Predicting academic performance of underprepared freshmen with high school GPA, ACT scores, learning styles, psychologic...

MISSING IMAGE

Material Information

Title:
Predicting academic performance of underprepared freshmen with high school GPA, ACT scores, learning styles, psychological type, and learning skills
Physical Description:
xv, 205 leaves : ; 28 cm.
Language:
English
Creator:
Moore, Robert L., 1941-
Publication Date:

Subjects

Subjects / Keywords:
Prediction of scholastic success   ( lcsh )
Academic achievement   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1985.
Bibliography:
Includes bibliographical references (leaves 199-203).
Statement of Responsibility:
by Robert L. Moore.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 14772001
ocm14772001
System ID:
AA00012888:00001

Full Text













This is an authorized facsimile, made from the microfilm
master copy of the original dissertation or master thesis
published by UMI.

The bibliographic information for this thesis is contained
in UMI's Dissertation Abstracts database, the only
central source for accessing almost every doctoral
dissertation accepted in North America since 1861.




-UMf Dissertation
Services

A Bell & Howell Company
300 N. Zeeb Road, Ann Arbor, Michigan 48106
1-800-521-0600 313-761-4700


Printed in 1997 by xerographic process
on acid-free paper



DPBT
06
UNIVERSITY OF FLORIDA LIBRARIES








8606742


Moore, Robert L.


PREDICTING ACADEMIC PERFORMANCE OF UNOEPREPARED FRESHMEN
WiTH HIGH SCHOOL GPA, ACT SCORES. LEARNING STYLES,
PSYCHOLOGICAL TYPE. AND LEARNING SKILLS


The University of Florida PH.D. 1985

University
Microfilms
International aON.ZebRoad.Ann Abr.MI 1O6

Copyright 1985

by

Moore, Robert L.

All Rights Reserved


















INFORMATION TO USERS


This reproduction was made from a copy of a manuscript sent to us for publication
and microfilming. While the most advanced technology has been used to pho-
tograph and reproduce this manuscript, the quality of the reproduction is heavily
dependent upon the quality of the material submitted. Pages in any manuscript
may have indistinct print In all cases the bestavailable copy has been filmed.

The following explanationof techniques is provided to help clarify notations which
may appear on this reproduction.

1. Manuscripts may notalways be complete. When It is not possible to obtain
missing pages. a note appears to indicate this.

2. When copyrighted materials are removed from the manuscript. a note ap-
pears to indicate this.

3. Oversize materials (maps. drawings. and charts) are photographed bysec-
tfrnlndg th nriginal hadeininn at the upper left h:::d ::::r d continu-
....... .. a -- ----- c- -
ing from left to right in equal sections with small overlaps. Each oversize
page Is also filmed as one exposure and is available, for an additional
charge. as a standard 35mm slide or in black and white paper format.

4. Most photographs reproduce acceptably on positive microfilm or micro-
fiche but lack clarity on xerographic copies made from the microfilm. For
an additional charge, all photographs are available in black and white
standard 35mm slide format."


*For more information about black and white slides or enlarged paper reproductions.
please contact the Dissertations Customer Services Department.








Bici ns
htratim rsd













PLEASE NOTE:


In all cases this material has been filmed in the best possible way from the available copy.
Problems encountered with this document have been identified here with a check mark .


1. Glossy photographs or pages
2. Colored illustrations, paper or print
3. Photographs with dark background_
4. Illustrations are poor copy
5. Pages with black marks, not original copy
6. Print shows through as there is text on both sides of page
7. Indistinct, broken or small print on several pages v
8. Print exceeds margin requirements
9. Tightly bound copy with print lost in spine

10. Computer printout pages with indistinct print
11. Page(s) lacking when material received, and not available from school or
author.

12. Page(s) seem to be missing in numbering only as text follows.
13. Two pages numbered_ Text follows.
14. Curling and wrinkled pages

15. Dissertation contains pages with print at a slant, filmed as received

16. Other





University
Microfilms
International



















PREDICTING ACADEMIC PERFORMANCE OF UNDERPREPARED
FRESHMEN WITH HIGH SCHOOL GPA, ACT SCORES,
LEARNING STYLES, PSYCHOLOGICAL TYPE,
AND LEARNING SKILLS








BY

ROBERT L. MOORE


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


1985
















In Loving Memory

OE Robert H. Moore

















ACKNOWLEDGEMENTS


The author wishes to thank the members of his doctoral

committee for their guidance and assistance in developing the

dissertation. Dr. Lee Mullally and Dr. William Hedges were

very supportive and helpful during the writing of the

dissertation, and the long-time friendship of Bill Hedges was

especially comforting. Dr. Mary McCaulley provided both warm

support and critical analysis during the writing stages,

particularly concerning the Myers-Briggs theories and

analyses. Most of all, the author appreciates the guidance

and encouragement from the chairman of the committee, Dr.

Gordon Lawrence. His gentle, yet insistent, support made the

final product possible. The author was honored to have his

mentor, Dr. Katherine Steele, read the final draft and be

present at the defense, and he wishes to have her know she

provided an impetus to finish even though living hundreds of

miles away.

The statistical analyses and computer runs would not

have been possible without the patient assistance of Rita

Berg and JoLynda Hoggard, and the author is extremely

grateful to both of them. Much of the typing was done by

Evelyn Fuller, and her long hours at the word processor are

very much appreciated. The staff at the Academic Development











Office was understanding of the author's preoccupation during

the writing of the dissertation, and they were constant

sources of ideas and encouragement. The staff of the Career

Planning and Placement Office made the author feel welcome,

and he appreciates the use of the word processor in that

office.

Throughout the long period of the research and writing

of the dissertation, the author has had the comfort and

support of three wonderful daughters. Andrea, Rebecca, and

Rochelle provided more than they realize.

For her unselfish, loving support, which never waivered

even when it should have, the author is grateful to Nancy

Noth. It will take awhile, but the weekends will oe repaid.
















TABLE OF CONTENTS


ACKNOWLEDGEMENTS ..

LIST OF TABLES .. ..

ABSTRACT .. ....

CHAPTERS


INTRODUCTION .

The Problem .

Background for the Problem

Purpose of the Study .

Assumptions and Delineations

of the Research .. ...

Definition of Terms .

Hypotheses ...

Methodology .

Data Analysis .

Summary .....

REVIEW OF THE LITERATURE .

Introduction .


Constructs of Learning Style and

Psychological Type .

Learning Styles and the Learning

Environment .

Choice of Instruments for the Research

Kolb's Learning Style Inventory .

Kolb-McCarthy Learning Styles Inventory


ONE


PAGE

. iv

. ix

. xiii


2
. L

. 1

. 2

. 11




4.. .

16

18

. 19

20

. 21

. 22

. 22










TABLE OF CONTENTS


PAGE

Research on the Learning Styles

Inventory .......... 32

Jung's and Myers' Theories

of Psychological Types ... .. 41

Myers-Briggs Type Indicator .. .45

Comparison of the LSI and MBTI 64

Summary .............. 65

THREE METHODOLOGY ............... 67

Population ......... .... 67

DVST 1012, Learning Skills 6

Learning Skills Assessment 71

Asessment of Learning Styles

and Psychological Type ....... 73

Selection of the Population to Study .76

Selection of Independent and Dependent

Variables ...... ... .76

FOUR DATA ANALYSIS ...... ..... .86

Population ........... .. 86

Success and Failure Analysis ... 92

Analyses of Additional Independent

Variables .............. 98

Analyses of Dependent Variables .08

FIVE SUMMARY ................. 160

Introduction ... ........ .160

Discussion .............. 163

Recommendations ........... 185










TABLE OF CONTENTS


APPENDICES PAGE

A ATTITUDES AND FUNCTIONS BY TYPE ..... 189

B ACCURACY AND RELEVANCE OF STYLE AND TYPE 190

C TITLES OF COURSES CHOSEN FOR ANALYSIS 196

D MEASURES OF CENTRAL TENDENCY ... .. .197

REFERENCES .. ..... .. ........ 199

BIOGRAPHICAL SKETCH .... .... .204


viii












LIST OF TABLES


TABLE PAGE

1-1 High-Risk Courses--Fine Arts 7

1-2 High-Risk Courses--Humanities ... 8

1-3 High-Risk Courses--Natural Sciences 9

1-4 High-Risk Courses--Social Studies .. 10

2-1 LSI Undergraduate Scores ... 35

2-2 Distribution of Learning Styles

for Chicago High School Students ... 37

2-3 Comparison of Learning Style Instruments 42

2-4 The Sixteen Psychological Types 47

2-5 Distribution of Type for Combined Samples of

11th and 12th Grade High School Students 49

2-6 Distribution of Preference for Combined Samples

of llth and 12th Grade High School Students 50

2-7 MBTI Compared with the Alport-Vernon-Lindsey

Study of Values, the Edwards Personality

Preference, and the Personality Research

Inventory .. .. 53

2-8 College Students' Perception of Type 63

3-1 College Distribution for Students in the Study 69

3-2 Disciplines, Departments, and Courses Chosen

for Analysis .. ... 83

3-3 Order of Entry for Independent Variables .84

4-1 Age Distribution for Students in the Study 87

4-2 ACT Scores for Students in the Study .. .89

4-3 High School Size and Rank in Class .... .90










LIST OF TABLES


TABLE PAGE

4-4 Comparison of Self-Reported and Actual HSGPA 91

4-5 Successful and Unsuccessful Attempts

for all Courses ............. .93

4-6 Successful and Unsuccessful Attempts

in the Four Core Disciplines ... 94

4-7 Successful and Unsuccessful Attempts

in Selected Departments ... 96

4-8 Successful and Unsuccessful Attempts

in Selected Courses ............ 97

4-9 Learning Styles for Students in the Study .100

4-10 Cumulative Grade Point Averages

for the Four Learning Styles .. 101

4-11 MBTI Types--Males .. 102

4-12 MBTI Types--Females ... .103

4-13 MBTI Types--Total Population ... 104

4-14 Rank Order Comparison of Predicted and Actual

Achievement. ... .107

4-15 Academic Status at the End

of the Freshman Year ........... 110

4-16 Final Parameter Estimates

For the Dependent Variable Academic STATUS L12

4-17 Final Parameter Estimates

For the Dependent Variable CUMGPA ..... 115

4-18 Final Parameter Estimates

For the FINE ARTS Discipline .. 118










LIST OF TABLES


TABLE

4-19 Final Parameter Estimates

For the HUMANITIES Discipline .

4-20 Final Parameter Estimates

For the NATURAL SCIENCES Discipline

4-21 Final Parameter Estimates

For the SOCIAL STUDIES Discipline


PAGE


S. 119



. 120



. 121


4-22 Final Parameter Estimates

For the FNAR Department .......... 123

4-23 Final Parameter Estimates

For the COMM Department .......... 124

4-24 Final Parameter Estimates

For the ENGL Department .......... 125

4-25 Final Parameter Estimates

For the HIST Department .......... 126

4-26 Final Parameter Estimates

For the BIOL Department .......... 128

4-27 Final Parameter Estimates

For the GEOL Department .......... 129

4-28 Final Parameter Estimates

For the MATH Department .......... 130

4-29 Final Parameter Estimates

For the PLSC Department .......... 131

4-30 Final Parameter Estimates

For the PSYC Department .......... 133

4-31 Final Parameter Estimates

For the SOCI Department ... .134










LIST OF TABLES


TABLE

4-32


Final Parameter Estimates

For the GEOG Department


4-33 Final Parameter Estimates


4-34

4-35

4-36

4-37

4-38

4-39

4-40

4-41

4-42

4-43

4-44

4-45

4-46


Final

Final

Final

Final

Final

Final

Final

Final

Final

Final

Final

Final

Final


Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter

Parameter


Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates

Estimates


4-47 Summary of Regressors for

4-48 Summary of Regressors for

Departments ..

