditions of creativity, In Sternberg, R. J. The nature of creativity: Contemporary psychological perspectives (pp. 11-38). Cambridge University Press. Hennessey, B. A., & Zbikowski, S. ( 1993). Immunizing children against the negative effects of reward: A further examination of intrinsic motivation training techniques. Creativity Research Journal 6, 297-308. Herrmann, N. (1989). The creative brain North Carolina: Brain Books, The Ned Herrmann Group. Hill, K. G. (1991). An ecological approach to creativity and motivation: Trait and environmental influences in the college classroom. (Doctoral dissertation, Brandeis University). University Microfilms International 9118702. Hockey, R. (1979). Stress and the cognitiv e components of skilled performance. In V. Hamilton, & D. Warbutin (Eds.), Human stress and cognition (pp. 141-177). New York: John Wiley & Sons. Honey, P., & Mumford, A. (1992). Manual for the learning style questionnaire. Maidenhead, Berkshire: Peter Honey. Honeyman, M. S., & Miller, G. S. (1998) The effect of teaching approaches on achievement and satisfaction of fiel d-dependent and field independent learners in animal science. Journal of Animal Science 76, 1710-1715.

PAGE 336

318 Jakubowski, T. G., & Dembo, M. H. (2002, April). Social cognitive factors associated with the academic self -regulation of undergraduate college students in a learning and study strategies course. Paper presented at the annual meeting of the American Educational Research Association, New Orleans, LA. James, W. B., & Gardner, D. L. (1995). L earning styles: Implic ations for distance learning. New Directions for Adul t and Continuing Education 67, 19-32. Jensen, E. (2004). Brain-based learning: The new science of teaching and training. San Diego: The Brain Store. Jones, C., Reichard, C., & Mokhtari, K. (2003). Are students’ learning styles discipline specific? Community College Journal of Research and Practice 27, 363-375. Judge, T. A., Erez, A., Bono, J. E., & T horesen, C. J. (2002). Are measures of self-esteem, neuroticism, locus of c ontrol, and generalized self-efficacy indicators of a common core construct? Journal of Personality and Social Psychology, 83(3), 693-710. Jung, C. (1960). The structure and dynamics of the psyche. New York: Bollingen Foundation. Keefe, J. W. (1960). Learning style: An overview. In NASSP’s Student learning styles: Diagnosing and prescribi ng programs (pp. 1-17). Reston, VA: National Association of S econdary School Principals. Kirton, M. J. (1976). Adaptors and I nnovators: A description and measure. Journal of Applied Psychology 61, 622-629. Kirton, M. J. (1978). Have adaptors and innovators equal levels of creativity? Psychological Reports 42, 695-698. Kirton, M. J. (1994). A theory of cognitive style. In Kirton, M. J. (Rev. Ed.). Adaptors and innovators: Styles of creativity and problem-solving. (pp. 133). New York: Routledge. Kirton, M. J. (1999). Kirton Adaption-Innovation Inventory Manual 3rd Edition. Occupational Research Centre: Berkhamsted, U.K. Kirton, M. J. (2003). Adaption – innovation: In th e context of diversity and change New York: Routledge. Knowles, K. (1999). Effects of teacher knowle dge and motivation on students Unpublished dissertation. University of Maryland, College Park, MD.