4-49 Summary of Regressors for


COMM 1302

ENGL 1003

ENGL 1023

HIST 2003

BIOL 1004

GEOL 1114

MATH 1203

PLSC 2003

SOCI 2013

?NAR 1062

COMM 1313

ENGL 1013

MATH 1033

PSYC 2003


STATUS and CUC


. 135

. 137

. 138

. 139

S. 140

.. 142

. 143

S. 144

S. 145

S. 147

148

S. 149

S. 150

S. 151

S. L53

IGPA 154


Disciplines and



Courses .. ...


4-50 Summary of Independent Variables Found

as Regressors .

5-1 Reference List of Terms and Definitions .

B-1 Perception of Style and Type ..

8-2 Actual and Predicted Learning Style .

B-3 Actual and Predicted Type .. ....


PAGE


. 155

. L56


S. 157

. 167

. L91

. 192

. 194
















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




PREDICTING ACADEMIC PERFORMANCE OF UNDERPREPARED
FRESHMEN WITH HIGH SCHOOL GPA, ACT SCORES,
LEARNING STYLES, PSYCHOLOGICAL TYPE,
AND LEARNING SKILLS






By

ROBERT L. MOORE

August, 1985


Chairman: Dr. Gordon Lawrence
Major Department: Educational Leadership


The problem examined was how to better predict, for

underprepared college freshmen, academic performance in core-

curriculum coursework, freshman-year grade point average, and

the resulting academic status. The Kolb-McCarthy Learning

Style Inventory (LSI), the Myers-Briggs Type Indicator

(MBTI), and a locally-developed Learning Skills Assessment

(LSA) were used to assess 171 students in a learning skills

class. Students' high school grade point averages and ACT

scores were also available. From this group, 128 entering


ciii









freshmen without prior college experience were identified as

the study population.

There were 31 dependent variables selected for analysis:

academic status and grade point average at the end of the

freshman year; 4 core-curriculum disciplines (Fine Arts,

Humanities, Natural Sciences, and Social Studies); and 11

departments and 14 courses from those disciplines. These

dependent variables were used to determine if predictability

of academic success would be improved when the LSI, MBTI, and

LSA were added to regression equations containing high school

grade point averages and ACT scores.

It was hypothesized that high school grade point

averages and ACT scores were insufficient predictors oE

academic success and that the additional independent

variables (LSI, MBTI, and LSA) would significantly improve

predictability. High school grade point averages were found

to be predictive in 39% of the 31 regression equations, and

ACT scores were found to be predictive in 42% of the

equations. The LSI significantly improved 29%, the MBTI 48%,

and the LSA 22% of the equations. Most regressors appeared

in expected directions, but several anomalous results led to

recommendations of replication with a larger sample of

underprepared students and a sample of all entering freshmen.

Of particular note was the lack of predictive power of

reflective observation and abstract conceptualization from

the LSI and sensing and intuition from the MBTI. These

findings were inconsistent with other research on these two










instruments, suggesting that this group of underprepared

students may have differed in important ways from college

students not considered underprepared. Cautious use of the

LSI and MBTI for advising was recommended, but the LSA was

found to be inadequate for prediction.


















CHAPTER ONE
INTRODUCTION

The Problem

High school grade point average and standardized

assessment, such as the American College Test (ACT) or the

Scholastic Aptitude Test (SAT), are commonly used to predict

academic success in college, and they are often used as

screening devices when admitting, advising, and placing

students. Yet, even after high school grade point averages

and standardized assessments have been used for prediction,

screening, advising, and placement, many students fail to

achieve academic success during their first year in college.

The early college record for such students is characterized

by very low freshman grade point averages and a resulting

academic standing of probation, suspension, or withdrawal

from college.

Upon closer examination, the lack of success during the

freshman year is often attributable to poor grades in certain

"high-risk" core curriculum courses. tE one accepts the

notion that grades of "A," "B," or "C" are considered

successful completion of a course and that grades of "D,"

"F," "W" (withdraw), and "I" (incomplete) are considered

unsuccessful completion, then high-risk courses can be

defined as those with a high rate of unsuccessful completion.










2

For the purposes of the present research, a course in which

30% or more of all freshman students enrolled receive grades

of "D," "F," "W," or "I" will be considered a high-risk

course. (See Blanc, Debuhr, and Martin, 1983, for a

rationale for using a 30% rate.) Students who are not

prepared to be placed in such high-risk courses, yet who are

advised to take one or more of them during their freshman

year, may suffer low grade point averages and academic

failure.

The question to be investigated by the present research

is: How may advisors and others better predict success or

failure in high-risk core-curriculum coursework, freshman-

year grade point average, and resulting academic status by

examining factors in addition to high school grade point

average and standardized (ACT) assessment?

Background for the Problem

General

If the present national trend of tightening academic

standards and the closing of the "open door" continues,

coupled with a decreasing pool of eligible, traditional-age

college students, colleges and universities will face

increasing problems of maintaining enrollments to meet

financial obligations. Retention of students under such

conditions is becoming increasingly important, and students

who are not prepared to meet higher standards pose a special

problem. Maxwell (1979) has asserted:

(U)nderprepared students will not disappear from college
classrooms, nor can most colleges expect to restrict
admission to the best prepared--there are too few of










3

them. (I)t is clear that colleges must continue to
to offer comprehensive and intense academic support
services to their students. (p. 25)

Popular national opinion is focusing on basic skills

preparation (or, more accurately, the lack of it). Concern

is also being expressed in the academic community. In a

recent study conducted by the City University of New York

(reported by Nielsen and Polishook, 1984), 85% of the 1,269

institutions responding to a survey of every American college

and university reported that inadequate preparation of their

entering freshmen was a common problem. By contrast, only

three percent of the responding institutions did not perceive

poor academic foundations to be a serious problem. The

responding institutions further reported that basic skill

deficiencies were found in reading (28% of the freshmen

students), writing (31%), and basic mathematics (32%). It is

not surprising that 80% of the reporting institutions,

ranging from open-door, two-year colleges to the most

selective universities, stated that they offer remedial

courses in basic skills.

Academic success in college is a result of many

factors and their interactions, such as, skills preparation

(basic skills, study skills, and thinking skills), personal

and social adjustment, perception of college, attitude and

motivation, personal values, academic standards of the

institution, quality of teaching, learning style preference,

the match between teaching and learning styles, personality











type, academic advising and guidance, hours taken, hours

employed, placement in the course sequence, balance between

freedom and responsibility, environmental setting, peer and

familial influence, relationship to faculty, goal and career

orientation, financial support, and extracurricular

activities. Conversely, Maxwell (1979) has categorized

student characteristics which might lead to academic

failure in college: lack of potential, inadequate

conception of the work involved in succeeding, importance of

other activities over study, interference from psychological

problems, failure to assume responsibility for one's own

Learning, inhibition of language functions (poor reading,

writing, and speaking skills), lack of understanding of

standards for high-quality performance, selection of

inappropriate major, vagueness about long-term goals, and

selection of wrong college.

An attempt to list success or failure factors is

certain to have omissions, and the problem becomes even more

complicated when the possible interrelationships among the

many factors are considered. However, the underprepared (or

"misprepared," Maxwell, 1979, p. 3) college student will

almost certainly exhibit problems in one or more of the above

factors and/or their interrelationships. While the extent

and problematic nature of these factors will certainly vary

by individual, all are clearly important success or failure

variables and all will not be reflected by students' high

school grade point averages and ACT scores.











University of Arkansas at Fayetteville

To help set the stage for the present research, success

statistics for a selected group of University of Arkansas

at Fayetteville students were examined by the author. For

the 128 entering freshmen enrolled in a learning skills

course in the fall of 1983, only 73.4% were in good academic

standing by the end of that first term. By the end of their

freshman year, that figure had dropped to 58.6%. The

remaining 41.4% were either on academic probation (14.8%),

had been suspended for one year (6.3%), or had withdrawn from

the University (20.3%).

A large-scale study (Donnally, 1984) of University of

Arkansas freshmen has substantiated a high rate of academic

attrition. In 1981, the University began a six-year

retention study based on the entering freshman class of that

year. By Fall, 1982, only 66% of that group were still

enrolled. By Fall, 1983, 51% were still enrolled, and by

Fall, 1984, only 44% of the original group were still

enrolled. Thus, over a three-year period, an attrition rate

of 56% had reduced the original 2,628 entering freshman

students to 1,156. Further, in Fall, 1984, after three

complete academic years, only 44% of the remaining 1,156

students were classified as seniors, and seven percent of the

students still enrolled were on academic probation.

A more detailed understanding of freshman attrition is

obtained when one examines success (A, B, or C) in the core-












curriculum disciplines--Fine Arts, Humanities, Natural

Sciences, and Social Studies--at the University of Arkansas.

Tables 1-1 through 1-4, High-Risk Courses, list 53 core-

curriculum courses in the four major disciplines listed above

which posted a success rate of 70% or below for entering

freshmen (N = 2,228) during the 1983-84 academic year at

the University of Arkansas. (Only those courses in which at

least 30 entering freshmen were enrolled were selected.) It

can be seen from the tables that courses in which large

numbers of freshmen students are traditionally enrolled are

also ones which have low (70% or below) rates of success.

The overall success rates for the selected courses in the

four disciplines were found to be 65% for fine arts, 65% for

humanities, 54% for natural sciences, and 61% for social

studies.

The University of Arkansas has traditionally been a non-

selective, open-door institution. Any student with a high

school diploma (or equivalent) has been able to enter the

University regardless of his or her high school grade point

average or ACT scores. A student has been placed on academic

probation only if his or her semester grade point average

dropped below a 1.25 (Colleges of Agriculture and Arts &

Sciences) or 1.50 (Colleges of Business Administration,

Education, and Engineering). However, beginning with the

Fall, 1984 term, higher standards for placing students on

probation were put into effect. All freshmen and sophomores

are to be placed on probation if their semester grade point

average falls below a 1.75, and all juniors and seniors are












Table 1-1
High-Risk Courses--Fine Arts

Course Title

ARCH 1812 Introduction to Architecture

ART 1003 Drawing Fundamentals I

ART 1313 Drawing Fundamentals II

FNAR 1022 Theatre Arts

FNAR 1072 Visual Arts

LARC 1812 Landscape Architecture

Totals


Success Rate

70%

68

70

68

63

55

65%


Notes for Tables 1-1 through 1-4:

Tables 1-1 through 1-4 represent all courses in the four
core-curriculum disciplines for which entering freshmen
enrollment was 30 or more and for which the success rate
was 70% or less.

The first digit in the course number indicates the
level of the course, e.g., 1 = freshman, 2 = sophomore,
etc. The last digit in the course number indicates the
credit hours for the course. Thus, ARCH 1812 is a
freshman-level, two-hour course.

The titles for the courses are from the University of
Arkansas' college catalogs.

N indicates the number of entering freshmen enrolled in
each of the courses.

Success Rate indicates the percentage of the entering
freshmen who received grades of A, 8, or C in each course.

Percentage figures were rounded to the nearest whole
number, and the total percentage for the success rate is a
weighted average for all the courses listed in the
discipline.












Table 1-2
High-Risk

Course

ANTH 2023

ENGL 1003

ENGL 1013

ENGL 1102

GERM 1003

HIST 2003

HIST 2013

JOUR 1023

PHIL 1202

PHIL 2003

PHIL 2203

SPAN 1003

WCIV 1003

WCIV 1013

WLIT 1113


I


Courses--Humanities

Title

Cultural Anthropology

English Grammar

Freshman Composition 1

Reading Development

Elementary German

American History, 1492-1877

American History, 1877-present

Intro. to Mass Communications

Reflective Thinking

Introduction to Philosophy

Logic

Elementary Spanish

Inst. & Ideas of West. Man I

Inst. & Ideas of West. Man II

Intro. to World Literature

Totals


See notes on Table 1-1.


6

43

86

7



49

50





1





3I

1

2!

17:


N Success Rate

i6 65%

13 56

i0 70

78 65

64 67

10 65

04 62

69 59

30 57

82 70

11 68

34 68

69 56

69 64

39 65

39 65%












Table 1-3
High-Risk Courses--Natural Sciences

Course Title

ASTR 2003 Survey of the Universe

BIOL 1004 General Biology

BIOL 1024 Biological Concepts

BOTY 1014 General Botany

CHEM 1094 Chemistry in the Modern World

CHEM 1104 General Chemistry I

CHEM 1104 General Chemistry II

GEOL 1004 Earth Science for Teachers

GEOL 1114 General Geology

MATE 1033 Remedial Math (Algebra)

MATH 1103 Patterns in Mathematics

MATH 1203 College Algebra

MATH 1213 Plane Trigonometry

MATH 1285 Precalculus Mathematics

MATH 2043 Survey of Calculus

MATH 2053 Finite Mathematics

MATH 2555 Calculus & Analytic Geometry

PHYS 1023 Physics and Human Affairs

PHYS 1044 Physics Eor Architects I

PHYS 1054 Physics for Architects II

PHYS 2053 University Physics

ZOOL 1004 General Zoology

ZOOL 2443 Anatomy

Totals


N

62

525

54

157

48

349

216

57

268

493

106

1102

220

196

176

304

387

50

64

36

44

107

36

5057


Success Rate

50%

60

61

45

60

60

59

60

64

33

60

49

50

65

66

67

62

56

56

61

59

49

56

54%


See notes on Table 1-1.












Table 1-4
High-Risk Courses--Social Studies

Course Title

ECON 2013 Principles of Macroeconomics

GEOG 1003 Physical Geography

GEOG 1123 Human Geography

PLSC 1503 Intro. to Political Science

PLSC 2003 American National Government

PLSC 2203 State and Local Government

PSYC 2003 General Psychology

SOCI 2013 General Sociology

SOCI 2033 Social Problems

Totals


N

159

94

243

100

501

59

663

541

109

2469


Success Rate

49%

44

63

64

67

54

58

66

63

61%


See notes on Table 1-1.










11

to be placed on probation whenever their semester grade point

average falls below a 2.00, although no student will be

placed on probation if his or her cumulative grade point

average is 2.00 or higher.

Further, beginning with the Fall, 1985 term, students

with less than a 2.00 high school grade point average are not

to be admitted to the University, and two classes of

admission are to be enforced: (1) Students with a high

school grade point average below 2.50 and an ACT

composite score below 18 will be admitted conditionally.

(2) Students with a high school grade point average of 2.50

or above or an ACT composite score of 18 or above will be

admitted unconditionally. Students admitted

conditionally must meet rigid academic requirements by the

end of their freshman year or they will not be allowed to

continue at the University. Entering freshmen will also be

expected to have earned high school credits in four years of

English, two years of mathematics, two years of natural

sciences, and two years of social studies. The high school

requirements will become even more rigorous in 1986 and again

in 1988, and any deficiencies in core-curriculum areas must

be cleared by students during their freshman year in college.

Purpose of the Study

The research was designed to determine if success

predictability (at the five levels explained below) for

underprepared entering freshmen is significantly enhanced by

combining additional selected assessment instruments with

high school grade point average and ACT scores in regression










12

formulas. Three additional assessment tools were examined to

determine if they significantly enhanced academic success

predictability:

1. Students' learning style as measured by the Kolb-

McCarthy Learning Style Inventory (LSI);

2. Students' psychological type as measured by the

Myers-Briggs Type Indicator (MBTI); and

3. Students' reading rate and comprehension, listening

comprehension, notetaking skills, and study habits and

attitudes as measured by the Learning Skills Assessment

(LSA). The LSA is locally developed, and its use in the

present research is to be considered exploratory since

verification of the instrument's validity and reliability had

not been undertaken prior to the research.

The five major levels examined in the regression

equations form a hierarchy ranging from the most general to

the most specific academic outcomes and yield 31 dependent

variables:

1. Academic status at the end of the freshman year;

2. Freshman-year grade point average;

3. Success in the four core-curriculum disciplines

(Fine Arts, Humanities, Natural Sciences, and Social

Studies);

4. Success in 11 selected departments within the core-

curriculum disciplines; and

5. Success in 14 selected courses within the

departments.










13

It was expected that the research would show that high

school grade point average and ACT scores are necessary but

insufficient predictors of success at all levels of the

hierarchy, but particularly in the high-risk core courses

required of freshmen. The expected findings, then, should

suggest that additional assessment is necessary to better

advise, place, and remediate students before serious academic

problems begin.

The research was an early stage in the development Later

of a process model for freshmen, particularly those admitted

conditionally, which involves the student, his or her

advisor, instructors, and the academic support staff:

The student will, through assessment and feedback,

learn more about his or her capabilities and limitations.

Rather than a passive acceptor of recommendations for college

work, the student will become an active, knowledgeable

partner in the initial advising experience.

The advisor will be able to more accurately place a

student in remedial, regular, or advanced sections of core

coursework. He or she will oe better prepared to guide

students to academic support services when needed and will

better facilitate the crucial freshman year for them.

The instructors will be provided with more precise

individual and group data with which to improve their

instructional methodology. By alerting instructors to

characteristics which signal success or failure, they will be












more sensitive to early indications of trouble that help

alert them to the need for appropriate action.

The academic support staff will be better able to

program for the needs of all entering students, not just

those flagged by admissions standards, and a closer

realization of student retention will be achieved.

The University has authorized the establishment of an

Academic Development Office (ADO) to aid conditionally-

admitted students and others in academic difficulty. The

author has been asked to design and direct the ADO, and the

findings of the present research will be used to help design

the assessment, feedback, advising, remediation, and

monitoring components of that office.

Assumptions and Delineations of the Research

At the time the study was initiated, an intact group of

conditionally-admitted and course-deficient students (as

defined by the new admissions standards described above) did

not exist at the University. However, as will be explained

more fully in a later section, a group of students was chosen

which, for the most part, were considered to be underprepared

for college. There was not an intent to choose a

representative sample of the the entire entering freshman

class, since an underlying assumption of the research was

that mispredicting academic success is particularly impactful

on underprepared students. However, while the research and

the findings are specific to the study population and the

University disciplines, departments, and courses under

investigation, the implications of the findings will be












useful in other settings. To the extent that other

institutions are under pressure to selectively admit students

and to administer scarce support resources, the findings will

suggest similar attention to early warning signs of possible

academic difficulty, particularly for underprepared students,

and should assist them also in building cost-effective models

of selection, assessment, feedback, advising, placement, and

supportive services. Background and demographic

characteristics of age, sex, high school size, rank in high

school, and the initial college selected, as well as a

description of the research setting, are provided to help

other researchers determine generalizability and the need for

replication.

Academic support programs are typically designed and

implemented following the traditional process/product methods

of assessment, prescription, remediation, advising, and

support because it "seems to work." (See Moore, 1981.) New

models of designing and evaluating developmental programs

need to employ theoretical research to "provide practitioners

with a systematic, predictable method for determining

[program] value and making value decisions" (p. 48). Of

particular concern in the present research are the

theoretical notions of learning styles, psychological type,

and basic and study skills. Specifically, the research will

show how these variables may help better predict academic

success for underprepared students. While the outcome of the

research will not be the actual design of an academic support

program based upon Learning style, psychological type, and










16

basic and study skills, it is assumed that the findings will

aid in the establishment of such design.

Definition of Terms

The following terms are defined for use in the research,

and they will be more fully explained in later sections:

LSI: Kolb-McCarthy Learning Style Inventory

ACCE: Perceiving Dimension

AC: Abstract Conceptualization Scale

CE: Concrete Experience Scale

AERO: Processing Dimension

AE: Active Experimentation Scale

RO: Reflective Observation Scale

Style I (Divergers): CE and RO

Style II (Assimilators): AC and RO

Style III (Convergers): AC and AE

Style IV (Accommodators): CE and AE

MBTI: Myers-Briggs Type Indicator

El Dimension

E: Extraverted attitude

I: Introverted attitude

SN Dimension

S: Sensing function

N: Intuitive function

TF Dimension

T: Thinking function

F: Feeling function












JP Dimension

J: Judging attitude

P: Perceiving attitude

MBTI Type: One of 16 possible four-letter

combinations:

ISTJ ISFJ INFJ INTJ

ISTP ISFP INFP INTP

ESTP ESFP ENFP ENTP

ESTJ ESFJ ENFJ ENTJ

Combined Types: Two- or three-letter combinations

implying two or more types, i.e., EST- implies ESTJ and ESTP,

and IN-- implies INFJ, INFP, INTJ, and INTP.

LSA: Learning Skills Assessment (locally developed)

RATE: LSA Reading rate

COMP: LSA Comprehension score

READ: LSA Textbook reading score

LISTEN: LSA Listening score

NOTES: LSA Notetaking score

STUDY: LSA Study Skills score

ACT: American College Test

ACTC: ACT Composite

ACTE: ACT English

ACTM: ACT Mathematics

ACTS: ACT Social Sciences subscore

ACTN: ACT Natural Sciences subscore

HSGPA: High school grade point average

STATUS: Academic status at the end of the freshman year










18

CUMGPA: Cumulative grade point average at the end of the

freshman year

The core-curriculum disciplines, departments, and

courses, are operationally defined as they are used in the

research. Chapter Two, Review of the Literature, provides a

detailed description of the LSI and MBTI. Chapter Three,

Methodology, provides an operational definition of the LSA.

Hypotheses

There are five research hypotheses, and they relate the

major independent variables (Learning Style Inventory (LSI),

Myers-Briggs Type Indicator (MBTI), and Learning Skills

Assessment (LSA)) to the dependent variable hierarchy

(STATUS, COMGPA, and the core-curriculum Disciplines,

Departments, and Courses). The five hypotheses are:

S1: A regression equation which uses high school

grade point average and ACT scores to predict STATUS at the

end of the freshman year will be significantly improved with

the addition of scores from the Learning Styles Inventory,

Myers-Briggs Type Indicator, and Learning Skills Assessment

to the equation.

H2: A regression equation which uses high school

grade point average and ACT scores to predict CUMGPA at the

end of the freshman year will be significantly improved with

the addition of scores from the Learning Styles Inventory,

Myers-Briggs Type Indicator, and Learning Skills Assessment

to the equation.

H3: A regression equation which uses high school

grade point average and ACT scores to predict success in the












four selected core-curriculum disciplines will be

significantly improved with the addition of scores from the

Learning Styles Inventory, Myers-Briggs Type Indicator, and

Learning Skills Assessment to the equation.

H4: A regression equation which uses high school

grade point average and ACT scores to predict success in the

11 selected core-curriculum departments will be

significantly improved with the addition of scores from the

Learning Styles Inventory, Myers-Briggs Type Indicator, and

Learning Skills Assessment to the equation.

85: A regression equation which uses high school

grade point average and ACT scores to predict success in the

14 selected core-curriculum courses will be significantly

improved with the addition of scores from the Learning Styles

Inventory, Myers-Briggs Type Indicator, and Learning Skills

Assessment to the equation.

Methodology

One hundred seventy-two students enrolled in the 10

sections of a learning skills class (DVST 1012) in the Fall,

1983 term at the University of Arkansas, Fayetteville, were

assessed using the Kolb-McCarthy Learning Style Inventory

(LSI), the Myers-Briggs Type Indicator (MBTI), and a

locally-developed Learning Skills Assessment (LSA). The

results of the LSI and MBTI were provided to students by the

investigator in their classroom groups, and the results of

the LSA were provided to students in individual feedback

sessions by their classroom instructors.