PAGE 337

319 Kolb, D. A., (1984). Experiential learning: Experien ce as the source of learning and development. Upper Saddle River, NJ: Prentice Hall, Inc. Koljatic, M., & Kuh, G. D. (2001). A longitudinal assessment of college student engagement in good practices in undergraduate education. Higher Education 42, 351-371. Kubes, M., & Spillerova, D. (1992). The e ffects of personality characteristics on communication patterns in R&D teams. Creativity and Innovation Management 1(1), 33-40. Kuh, G. D. (2001). The national survey of student engagement: Conceptual framework and overview of psychometric properties. Bloomington, IN: Indiana University Center for Postsecondary Research and Planning. Kuh, G. D. (2003). What weÂ’re learning about engagement from NSEE. Change 35(2), 24-33. Kuh, G. D., Hayek, J. C., Carini, R. M., Ouimet, J. A., Gonyea, R. M., & Kennedy, J. (2001). NSSE Technical and Norms Report. Bloomington, IN: Indiana University Center for Postsecondary Research and Planning. Kuhn, T. S. (1970). The structure of scientific revolutions. 2nd Edition. Chicago, IL: University of Chicago Press. Larkin-Hein, T. (2000, August). Learning styles in introductory physics: Enhancing student motivation, interest, & learning. Paper presented at the meeting of the International Conf erence on Engineering and Computer Education, Sao Paulo, Brazil. Laurillard, D. (1979). The pr ocess of student learning. Higher Education 8, 395409. Lewin, K. (1951). Field theory in the social sciences New York: Harper & Row. Linnenbrink, E. A., & Pintrich, P. R. ( 2003). The role of self-efficacy beliefs in student engagement and l earning in the classroom. Reading and Writing Quarterly 19, 119-137. Lunde, J. P., Baker, M., Buelow, F. H., & Hayes, L. S. (1995). Reshaping curricula: Revitalizing programs at three land grant universities. Bolton, MA: Anker Publishing Company, Inc. Marton, F., Hounsell, D., & Entwistle, N. (1984). The experience of learning Edinburgh, Great Britain: Scot tish Academic Press. Marton, F., & Slj, R. (1976). On qualitati ve differences in learning: I-outcome and process. British Journal of Educational Psychology 46, 4-11.

PAGE 338

320 Marzano, R. J., Pickering, D ., & Pollock, J. E. (2001). Classroom instruction that works:Research-based strategies for increasing student achievement. Alexandria, VA: Association fo r Supervision and Curriculum Development. McCarthy, R. M. (1988). An investigation of role conf lict and coping behavior in women managers. Unpublished thesis. Universi ty of Hertforshire, UK. McKeachie, W. J., Chism, N., Menges, R. Svinicki, M., & Weinstein, C. E. (1994). Teaching tips: Strategies, research, and theory for college and university teachers, Lexington, MA: D. C. Heath and Company. McKeachie, W. J., Pintrich, P. R., Lin, Y., & Smith, D. A. (1986). Teaching and learning in the college classroom: A re view of the research literature. Technical Report No. 86-B-001.1. A nn Arbor, MI: University of Michigan. Messick, S. (1984). The nature of cogn itive styles: Problems and promise in educational practice. Educational Psychologist 19(2), 59-74. Miller, L. E. (1998). Appropriate analysis. Journal of Agricultural Education (39)2, 1-10. Misra, R., & McKean, M. (2000). Colleg e studentsÂ’ academic stress and its relation to their anxiety, time ma nagement, and leisure satisfaction. American Journal of Health Studies 16(1), 41-51. Morris, C. G. (1990). Contemporary psychology and effective behavior. (7th Ed.). Glenview, IL: Scott & Foresman. Nickerson, R. S. (1999). Enhancing cr eativity, In Sternberg, R. J. (Ed.). Handbook of creativity. (pp. 393-430). New York: Cambridge University Press. Pace, C. (1980). Measuring the quality of student effort. Current Issues in Higher Education 2, 10-16. Paris, S. G., & Turner, J. C. (1994). S ituated Motivation. In Pintrich, P. R., Brown, D. R. & Weinstein, C. E. (Eds.). Student motivation cognition, and learning: Essays in honor of Wilbert J. McKeachie. (pp.213-237). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. Pascarella, E. T. (1980). Student-faculty informal contact and college outcomes. Review of Educational Research 50, 545-595. Pascarella, E. T., & Terenzini, P. T. ( 1978). Student-faculty informal relationships and freshman year educational outcomes. Journal of Educational Research 71, 183-189.