20

Using the University of Arkansas' On-line Administrative

Student Information System (OASIS) and the Statistical

Analysis System (SAS) to determine core-curriculum

enrollments from the population under study, 4 disciplines,

11 departments, and 14 courses were chosen for analysis. To

ensure a sufficient N for the study, only those

disciplines, departments, and courses in which at least 20

enrollments from the study population occurred were chosen.

The selected core-curriculum disciplines, departments and

courses, along with freshman-year grade point average and

resulting academic status, became the focus of the research.

From the group of 172 students who received the initial

assessment, 128 entering freshmen without prior college

experience were selected as the target population for the

research. The three major independent variables (LSI, MBTI,

and LSA) and their subcomponents were introduced into

stepwise regression formulas along with HSGPA and ACT

composite and subscores to test the research hypotheses.

Chapter Three, Methodology, contains the research procedures

in greater detail.

Data Analysis

The data were analyzed using a logistical regression

analysis procedure, LOGIST, run under the Statistical

Analysis System (SAS) at the University of Arkansas Computing

Center. There were 30 independent variables (LSI, MBTI, LSA,

and their subcomponents) and 31 dependent variables (STATUS,

CUMGPA, and 4 core-curriculum disciplines, 11 departments,

and 14 courses). A stepwise regression analysis was










21

performed for each of the 31 dependent variables to determine

which and how many of the 30 independent variables

contributed significantly to the regression model variance.

Since there was a large number of independent variables, and

because the loss of degrees of freedom was a concern, the

independent variables were introduced into the model as

possible regressors in small sections, no more than four new

variables at a time. The variables which remained in the

model as regressors at the end of each run were retained as a

part of the possible regressor pool for the next run. A test

of model "adequacy of fit" (provided by the LOGIST program)

was performed after all the independent regressors were

identified for a particular dependent variable. Chapter

Four, Data Analysis, provides a more detailed description of

the analyses employed.

Summary

Academic success in college is dependent upon many

factors and their interrelationships, and success is not

easily or accurately predicted from high school grade point

averages and ACT scores alone. The present research was

designed to determine if success predictability for

underprepared students can be significantly enhanced by

adding measures of learning style, psychological type, and

learning skills to high school grade point average and ACT

scores in regression formulas. The research will provide the

basis for the later development of a process model to aid in

assessing, remediating, and monitoring entering underprepared

college freshmen.


















CHAPTER TWO
REVIEW OF THE LITERATURE

Introduction

The Review of the Literature begins with a brief

overview of the "Constructs of Learning Style and

Psychological Type" and "Learning Styles and the Learning

Environment" to provide theoretical and practical bases for

the section which follows on "Choice of Instruments for the

Research." The remainder of the chapter is devoted to the

theoretical bases for, a description of, the research on, and

a comparison of the Learning Style Inventory (LSI) and the

Myers-Briggs Type Indicator (MBTI). Research data do not

exist on the Learning Skills Assessment (LSA), and that

instrument is described operationally in Chapter Three,

Methodology.

Constructs of Learning Style and Psychological Type

Davies (1981) classified learning styles constructs as

theoretical (from a specific two-dimensional psychological

orientation), response-based (from empirical studies of

learning behavior), and integrated (from a range of sources

including learning theory, developmental theory, observation,

and analysis). He described theoretical constructs of

learning style by citing three examples of bi-polar

psychological theories. In the first of these theories he

cited research on two types of individuals: field-dependent

22












(viewing the world in a global, gestalt way) and field-

independent (viewing the world in an analytic, detached way).

A second two-dimensional theoretical approach cited by Davies

involves perceptive learners (those who focus on clues in the

data, examine relationships, and jump from part to part of

the material they are trying to study) and receptive learners

(those who focus on details in a logical order, avoid

judgement by preconceived ideas, and wait to get all the data

before reaching conclusions). Davies' third two-dimensional

theoretical approach involves two ways of choosing between

alternatives: impulsive and reflective.

Davies' second major classification of learning styles

constructs, described as response-based, are empirically-

observed notions of how students learn. As an example of

such studies, he cited research on observations of classroom

learning behaviors and the resulting categories: compliant,

anxious-dependent, discouraged workers, independent heroes,

snipers, attention seekers, and silent types. A second

example of response-based learning styles cited by Davies was

research on University of Chicago students who were

categorized as authoritarian, anti-authoritarian, or rational

types.

The integrated approaches to constructs of learning

styles, and the final major classification cited by Davies,

are: Hill's Cognitive Style Interest Inventory (symbols of

meaning, cultural determinants, modalities of inference, and

modes of memory), Kolb's Learning Style Inventory (concrete

experience, reflective observation, abstract












conceptualization, and active experimentation), and the

Myers-Briggs Type Indicator (extravert-introvert, sensing-

intuitive, thinking-feeling, and judging-perceiving).

Lawrence (1984b) asserted that "the term 'learning

styles' is used variously and loosely in educational

literature" (p. 7), and he broadly applied the term to

encompass four aspects of psychological makeup:

1. Cognitive style in the sense of preferred or
habitual patterns of mental functioning:
information processing, formation of ideas, and
judgements.
2. Patterns of attitudes and interests that influence
what a person will attend to in a potential learning
situation.
3. A disposition to seek out learning environments
compatible with one's cognitive style, attitudes,
and interests and to avoid environments that are not
congenial.
4. Similarly, a disposition to use certain learning
tools and avoid others. (p. 7)

Learning Styles and the Learning Environment

While many forms of imparting learning have been

identified (such as, independent study, laboratory methods,

mediated instruction, personalized systems of instruction,

traditional Lecture methods, seminars, peer tutoring,

computer-assisted instruction, discussion groups, programmed

instruction, discovery learning, etc.), no single method is

"best" because of the individual differences learners

possess. An identification of the preferred ways of

perceiving and processing information, a construct of

learning according to Kolb (1976), then, seems to be an

important first step in the learning process. Students

usually cannot exert much control over the instructional

methodology used in college courses, but helping them to











recognize their preferences for perceiving and processing

information will aid in their understanding of why they learn

more in some subjects and with some methods than others.

In an examination of instructional versus learning

styles (using the Canfield Instructional Style Inventory and

the Canfield Learning Style Inventory), Efurd (n.d.) found

that Westark Community College (Arkansas) students in speech,

anatomy, and physiology classes demonstrated the following:

male instructors and male students matched teaching and

learning styles more often than any other pairing; male

students were more competitive than females; male students

preferred working with peers more than female students; male

students preferred contact to be qualitative; females

preferred inanimate contact; male students preferred reading

more than female students; and instructors believed that

changing instructional methods can create positive

differences in learning.

In an interview with Boylan (1981), Canfield reported

that research on any one technique of matching teaching and

learning styles is going to be inconclusive as long as a

homogenyy of students" (p. 25) and their attendant individual

styles exist in a particular Learning situation. However,

Boylan reported Canfield as saying that students who do

receive preferred treatment tend to perform better in the

classroom, and that matching of teaching to learner styles

does enhance learning. In that same interview, Canfield

asserted that, based upon work done in achievement motivation












at some community colleges, it was possible to change

learning styles to achieve academic success.

Based upon observations, interviews, and experimental

studies over a decade, Dunn and Dunn (1979) offered the

opinion that "regardless of their age, ability, socioeconomic

status, or achievement level, individuals respond uniquely to

their immediate environment" (p. 239). Citing research on

matching teachers and learners, they reported significant

improvement in both achievement and motivation when matching

occurs. Further, they asserted that students are able to

accurately predict the modality in which they could best

achieve the desired performance. Also, according to Dunn and

Dunn, when students are taught by the method they predicted

was best for them, they score higher on tests, fact

knowledge, attitude, and efficiency of work than students

taught by methods dissonant with their preferences. However,

they further reported that mostot teachers can respond to

differences in students' learning styles. That is preferable

to trying to match students with teachers" (p. 238).

Commenting on Bill's Cognitive Style Mapping, Maxwell

(1979) cited cost, time, instructor incentive and energy,

training difficulties, traditional classroom methods, and

publisher resistance as obstacles to attempts to match

teaching and learning styles. Significantly for the present

research, Davies (1981) asserted that "(i)f the goal is

developmental, where learning to learn and individual

development are essential, then matching may be

inappropriate" (p. 4).












Bernice McCarthy, co-author of the Kolb-McCarthy

Learning Style Inventory, in an interview with Leflar (1982),

reported that in her early research she found that when

students' learning styles were reinforced during part of the

day the reinforcement carried over into the rest of the day

when styles were not matched. However, McCarthy (1980a)

elsewhere stated: "(A)ny 'matching' operation in a school

necessitates labels, and we already have too many labels in

education now. Do we need more?" (p. 84). Further, grouping

and labeling students with learning styles, and teaching to

those styles, "would be even more stultifying than the old IQ

groupings and trackings" (p. 29).

While there may be disagreement on the advisability of

matching teaching and Learning styles, there is utility in

helping both the instructor and student understand the

learning process. Hill, Canfield, Kolb, Dunn, and McCarthy

have all stressed the importance of gaining such an

understanding. (See Boylan, 1981; Davies, 1981; Dunn, 1981,

1983; Dunn and Dunn, 1979; Kolb, 1976, 1980, 1981, 1984;

Leflar, 1982; McCarthy, 1980a, 1980b; and Mentkowski, 1981.)

Choice of Instruments for the Research

Of the various learning styles constructs described in

the above categorization by Davies, the integrated approach

has appeal because of the sources from which the construct is

developed. The complex notion of learning styles is, in this

researcher's opinion, best derived by integrating Learning

theory, theories of human personality and development,

observation in Learning settings, and analyses of the












interrelationships among these factors. Two of the

integrated approaches cited by Davies--the Kolb Learning

Style Inventory (LSI) and the Myers-Briggs Type Indicator

(MBTI)--were selected for the present research because they

provide the necessary integration, theory, observation, and

interrelationships. (A slightly modified version of the LSI,

the Kolb-McCarthy Learning Style Inventory, was used to

assess students in the study. The Kolb-McCarthy LSI is

described in a later section of this chapter.)

Both the LSI and MBTI are grounded in Jungian theory of

psychological type, the MBTI more faithfully so, and thus

satisfy the need to examine human personality and

development. Both were derived from thousands of

observations in learning settings and thus have empirical

validity. Considerable research on the two instruments has

demonstrated practical applications of Jung's theories of

type and has provided analyses of learning theory and the

integrated construct of learning styles. The following

sections describe the LSI and MBTI, cite relevant research on

the two instruments, and provide comparisons with other

learning styles instruments where appropriate to confirm

validity.

Kolb's Learning Style Inventory

Kolb's Learning Style Inventory (LSI) is an assessment

of a person's preferred modes of perceiving (from concrete

to abstract) and processing (from reflective to active)

information. The underlying theory is










29
a dialectic one, founded on the Jungian .concept of
styles or types and experiential learning grounded
in the intellectual origins of Kurt Lewin in the
'40's and the sensitivity and laboratory education work
of the '50's and '60's. [At] the core of the model is a
simple description of the learning cycle, or how
experience is translated into concepts which in turn are
used as guides in the choice of new experience. (Kolb,
1976, p. 2).

Kolb has thus combined personality type (Jung), learning

theory (Lewin), and empirical observation into a unified

construct of learning style.

The determination of preference on each of the two

dimensions (perceiving and processing) results in an

identification of an individual's relative emphasis on four

learning abilities--Concrete Experience (CE), Reflective

Observation (RO), Abstract Conceptualization (AC), and Active

Experimentation (AE)--plus two combination scores that

indicate the extent to which an individual perceives

information abstractly or concretely (ACCE) and the extent to

which an individual processes information actively or

reflectively (AERO).

There are four responses, which the learner is asked to

rank order, to each of nine stem items on the LSI. These

responses correspond to the four end points on the two

dimensions. When the combination scores are computed and

"mapped," an individual learning style emerges. (See Figure

2-1, Kolb's Learning Styles.) These four styles are

described below:

Style I: Preference for CE and RO

Kolb has labeled this type learner the Diverger. He

or she has strength in "imaginative ability," "ability to






















Concrete Experience (CE)

Accommodator Diverger

IV I
Active I Reflective
Experimentation--------------------------Observation
(AE) (RO)
III I I

Converger 1 Assimilator


Abstract Conceptualization (AC)

Figure 2-1
Kolb's Learning Styles












view concrete experiences from many perspectives," and the

"ability to organize many relationships into a meaningful

'gestalt'" (1976, p. 5).

Style II: Preference for AC and RO

The Assimilator has strengths in "the ability to

create theoretical models," "inductive reasoning," and "the

ability to assimilate disparate observations into an

integrated explanation" (p. 6).

Style III: Preference for AC and AE

The Converger has his or her greatest strength in

"the practical application of ideas" (p. 5).

Style IV: Preference for CE and AE

The Accommodator has strengths in "doing things,

carrying out plans and experiments, and involving himself in

new experiences" (p. 6).

Kolb-McCarthy Learning Styles Inventory

McCarthy (1980a) with David Kolb, Paul Torrance, and

others combined the Kolb Learning Style Inventory and the

Torrance assessment of hemispheric preference into a two-part

assessment she called the 4MAT Survey Battery. Part One of

4MAT is referred to as the Kolb-McCarthy Learning Style

Inventory and is the version used in the present research.

The Kolb-McCarthy LSI is faithful to Kolb's construct of

learning style and modifies the original Kolb LSI only by

putting the response items into a sentence completion form

rather than single-word choices. All research cited below is

based upon the earlier Kolb LSI unless specifically

designated as Kolb-McCarthy or 4MAT.










32

In an interview with Leflar (1982b), McCarthy asserted:

The importance of the 4MAT system is not deciding what
the individual's learning style is, but rather the
purpose is to help the teacher to become more skilled in
[teaching] to all learning styles. Thus the cycle of
learning which Kolb developed is a natural cyclical
progression moving through all four learning styles.
[Teaching "around the circle" lets students "shine" some
of the time during their learning, but to] also move
away from their natural preferences some of the time to
acquire new skills and stretch their abilities. (p. 17)

Research on the Learning Styles Inventory

Item Analysis

Kolb (1976) showed in an item analysis of the LSI for

287 managers and management students that none of the

possible 36 choices correlated less than .45 with the

appropriate scale total, and most correlates fell between .50

and .60. Further, choices on a particular scale correlated

negatively with the theoretical opposite (e.g., CE choices

correlated negatively with the AC scale total). Also, no

significant correlations were found between choices and their

theoretically orthogonal scales (e.g., RO choices did not

significantly correlate with the CE or AC scale totals).

These data indicate that the item choices on the four LSI

scales (CE, RO, AC, and AE) have high convergent and

discriminant validity.

Intercorrelation of LSI Scales

Kolb (1976) showed that for a sample of 807 people CE

and AC were negatively correlated (-.57, p < .001), and RO

and AE were also negatively correlated (-.50, p < .001).

Orthogonal correlations were low but significant because of

the large sample size (CE with RO, .13; RO with AC, -.19; AC










33
with AE, -.12; and AE with CE, -.02). All correlations but

AE/CE were significant at p ( .001. Because of the

intercorrelation data, Kolb justified the creation of two

combination scores to measure continuous dimensions (ACCE and

AERO). With the ACCE dimension, AC correlated .90 and CE

correlated -.85. With the AERO dimension, AE correlated .85

and RO correlated -.84.

Reliability--Test-Retest

Kolb (1976) pointed out that since the four basic

learning modes assessed by the LSI are theoretically

interdependent, any action, including responding to the test,

is determined in varying degrees by all four learning modes.

Responses to the choices indicating preference will be

variable according to the individual's interpretation of the

situation presented. Further, since there are few

theoretically "pure types" (p. 12), reliability will be

dependent upon the extent a person has attained a preferred

style. Retesting will vary to the extent the individual has

approached or moved away from his or her stated preference.

Because of this test-retest problem, Kolb cautioned that the

LSI will be of limited use for assessment and selection of

individuals "without additional detailed knowledge of the

person and his situation" (p. 13). With these cautions in

mind, Kolb applied test-retest assessments to four groups who

ranged in discontinuity from three to seven months. The

combined scores showed a high correlation between test-retest

on the continuous scales (ACCE = .91; AERO = .71). The

individual scales showed similarly high test-retest












correlations, with the exception of CE which reached a

correlation of only .48.

Reliability--Split-Half

Kolb (1976) used the Spearman-Brown formula between

halves of the LSI for five different groups (total N =

690). Correlations of approximately .80 were consistent

across all five samples for the combination scores, but

somewhat less satisfactory for the individual scales, perhaps

due to the shortness of the scales. From the split-half

studies, Kolb concluded that the continuous scales were

highly reliable indices suitable for most research

applications. The greater variability in the individual

scales suggested greater caution.

LSI Population Data and Applications

Table 2-1, LSI Undergraduate Scores, is a summary of

scores for five selected undergraduate populations supplied

to Kolb (1976) by other researchers. All populations in that

summary demonstrated a preference for Style I, Divergers,

with the Lesley female undergraduates showing the strongest

preference for that style.

Kolb (1976), after examining data from a sample of

practicing managers and graduate students in management,

identified a correspondence between LSI styles and

undergraduate academic specialization:

Divergers: History, Political Science, English, and

Psychology.
















Table 2-1
LSI Undergraduate Scores

School N CE RO AC AE Style Preference


Kent St.a 135

MITb 342

U. Mass.c 284

Lesleyd 66

Alvernoe 213

Combined 1040


14.9

14.8

14.5

16.2

15.3

14.9


13.8

13.6

14.7

14.4

14.7

14.2


17.2

17.4

17.0

16.0

15.6

16.8


16.0

15.0

15.7

15.0

15.7

15.5


(Diverger)

(Diverger)

(Diverger)

(Diverger)

(Diverger)

(Diverger)


aUndergraduate population

bSeniors only

CEngineering undergraduates

dFemale undergraduates

etiberal arts females












Assimilators: Economics, Mathematics, Sociology,

Chemistry, and Physics.

Convergers: Nursing and Engineering.

Accommodators: Business.

McCarthy (1980a) administered Kolb's Learning Style

Inventory to 329 Chicago high school students (ages 17-18;

183 females, 146 males) to determine their learning styles

and found that approximately equal percentages of females and

males fell into each of the four quadrants. Her findings are

shown in Table 2-2, Distribution of Learning Styles for

Chicago High School Students. She further reported that at

this age, students tended to favor the Concrete Experience

scale (60% for CE as opposed to 40% for AC) and that students

slightly favored the Reflective Observation scale (57% for RO

as opposed to 43% for AE). McCarthy reported that her

findings on the processing dimension (AERO) conflicted with

Kolb's findings "who has reported a more active orientation

in the age group 16-35 and then a tapering off toward a more

reflective orientation in later years" (p. 81).

In examining the results of combined learning styles

studies over several years, McCarthy (1980b) expressed

concern that schools teach primarily to one style learner,

the Assimilator. Since her data indicated that only

approximately 28-30% of the general population may be

Assimilators, she cautioned that 70% of the students may not

be getting the education best suited to their abilities.















Table 2-2
Distribution of Learning Styles for
Students


Chicago High School


Styles

Divergers

Assimilators

Convergers

Accommodators


Females (%)


(32)

(23)

(19)

(25)

(100)


Males (%)


Total (%)


(38)

(20)

(16)

(25)

(99)


(35)

(22)

(18)

(25)

(100)


Chicaao Hiqh School










38

Mentkowski (1981) administered the Kolb LSI to 679 women

who entered a two-year nursing program at Alverno College in

a longitudinal study to determine the effect of the college

experience on learning styles. The women were either

"weekday" students (n = 412, average age = 22) or

"weekend," and thus considered non-traditional, students

(n = 267, average age = 33). Entering women in both

groups were similar in their overall preference for style

(Divergers), but entering weekend women had a significantly

greater preference for CE and AE and a significantly lesser

preference for RO over entering weekday women. Upon

graduation two years later, compared to entering women in

their respective groups, weekend students were significantly

less likely to prefer CE; both groups were less likely to

prefer RO; and both groups were more likely to prefer AC than

their entering counterparts.

Weekend students made an overall shift from Divergers to

Accommodators, while weekday students remained Divergers, but

were not so strong in that style. Since the weekend students

were on the average 11 years older than weekday students, and

significantly higher on the CE scale upon entering college,

Mentkowski concluded that "life experiences" prior to college

for those women did not cause a shift toward AC, but actually

enhanced the preference for CE. Similarly, since weekend

students were significantly higher on the AE scale, she

concluded that life experiences prior to college may have

enhanced the AE preference. College, in contrast to life












experiences, according to Mentkowski, does shift the

preference from CE to AC and does promote a more "balanced"

learning style.

In correspondence with this author, Kolb has referred to

one's "adaptive style," and he has developed an instrument,

the Adaptive Style Indicator (ASI), to measure this ability.

(See Kolb, 1980 and 1984 for descriptions of the ASI and its

implications for education.) Apparently, the successful

Alverno students had learned to adapt to a variety of

learning situations, thus a shift to the "more balanced"

style.

Leflar (1982a) reported, after administering McCarthy's

4MAT Battery, that freshmen at the College of the Ozarks

(Arkansas) were significantly, "overwhelmingly," concrete in

their learning preference, indicating they would profit most

from simulations, group discussions, dramatizations, peer

tutoring, and independent explorations. Traditional,

abstract classroom methods of Lecturing and assigning

readings were used in the majority of the courses in the

College at the time, and, subsequent to Leflar's findings,

training of faculty in the use of McCarthy's 4Mat system for

advising and teaching has resulted at the faculty's request.

LSI Comparative Studies

Concerned about the construct validity of learning

styles instruments, Ferrell (1983), analyzed four

instruments: the Grasha-Riechmann Learning Style Scales

(SLSS), Kolb's Learning Style Inventory (LSI), Johnson's

Decision Making Inventory (DMI), and the Dunn Learning Style










40

Inventory (DLSI). Subjects (all from Southern Illinois) were

260 high school students and 211 community college students

ranging in age from 17 to 21, with an equal representation of

males and females in both populations. Data obtained from

each of the subjects for the four instruments were separately

factor analyzed. Three factors were obtained from the SLSS

and four from each of the others. For the LSI, 23 (of the

36) items loaded on four single factors and seven items did

not have salient loadings on any factor. Ferrell reported

that only for the Kolb LSI did the revealed factors match the

described learning styles. Items comprising the four factors

extracted matched the four learning preferences as described

by Kolb and "supported Kolb's conceptualization of learning

styles" (p. 36). In the same 1983 study, Ferrell also

examined the four instruments in relation to Keefe's (1979)

conceptualization of learning style (comprised of cognitive,

affective, and physical/physiological behaviors). None of

the instruments were found to completely fit Keefe's

paradigm. The LSI and DMI were cognitive in nature; the SLSS

was comprised of cognitive and affective factors; and the

DLSI reflected the affective and physical/physiological

aspects of learning style. A factor representing a positive

attitude toward Learning was found on both the SLSS and the

LSI. Analytical/abstract orientations were found in factors

comprising the LSI and the DMI. In spite of the overlapping

factors across the four instruments, "(they) clearly were not

measuring the same thing" (p. 37). Therefore, Ferrell












concluded that "learning styles literature has not

established a single unified conception of learning style"

(p. 39). (See also Lawrence, 1984a, 1984b.)