PAGE 339

321 Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students. San Francisco: Jossey-Bass Publishers. Pask, G. (1976). Styles a nd strategies of learning. British Journal of Educational Psychology 46, 128-148. Pask, G., & Scott, B. C. E. (1972). Lear ning strategies and in dividual competence. International Journal of Man-Machine Studies 4, 217-253. Pfeiffer, D. (2001). Academic and environmental stress among undergraduate and graduate college students : A literature review. Unpublished thesis. University of Wisconsin-Stout, Stout, WI. Piaget, J. (1970). Genetic epistemology New York: Columbia University Press. Pike, G. R. (1999). The constant error of the halo in educational outcomes research. Research in Higher Education 40, 61-86. Pike, G. R. (2004). Measuring quality: A comparison of U.S. News rankings and NSSE benchmarks, Research in Higher Education 45(2), 193-208. Pike, G. R., Kuh, G. D., & Gonyea, R. M. (2003). The relationship between institutional mission and studentsÂ’ in volvement in educational outcomes. Research in Higher Education, 44, 243-263. Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology 82(1), 33-40. Pintrich, P. R., & Schunk, D. H. (2002). Motivation in education: Theory, research and application (2nd Ed.). Columbus, OH: Merrill Prentice Hall. Pintrich, P. R., Smith, D. A. F., Ga rcia, T., & McKeachie, W. J. (1991). A manual for the use of the motivated strategi es for learning questionnaire (MSLQ) (Technical Report No. 91-B-004). A nn Arbor, MI: University of Michigan. Price, L. (2004). Individual differences in learning: Cognitive control, cognitive style, and learning style. Educational Psychology 24(5), 681-698. Price, B., Dunn, R., & Dunn, K. (1990). Productivity environm ental preference survey: An inventory for the identific ation of individual adult preferences in a working or learning environment Lawrence, KS: Price Systems, Inc. Puccio, G. J., Talbot, R. J., & Joniak, A. J. (1993). Person-environment fit: Using commensurate scales to predict student stress. British Journal of Educational Psychology 63(3), 457-468.

PAGE 340

322 Romano, J. L. (1992). Psychoeducational in terventions for stress management and well-being. Journal of Counseling and Development 71, 199-202. Ross, S. E., Niebling, B. C., & Heckert, T. M. (1999). Sources of stress among college students. College St udent Journal, 33(2), 312-321. Rudd, R., Baker, M,, & Hoover, T. (2000) Undergraduate agriculture student learning styles and critical thinking abilities: Is there a relationship? Journal of Agricultural Education 41(3), 2-12. Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internationalization: Examining reasons for acting in two domains. Journal of Personality and Social Psychology 57, 749-761. Sadler-Smith, E. (2001). A reply to Re ynoldÂ’s critique of learning style. Management Learning 32(3), 291-304. Sandler, M. E. (2000). Career decision-mak ing self efficacy, perceived stress, and an integrated model of student persiste nce: A structural model of finances, attitudes, behavior, a nd career development. Research in Higher Education 41(5), 537-580. Schlechty, P. C. (2002). Working on the work: An action plan for teachers, principals, and superintendents San Francisco: Jossey-Bass. Schmeck, R. R., Ribich, F. D., & Ramanaiah, N. V. (1977). Development of a self-report inventory for assessing individual differences in learning processes. Applied Psychological Measurement 1, 413-431. Schmitt, N. (1996). Uses and abuses of coefficient alpha. Psychological Assessment 8(4), 350-353. Schunk, D. H. (1991). Self-efficacy and academic motivation. Educational Psychologist 26, 207-231. Schunk, D. H. (2000). Learning theories: An educational perspective. Upper Saddle River, NJ: Prentice-Hall, Inc. Shields, N. (2001). Stress, active coping, and academic performance among persisting and nonpersisting college students. Journal of Applied Biobehavioral Research, 6(2), 65-81. Shuell, T. J. (1986). Cognitive conceptions of learning. Review of Educational Research 56, 411-436. Skinner, B. F. (1953). Science and human behavior New York: Free Press.