In an attempt to define measurements and outcomes of

certain learning styles instruments, Dunn and others (1981)

compared the definitions and applications of eight different

learning style instruments, including the LSI. The results

of that comparison are summarized in Table 2-3, Comparison of

Learning Style Instruments. Dunn et al. reported that while

there were overlapping characteristics among the eight

approaches to measuring learning style (structure,

motivation, sociological needs, perceptual modes, and thought

processing), there were important differences, as shown in

Table 2-3. Further, disagreements existed among the

researchers on whether learning styles are inherited or

culturally determined, whether teaching should focus on

strengths exclusively or on both strengths and weaknesses to

build adaptability, and to what extent learning styles are

fixed or change over time.

Jung's and Myers' Theories of Psychological Types

To better understand the Myers-Briggs Type Indicator, it

is important to have an appreciation of Jung's and Myers'

theories of psychological types. Jung developed three

bi-polar psychological orientations. They are: Extraversion

and Introversion, Thinking and Feeling, and Sensing and

Intuiting. (See Campbell, 1971.) A fourth dimension,

Judging and Perceiving, was implicitly derived from Jung's

Theory by Isabel Myers and her mother, Katharine Briggs.





Table 2-3
Comparison of Learning Style Instruments


Author

Can field and
Lafferty


Dunn, Dunn,
and Price



Gregorc




Hill
IliLL




Hunt


Kolb






Ramirez and
Canteneda


Sclhmck


Instrument

Learning Style
Inventory


Learning Style
Inventory


Transactional
Ability
Invoetory


Cognitive Style
Interest
Inventory


Teacher Aaaess-
ment of Student
Learning Styles

Learning Style
Inventory





Child Hating Form



Inventory of
Learning Process


Definition


Application


Style derived from ac- Used to develop materials for class
ademic, structural, or individuals to understand stu-
and achievement con- dent difficulties and to aid couna-
ditions, content, mode eling. Emphasis on attltdue and ef-
and expectations. feet.
Strengths identified Used to diagnose individual learn-
in environmental, emo- ing characteristics. Suggests mater-
tional, sociological, lals to facilitate achievement.
and physical dimen-
sione.
Mind dualities of con- Emphasis on matching of materials
crete-abstract and and methods to preferences. Encour-
sequential-random are ages strengthening of non-
palrod to determine preferences.
meaning.
Meaning reflected by Identifies major, minor, and negoti-
how qualitative and able categories. Serves to develop
theoretical symbols Personalized Educational Program
are handled, influ- (PEP).
enced, and perceived.
Describes most likely Used to enhance development of con-
learning conditions ceptual level, which ranges from
and amount of struc- unsocialized to independent.
tura needed.
Describes orientation Emphasis on awareness of style and
on dimensions of con- alternative modes. Focus on non-
crete experience, ab- dominant orientations as well as
stract conceptualiza- strengths.
tion, reflective ob-
servation, and active
experimentation.
Style determined by Used to match and mismatch learn-
field dependency and Ing and teaching to enhance "bi-
cultural differences, cognitive ability" and reduce style
preferences.
Measures information Used to encourage development of
processing ranging deep, thoughtful, and elaborative
from shallow and re- processing.
iterative to deep and
elaborative.










43

These four dimensions were used to develop scales which form

the Myers-Briggs Type Indicator. (See Myers, 1980.)

Extraversion and Introversion

Myers (1980) summarized the relative interests of the

introvert and the extravert:

The introvert's main interests are in the inner
world of concepts and ideas, while the extravert is
more involved with the outer world of people and things.
Therefore, when circumstances permit, the introvert
concentrates perception and judgement upon ideas, while
the extravert likes to focus them on the outside
environment. (p. 2)

It should be made clear that neither Jung nor his

followers have suggested that persons are exclusively one

extreme or another in the extravert-introvert dimension, nor

any of the other bi-polar dimensions described below. On the

other hand, Jung devoted much of his life to examining

neuroses resulting from the "falsification" of attitude type

(such as parental influence). Such examinations are clearly

(and thankfully) beyond the scope of this investigation.

Sensing and Intuiting

In describing the sensing-intuition dimension Myers

(1980) stated:

One form of perception is the familiar process of
sensing, by which we become aware of things directly
through our five senses. The other is the process of
intuition, which is an indirect perception by way of
the unconscious, incorporating ideas or associations
that the unconscious tacks on to perceptions coming from
the outside. (p. 2)

Thinking and Feeling

Myers (1980) described the thinking-feeling dimension as

two distinct and sharply contrasting ways of coming to
conclusions. One way is by the use of thinking, that












is, by a logical process, aimed at an impersonal
finding. The other is by feeling, that is, by
appreciation--equally reasonable in its
fashion--bestowing on things a personal, subjective
value. (p. 3)

Judging and Perceiving

The definitions Myers provided (above) for the sensing-

intuiting and thinking-feeling dimensions are concise

explanations of the perceiving and judging attitudes,

respectively. However, she further stated:

[These attitudes are] a way of life, a method for
dealing with the world around us. (B)oth cannot be
used at the same moment, [and] most people find one
attitude more comfortable than the other This
preference makes the difference between the judging
people, who order their lives, and the perceptive
people who just live them. (pp. 8-9)

Dominant Function

As described above, perception occurs either through the

senses or through intuition, and judgement occurs either

through thinking or through feeling. While all four

functions are distinctive elements of one's deep structure of

psychological type and are all used regularly, one is

considered "dominant and forms the centerpost of the

mental system. It provides psychological consistency, (and)

basic attitudes, values, and interests can be seen as flowing

from the arrangement of preferences of the four functions"

(Lawrence, 1984b, p. 2, emphasis added). (Jung referred to

the function opposite of the dominant function as "inferior,"

and von Franz (1979) has provided a detailed characterization

of the this function.)










45

Schemel and Borbely (1982) graphically discuss the two

preferred attitudes (which must be chosen from El and JP)

and the two preferred functions (which must be chosen

from SN and TF). See Appendix A, Attitudes and Functions by

Type, for a complete Listing of all the attitudes and

functions, including dominants and auxiliaries, for each of

the 16 types.

Myers-Briggs Type Indicator

Introduction

The Myers-Briggs Type Indicator (MBTI) has been in use

since first published by the Educational Testing Service in

1962. Its purpose was and is to implement Jung's theory of

psychological type.

Myers (1962) acknowledges her mother, Katharine C.

Briggs, for the original theory of type based upon Jung's

work, but it is Isabel Briggs Myers who is credited with the

authorship of the MBTI. Myers, educated at home after

becoming a "first grade dropout" (Lawrence, 1982, p. 14),

began working on her theories of type during World War II to

"promote human understanding [and] to help people choose

careers which allowed them to use their best abilities" (p.

14). During the next 20 years, by testing thousands of high

school students then tracing their career choices, and by

testing over 5,000 medical students then checking to see if

they had been successful in their chosen careers, Myers

slowly developed the MBTI.

The Educational Testing Service learned of Myers' work

through one of the medical schools used in her research and











published the first MBTI (Form F) in 1962, primarily for

psychologists and other professionals interested in human

behavior. The data for the 1962 publication were from a

large data base collection which terminated in 1957. Twenty

years later, Myers repeated the normative measures to

determine if temporal or cultural changes had eroded the

validity of the Type Indicator and to make some minor

modifications which two decades of experience with the

instrument suggested were desirable. (See Myers, 1977). In

1975, the MBTI (Form G), then published by Consulting

Psychologists Press, was considered ready for redistribution.

Description of the MBTI

The MBTI (Form G) is a 126-item, forced-choice

questionnaire (scaled down from the original 166-item, Form

F) designed as a tool to determine and utilize Jung's

psychological types. The result of an individual assessment

is a determination of one's preference on each of four

dimensions: Extraversion (E) or Introversion (I), Sensing (S)

or Intuition (N), Thinking (T) or Feeling (F), and Judging

(J) or Perceiving (P). Sixteen combinations of type are

possible from the preferences on the four dimensions. These

types are shown in Table 2-4, The Sixteen Psychological

Types.

Since the MBTI focuses on preferences (strengths) and

not weaknesses, the emphasis is on what is unique, yet

normal, about the individual being tested, and not an

emphasis on abnormality, such as is often the case with other

psychological measurements. The non-judgmental nature of the










47







Table 2-4
The Sixteen Psychological Types

ISTJ ISFJ INFJ INTJ

ISTP ISFP INFP INTP

ESTP ESFP ENFP ENTP

ESTJ ESFJ ENFJ ENTJ












Type Indicator aids in the understanding of individuals

without the adverse labeling effects.

Distribution of Type

McCaulley (in press) has combined the results of Myers

(1982) samples of 4,933 male and 4,387 female llth and 12th

grade high school students. These students were largely

college-preparatory and were above average in socio-economic

status. The distribution of types from those samples is

shown in Table 2-5, Distribution of Type for Combined Samples

of llth and 12th Grade High School Students.

From these data, distributions on the four dimensions

can be computed for students about to enter college. Table

2-6, Distribution of Preferences for Combined Samples of llth

and 12th Grade High School Students, provides percentage

distributions for males and females for the above samples.

Males and females showed the same overall preferences (though

slightly different percentages) for three of the four

dimensions--EI, SN, and JP. For the TF dimension, a reversal

occurred by sex. Males preferred thinking over feeling 61%

39%; females preferred feeling over thinking two to one.

These data are consistent with Myers (1977) earlier findings.

Reliability

Myers (1962, 1977) and McCaulley and Natter (1980) have

reported high internal consistencies (split-half

correlations) for the MBTI on samples from high school

through college, ranging from .77 to .87 for EI, .70 to .87

for SN, .44 to .86 for TF, and .71 to .84 for JP. McCaulley

and Natter (1980) reported that underachieving samples showed












Table 2-5
Distribution of Type for Combined
Grade Hiqh School Students

Type Males
(n = 4,933)

ISTJ 8.8%

ISTP 6.1

ISFJ 4.5

ISFP 5.2

INFJ 1.6

INFP 3.5

INTJ 3.7

INTP 4.8

ESTP 8.9

ESTJ 17.1

ESFP 7.2

ESFJ 8.2

ENFP 6.0

ENFJ 2.8

ENTP 6.4

ENTJ 5.3


Samples of llth and 12th


Females
(n = 4,387)

4.9%

2.0

9.5

5.7

2.0

4.4

1.4

2.1

3.9

12.6

11.8

20.5

9.4

4.5

3.2

2.4










50







Table 2-6
Distribution of Preferences for Combined Samples of llth
and 12th Grade High School Students

E I S N T F J P

Males 61.9 38.2 66.0 34.1 61.1 39.0 52.0 48.1
(n = 4,933)


Females 68.3 32.0 70.9 29.4 32.5 67.8
(n = 4,387)


57.8 42.5


Note: Percentages do not equal 100 due to rounding.










51
lower internal consistency correlations, particularly on the

TF dimension.

Reporting on research on test-retest reliability,

McCaulley and Natter (1980) cited Strickler and Ross (1962)

who found in a study of 41 Amherst students reliabilities of

.73 for EI, .69 for SN, .48 for TF, and .69 for JP.

McCaulley and Natter (1980) also reported that a study of

Howard University Graduates by Levy, Murphy, and Carlson

(1972) showed a higher test-retest reliability (after a

two-month interval) than the 1962 Strickler-Ross comparison,

and the correlations were slightly higher for females than

for males.

Carskadon (1979b) found most test-retest reliabilities

on Form G of the MBTI to be good to satisfactory. However,

he also reported that test-retest reliability for males on

the TP scale tended to be poor, replicating findings on Form

F of the MBTI. Later, in a replication of the study of

reliability of Form G, Carskadon (1982b) found a puzzling

reversal of the TF reliability between the sexes. Males

showed a .91 test-retest reliability, females showed only

.56, a highly significant difference. Carskadon suggested

further research on the interplay between sex and type and on

how the same type preferences may be manifested differently

for the two sexes.

Myers (1962) spoke to the issue of change in type from

test to retest:

Type development, i.e., the extent to which a person
actually has developed the processes and attitudes which










52

he prefers, enters every equation as an unknown quality.
[Nor is it known how many persons in a sample would bel
answering virtually at random because their type is
insufficiently developed to govern their responses. (p.
19)

Summarizing MBTI reliability, McCaulley (1981) reported

that when changes in type occur on retest, most changes

affect only one preference, and those preferences with low

original endorsement are the most likely to change. (c.f.,

Kolb, 1976. See also Carskadon, 1982a.)

MBTI Validity Studies

McCaulley and Natter (1980) summarized research on

comparison of the MBTI with the Gray-Wheelright Psychological

Types Questionnaire (which determines Jungian types) and with

a battery of 32 tests measuring ability, interest, and

personality variables. The MBTI measured the same dimensions

as the Gray-Wheelright, and the scales from the 32-test

battery significantly correlated with the MBTI in the

predicted directions. Citing earlier work by Myers (1962),

they summarized the comparison of the MBTI with the Alport-

Vernon-Lindsey Study of Values (AVL), Edwards Personal

Preference Scale (EPPS), and the Personality Research

Inventory (PRI). Significant correlations are shown in Table

2-7, and the strongest correlation in each comparison is

underscored.

From all the above findings, McCaulley and Natter con-

cluded that there is construct, concurrent, and predictive

validity for the MBTI personality, academic, and behavioral

measures. They further reported that intuitives consistently

score higher on aptitude measures than sensing types. In











Table 2-7
MBTI Compared with the Alport-Vernon-Lindsey Study of
Values, the Edwards Personality Preference Scale, and the
Personality Research Inventory

Alport-Vernon-Lindsey Study of Values

Theoretical INTJ

Economic ESTJ

Aesthetic IN-P

Social -SF-

Political EST-

Religious -NF-

Edwards Personality Preference Scale

Intellectual Achievement INT-

Order ISTJ

Exhibition E--P

Autonomy -NTP

Affiliation E-F-

Dominance E-T-

Nurturance --F-

Change -NTP

Endurance --TJ

Personality Research Inventory

Complexity Tolerance -NFP

Impulsiveness EN-P

Talkativeness E---

Gregariousness ES--

Work Attitude E--J

Artistic -NFP

Liking to use Mind -NT-

Note: Strongest correlations for each type are underscored.










54

ability measures, intuitives consistently score higher than

sensing types in English, vocabulary, verbal, and reading

tests.

Educational Applications

McCaulley and Natter (1980) reported that University of

Florida ESTPs ranked first in how East they read, but low in

comprehension. INTPs, who ranked second in reading speed,

had a high comprehension rate.

In a study reported by Myers (1980), Florida high school

students showed a significant difference between intuitives

and sensing types on mathematics and science. Ns outscored

Ss on five out of eight mathematics measures and two out of

three science measures. The primary difference occurred with

problem-solving ability, not computational skills. In

comparing intuitive and sensing types' academic aptitude and

study skills, intuitives scored significantly higher on all

measures (overall grade average, 8th grade study skills, L2th

grade aptitude, and California Test of Mental Maturity).

In contrast, according to Myers (1980), sensing types

significantly outscored intuitives in practical knowledge

applications (such as, electronics and mechanics).

Similarly, thinking types outscored feeling types in

technical, clerical, administrative, electronics, general

mechanics, and motor mechanics. She further asserted that

the most conspicuous consequence of type preference in

educational settings is in the choice between the two kinds

of perceptions: sensing and intuition.

It is therefore understandable that, as most schools are
now run, sensing children have less use for school than










55
intuitive children do (often no use at all), that on the
average they make lower grades and score lower on
intelligence tests (though not enough lower to account
for their grades), and that they far more frequently
drop out. (p. 127)

Subtantiating the above claim, Myers (1980) compared the

671 finalists for National Merit Scholarships and found

percentages of 17 and 83 for S and N, respectively. She also

assessed 500 students who did not finish 8th grade and found

dropout rates of 99.6% and 0.4% for S and N, respectively.

McCaulley and Natter (1980) used the MBTI to type

students at the Florida State University Developmental

Research School. Also available were results from a 187-

item "learning activities" questionnaire and selected high

school measurements (grades in school, eighth grade test,

ninth grade test, twelfth grade test, vocational aptitude,

and IQ). From these data they concluded the following:

Extraverts may do better on oral than written tests

and on tests applying knowledge than on tests of concepts and

ideas. While confident and willing to make oral

presentations, they scored lower than introverts on several

academic measures including aptitude, reading, and

mathematics. Introverts may do better on written tests

than on oral tests, and better on tests measuring concepts

and ideas than on tests of practical application. They may

have an advantage in college work where the focus is on

understanding concepts.

Sensing types have a greater interest in the "real

thing" than intuitives, and test scores may underestimate

their true knowledge. At a disadvantage with timed tests,










56

written tests, or tests requiring a knowledge of theory, they

may prefer objective tests which present, rather than ask for

the creation of, choices. They may do well on power,

performance, motor-skill, spatial, or perceptual intelligence

tests, but they may read less and write more poorly than

intuitives. Intuitive types, with their interest in

reading, tend to do well on tests requiring writing on the

meaning of words (comprehension) or symbols (problem

solving). Their perceptual ability serves them well on timed

tests, and they may score higher on academic measures than

sensing types. Intuitives may prefer independent study,

indicate problems with time management, and like spending

time on non-required reading.

Thinking types may have an edge over feeling types in

mathematics and science courses and in courses dealing with

equipment. They indicate a seriousness about education and

learning strategies, and they may score higher than feeling

types in electrical, mechanical, and technical areas.

Feeling types tend to score higher than thinking types on

social awareness and sensitivity and lower on mathematics,

science, and technical skills. They may achieve mastery in

weaker subjects, though, if the goal is value-related,

especially to human relationships.

Judging types may have grades that are Likely to be

better than would be predicted on the basis of aptitude

scores. They like scheduling and planning their work, and

they profess an enjoyment for "the work of school," such as,

studying, preparing for exams, making reports, and doing










57

projects. They may score slightly higher than perceptives in

making grades, while scoring Lower in all other academic

measures. Perceiving types pick up information through

curiosity, and their aptitude scores may indicate grades

higher than actually received. They tend to procrastinate

and not plan their work. Searching for freedom, they may

daydream. They may score higher on every measure of academic

achievement except grades in school.

In discussing self-concept, McCaulley and Natter (1980)

reported that I--P types are Likely to turn inward and feel

inadequate, while E--J types focus on the outer world and are

likely to have confidence in managing it. Intuitive types,

who tend to achieve higher academic scores, say they feel

academically superior, as do thinking types. Perceptive

types, who desire spontaneity and freedom, may feel they are

"stuck in a rut" (p. 180). Measures of self-concept, or

confidence, were the best predictors of good grades in the

McCaulley and Natter study. Confident students were not test

anxious, but they were concerned about exams. Such students

had a study plan and were serious about their work. There was

no difference in high predictors of good grades for black and

white students.

McCaulley and Natter (1980) categorized four different

knowledge approaches based upon the interrelationships of

extraversion-introversion and sensing-intuition:

IS--: Careful Compilers for whom knowledge is

important to establish truth.










58

ES--: Pragmatists for whom knowledge is important for

practical use.

IN--: Academics for whom knowledge is important for

its own sake.

EN--: Innovators for whom knowledge is important for

innovation.

They further described the interrelationship of sensing-

intuition and thinking-feeling:

-ST-: Practical, matter-of-fact; collects data which

can be verified by the five senses and can be used to make

logical decisions.

-SF-: Sociable, friendly; collects sense data, but

relies upon feeling rather than logic for decision-making.

-NF-: Enthusiastic, insightful; uses possibilities to

judge with personal warmth.

-NT-: Logical, ingenious; uses possibilities to make

logical decisions.

Considerable research, in addition to the major studies

cited above, exists on applications of type theory in

educational settings. The remainder of this section is a

selection of brief research reports to provide a further

foundation for how type theory may help explain students'

success (or lack of it) in the areas chosen for analysis in

the present research.

Contessa (1981) found that eighth graders (an age when

most students waiver between concrete operational and formal

operational stages of cognitive development, according to

Piagetian concepts) demonstrated a significant relationship











between sensing and intuition and level of cognitive

development, with a proportionately greater number of Ns

classified as formal operational thinkers than Ss. He also

reported that formal operational thinkers performed

significantly higher than operational thinkers on a posttest

of the science concept of model building. (Lawrence (1982)

has warned: "Correcting the biases of instruction that harms

ES type children is perhaps the most crucial unrecognized

problem of American Education" (p. 42). It is ironic that ES

teachers may be the most influential in changing students'

attitudes toward school. (See Cohen, 1981.))

Todd and Roberts (1981) found no difference between art

majors and music education majors on the El and TF

dimensions. They reported, however, a significant difference

on the SN and JP dimensions. Two cells, ENFP and INFP, held

58.5% of the art majors, while music majors seemed to be more

randomly distributed among the 16 types, with the greatest

percentage (20) represented by ENFJ. Music majors preferred

sensing (42%) and judging (58%). Art majors preferred

intuition (80%) and perception (74%).

In examining the relationship between type and

performance of practicing speech pathologists on the National

Examination in Speech Pathology, Middleton and Roberts (1981)

found the predominant preference to be ENFJ. The N function

was the best predictor of scores on the exam, accounting for

10.4% of the variance in the sample.

Hoffman, Waters, and Berry (1981) reported a high

dropout rate among EPs (53%) from a computer-assisted program












to teach Morse code. The EPs who dropped comprised 38% of

all dropouts from the program compared to an expectancy of

less than 20%.

When offered the opportunity to volunteer for an

innovative, non-structured humanities program at Walt Whitman

High School, Barberousse (c. 1970) found that among the 100

students who volunteered, 6% were STs, 15% were SFs, 19% were

NTs, and 60% were NFs.

Roberts and Butler (1982) found that among 100 upper-

division and graduate students enrolled in a reading program

at Texas Tech University, there was a significant positive

correlation on the SN dimension with vocabulary,

comprehension, and total reading scores. No correlation was

found with reading rate and SN, and no other MBTI dimensions

correlated with any of the reading scores.

Roberts and Butler (1982) found that intuitives

significantly preferred reading as their choice of a medium

of instruction, while sensing types did not. One type, ESFP,

significantly rejected reading as an instructional choice.

ESTJs and INFPs significantly preferred lecturing. To

explain the apparent contradiction of these two types,

Roberts asserted: "ESTJs like to lecture, and INFPs like to

listen to good lectures, so both rated the lecture high as a

medium of instruction" (p. 86). Fourteen of the 16 types

rejected audio as a desired medium of instruction. ISFPs and

ESFPs did not reject audio, nor did they show it as a

preference.