PAGE 341

323 Smith, D. (1977). College classroom in teractions and critical thinking. Journal of Educational Psychology 69, 180-190. Smith, J. (2002). Learning styles: Fa shion fad or lever for change? The application of learning style theory to inclusive curriculum delivery. Innovations in Education and Teaching International, 39(1), 63-70. Smith, W., Sekar, S., & Townsend, K. (2002). The impact of surface and reflective teaching and learni ng on student academic success In M. Valcke & D. Gombeir (Eds.), Learning styles: re liability and validity 407-418. Proceedings of the 7th Annual European Learning Styles Information Network Conference, June 26-28, Ghent. Ghent: University of Ghent. Smulders, F. E. (2004). Co-operation in NPD: Coping with different learning styles. Creativity and Innovation Management 13(4), 263-273. Soden, R. (1994). Teaching problem solving in vocational education. New York: Routledge. Sternberg, R. J. (1988). Ment al self-government: A theory of intellectual styles and their development. Human Development, 31, 197-224. Study Group on the Conditions of Excelle nce in American Higher Education. (1984). Involvement in learning: Real izing the potential of American higher education Washington, DC: U.S. Department of Education. Swartz, R. J., & Perkins, D. N. (1990). Teaching thinking: Issues and approaches. Pacific Grove, CA: Midwest Publications. Terenzini, P., & Pascarella, E. (1980). St udent/faculty relationships and freshman year educational outcomes: A further investigation. Journal of College Student Personnel 21, 521-528. Umbach, P. D., & Wawrzynski, M. R. (2005). Faculty do matter: The role of college faculty in student learning and engagement. Research in Higher Education, 46(2), 153-184. Vermunt, J. D. (1992 ). Learning styles and directed learning processes in higher education: Towards a process-orie nted instruction in independent thinking. Lisse: Swets and Zeitlinger. Wallas, G. (1926). The art of thought New York: Franklin Watts. Walling, M. (1987). Relationship of achievement motivation to the Kirton Adaption Innovation Inventory. Unpublished thesis. East Texas State University, Commerce, TX.

PAGE 342

324 Whittington, M. S. (1998). Improving the cognitive level of college teaching: A successful faculty intervention. Journal of Agricultural Education, 39(3), 31-40. Williams, D. D., Cook, P. F., Quinn, B. & Jensen, R. P. (1985). University class size: Is smaller better? Research in Higher Education 23(3), 307-318. Wingenbach, G. J. (2000). Agriculture st udentsÂ’ computer sk ills and electronic exams. Journal of Agricultural Education 41(1), 69-78. Winne, P. H. (1995). Self-regulation is ubiquitous but its forms vary with knowledge. Educational Psychologist 30(4), 223-228. Witkin, D. B. (1973). The role of cognitive style in academic performance and in teacher-student relations. Research Bulletin Educational Testing Service, Princeton, NJ, 73-101 Witkin, H. A., & Goodenough, D. R. (1977) Field dependence and interpersonal behavior. Psychological Bulletin 84, 661-689. Witkin, H. A., Oltman, P. K., Raskin, E., & Karp, S. A. (1971). A manual for the embedded figures tests. Palo Alto, CA: Consulting Psychologists Press. Zimmerman, B. J. (2000). Self-efficacy: An essential motive to learn. Contemporary Educational Psychology 25, 82-91.

PAGE 343

325 BIOGRAPHICAL SKETCH Curtis Friedel was born in Norfolk, Nebr aska, in April of 1976. He grew up on the family farm located south of Stuart, Nebr aska. Curtis graduated from Atkinson West Holt High School in May of 1995. Curtis attended Northwest Missouri State University in Maryville, Missouri, to obtain a Bachelor of Science in agricultural education. He fulfilled his student teaching requirements at Randolph Public Schools in Randolph, Nebraska, unde r the direction of Dennis Bazata and Dr. Marvin Hoskey. He completed his degree in May of 1999 with the honor of cum laude. After the completion of his B.S. degr ee, Curtis accepted a position as an agricultural instructor at La urel-Concord Public Schools in Laurel, Nebraska. There he started an agricultura l education program after a sixt y year absence of high school agricultural instruction of any form. Afte r four years, the program had won national awards, but more importantly had created opportunities for high school students to become successful. In June, 2003, Curtis accepted a graduate student assistantship in the Department of Agricultural Education and Communication at the Univers ity of Florida to obtain a Doctor of Philosophy degree. The degree in cludes specializations in teaching and learning, leadership and volunt eer development, and a minor in food and resource economics. His work at the University of Fl orida included serving as a co-instructor of

PAGE 344

326 undergraduate and graduate courses, as we ll as conducting resear ch in creative and critical thinking. Curtis is married to Elisabeth Friede l. He remains active in community and church and many professional organizations.