61
To demonstrate difference in one pair of type extremes,

McCaulley (1976) reported that from a sample of 3275

University of Florida entering freshmen in 1972, ESFJ types

ranked 16th among all the types on the Florida 12th grade

Placement Test scores, 16th on the SCAT, and 12th among their

class in first-quarter grades. INTP types, in contrast,

ranked first on the 12th grade placement test, first on SCAT

and SAT, and first on first-term grades. Focusing on the SN

dimension, McCaulley reported that sensing types were 78% of

a sample of 135 underprivileged university students; Ns were

22% of that group. By contrast, 18% of 1101 National Merit

finalists were Ss; 82% were Ms. "College," she reported,

"'lith its demand for complex problem solving, and for working

at an abstract, theoretical, or imaginative Level, suits the

interests of the intuitive type" (p. 7).

Many of the above research findings lead to the

conclusion that certain type combinations may be very

important in educational research. In addition to examining

the individual scales and the dichotomous dimensions,

McCaulley (1977) has recommended an examination of certain

groupings of type while conducting research:

Group 1: I--J, I--P, E--P, and E--J to examine the

perceiving function as dominant (I--J and E--P) and the

judging function as dominant (I--P and E--J).

Group 2: -ST-, -SF-, -NF-, and -NT- to examine the

combinations of perception (S and N) and judging (T and F).

Group 3: -S-J (practical, organized), -S-P (practical,

spontaneous), -N-P (innovative, spontaneous), and -N-J

(innovative, organized).












Group 4: --TJ (analytic, executive), --TP (analytic,

adaptable, --FP (sympathetic, adaptable), and --FJ

(sympathetic, executive).

Group 5: IN-- (academic), EN-- (innovative), IS--

(realistic), and ES-- (pragmatic).

Students' Perception of Type

Carskadon (1982a) reported that 129 college students

significantly chose a description of their actual type or a

description of a type in which their weakest preference was

reversed over descriptions which had preferences on El and JP

reversed, descriptions which had preferences on SN and TF

reversed, and descriptions which had all four preferences

reversed. (See Table 2-8, College Students' Perception of

Type.) In the same study, Carskadon determined that students

could accurately predict their preferences: 18% accurately

predicted all four functions; 44% were correct on three

functions; 24% were correct on two functions; 12% were

correct on one function; and only 2% were wrong on all four

functions. Broken down by dimension, students accurately

predicted their El dimension 68% of the time, their SN

dimension 66% of the time, their TF dimension 63% of the

time, and their JP dimension 72% of the time. In a

replication study, Carskadon (1982c) reported that 50% of 118

college students ranked their actual type descriptions as

their first choice. Further, students rated their actual

descriptions as "very true" (27%) or "mostly true" (37%),

while rating the description with all scales reversed as

"partly true" (39%) or "not very true at all (34%). As in












Table 2-8
College Students' Perception of Type
(N = 129)

Description of Type Percentage Who Ranked It Highest

Actual description 35

Weakest preference reversed 31

E-I and J-P reversed 23

S-N and T-F reversed 7

All scales reversed 4












the earlier study, Carskadon found that students were much

more sensitive to reversals on their dominant/auxiliary (SN

and TP) dimensions than they were on their attitude (El and

JP) dimensions.

Comparison of the LSI and MBTI

The MBTI and the LSI variables can be expected to appear

in combinations and patterns in the regression equations

which form the bases of the research hypotheses; it is

instructive, therefore, to examine research which compares

the two instruments. From a comparison of descriptions of

LSI styles and MBTI types, Kolb (1984) predicted the

following relationships:

Divergers are similar to I-F- types.

Assimilators are similar to IN-- types.

Convergers are similar to E-T- types.

Accommodators are similar to ES-- types.

In summarizing data from two populations, Kent State

undergraduates and University of Wisconsin M.B.A.s, Kolb

(1984) found significant correlations between the LSI and

MBTI as shown below:

El: Positively correlated with RO; negatively correlated

with AE.

SN: Positively correlated with AC; negatively correlated

with CE and AE.

TF: Positively correlated with CE; negatively correlated

with AC.

JP: No significant correlations.












Using data from a study of 220 managers and M.B.A.

students, Kolb reported the following relationships among

MBTI attitudes functions and LSI scales:

E is associated with AE and AC (Converger).

I is associated with AC and RO (Assimilator).

S is associated with AE and CE (Accommodator).

N is associated with AC and RO (Assimilator).

T is associated with AC and AE (Converger).

F is associated with CE and RO (Diverger).

J is associated with AC and AE (Converger).

P is associated with CE and AE (Accommodator).

While Kolb (1984) has indicated that correlations

between group scores on the LSI and MBTI offer some empirical

indication of validity of relationships between the two

instruments, he cautioned that:

(B)oth the LSI and MBTI instruments are based upon
self-analysis and report. Thus we are testing whether
those who take the two tests agree with our predictions
of the similiarity between Jung's concepts and those of
experiential learning theory; we are not testing, except
by inference, their actual behavior. (p. 80)

Summary

Learning Styles research has not resulted in a unified

theory of how individuals learn. However, various constructs

have emerged which help with an understanding of the

individual preferences learners exhibit. The literature

review above was provided to establish three major learning

style constructs (theoretical, response-based, and

integrated); to show how learning style and an individual's

learning environment are related; to establish a rationale










66

for choosing for the present research two of the integrated

approaches to Learning style (the Kolb-McCarthy Learning

Style Inventory and the Myers-Briggs Type Indicator); to

explain the theoretical bases for the LSI and MBTI and cite

validity, reliability, and application research on the two

instruments; and to provide a theoretical comparison of the

LSI and MBTI in order to better understand expected

combinations and patterns of predictor variables from these

two instruments in the regression equations.




















CHAPTER THREE
METHODOLOGY

Population

The initial population available for the study was a

group of 188 students enrolled in the 10 sections of DVST

1012, Learning Skills, during the Fall, 1983 term at the

University of Arkansas. (DVST 1012 is a semester-long,

two-hour course offered through the College of Education and

is described in the next section.) Of the 188 students

enrolled in the course, 172 were available for the initial

assessment (described in a later section). The remaining 16

had either dropped the course or were unavailable for

assessment.

From the group of 172 students who received the initial

assessment, 128 were chosen to become a part of the study.

Students were chosen who were new freshmen entering the

University of Arkansas for the first time. Of the 128

students in the study 123 were 17-, 18-, and 19-year olds who

had graduated from high school the previous spring (1983),

but five students had one or more years intervening between

high school and entrance into college. The population is not

to be considered a representative sample of all entering












freshmen at the University. As explained in the next

section, students enrolled in DVST 1012, Learning Skills are

usually there because of an indication of poor preparation

for college. As will be shown later, however, not all

students in the study were academically underprepared in high

school--both the high school grade point average and ACT

scores for some of the students indicated high potential

ability.

There were 75 males (58.6%) and 53 females (41.4%) in

the study, and they represented all of the

undergraduate colleges and schools at the University except

the School of Nursing. (See Table 3-1, College Distribution

for Students in the Study.) The College of Business

Administration is overrepresented in the population when

compared to the entire entering-freshman population. For the

study population, the College of Business Administration

students represented 42.2% of the total; in the larger

entering-freshman class, approximately one third of the

students had selected Business Administration as a major. It

is instructive to note that the College of Business

Administration consistently reports above average rates of

probation, suspension, dismissal, and withdrawal, yet

underprepared students selected that college at a relatively

high percentage.

DVST 1012, Learning Skills

DVST 1012, Learning Skills, is one of five developmental

courses offered through the Department of Developmental










69
Table 3-1
College Distribution for Students in the Study

College Males Females Total

College of Agriculture & Rome Econ. 4 5 9

School of Architecture 6 0 6

Fulbright College of Arts & Sciences 17 22 39

College of Business Administration 33 21 54

College of Education 5 5 10

College of Engineering 10 0 10

School of Nursing 0 0 0

Totals 75 53 128










70

Studies (which the author chairs) in the College of Education

at the University of Arkansas. The courses offered are:

DVST 1012, Learning Skills

DVST 1031, Survival Skills

DVST 1041, Human Potential Seminar

DVST 1051, Career Development

DVST 1071, Independent Developmental Studies

All DVST courses are open to all students and are one

credit hour, except DVST 1012, which is a two-hour course.

Except in the College of Agriculture and Home Economics and

in the College of Education, the courses do not meet degree

requirements. However, in all colleges the courses carry

institutional credit, and they are used to compute term and

overall grade point averages. Students are assigned to one

or more of the developmental courses through summer

orientation, academic advising, faculty referral, in-house

remediation, and self-selection. The demand for the courses

is greater than the space available, and attempts are made to

hold space for students with the greatest need.

?or the most part, students in the Lcarning Skills

course are referred by faculty who advise students at

pre-registration. Students expressing poor preparation in

study skills, habits, and attitudes in high school are the

primary candidates for the course. Freshmen comprise

approximately 75% of the enrollment in the course.

The Learning Skills course provides instruction with

three major goals: study skills knowledge, practice in using

the learned skills, and the study attitudes needed for










71
successful college work. Major topics included in the course

are time management, concentration, memory, reading college

textbooks, listening and notetaking skills, preparing for and

taking exams, and, in general, identifying and practicing

good learning and studying behaviors. A point system is used

for grading and includes attendance, homework assignments,

class quizzes, Learning Lab work, special-topic workshops,

and a final exam.

Students are given a Learning Skills Assessment (LSA) at

the beginning of the term to determine potential problems in

reading, writing, mathematics, listening and notetaking,

spelling, and study skills. As a result of this assessment,

students are given the opportunity to attend the Learning Lab

and/or special-topic workshops as options in addition to

other class requirements.

Learning Skills Assessment

During the first week of classes, 172 of the 188

students enrolled in the ten sections of DVST 1012, Learning

Skills, were assessed as a group to determine potential

problems i; basic skills and study skills. The Learning

Skills Assessment (LSA) has eight components and all eight

were administered:

A. Reading rate and comprehension. An untimed

reading selection was presented to the students. They were

asked to record their reading time when they were finished,

and this time was converted to a words-per-minute rate. They

were then asked to answer ten questions to determine their

comprehension of the material.












B. Textbook reading. Four untimed reading

selections were presented to the students. They were asked

to answer three questions to determine their understanding of

subject matter, main idea, and supporting detail. The

answers to the 12 questions were then grouped and converted

to percentage figures for each of the three areas tested.

C. Listening skills. A recorded "mini lecture,"

approximately five minutes long, was presented to students

via a recording, and they were instructed to listen and to

not take notes. At the end of the recording, the students

answered five questions on the selection. Correct answers

were converted to a percentage correct.

D. Academic writing. Students were asked to choose

between two possible writing topics, both of which related to

freedom in the New World. Students were then asked to write

at least two paragraphs, and their essays were judged on

awareness of audience, conception of subject, writer image,

overall organization, paragraph construction, sentence

patterns, vocabulary, spelling, and grammar.

E. Notetaking. A sample -mini lecture,"

approximately seven minutes long, was presented to students

via a recording, and they were asked to take notes. The set

of notes was evaluated to determine if the student was able

to pick out the five main ideas and the ten supporting

details (two for each main idea) in the selection.

F. Spelling. Twenty commonly-misspelled words were

presented to students via a recording. Correct answers were

converted to a percentage correct.












G. Mathematics. Six basic mathematics questions

and five basic algebra questions were presented to students

in an untimed test to provide a gross-level assessment of

potential problems in those two areas.

H. Study habits. A 50-item, untimed study habits

questionnaire was administered to students to determine their

strengths and weaknesses in concentration, memory, organizing

time, studying textbooks, listening and notetaking, taking

tests, and motivation.

Data were available for research only for Sections A, B,

C, E, and H of the LSA, since those sections relate directly

to the goals of the course, and only those five areas were

used in the study. Sections D, F, and G were remediated, to

the extent possible given time and resources available,

outside of class through Learning Lab work and/or special-

topics workshops. (A primary value of the LSA, particularly

for sections not directly related to the course, is the

feedback given to students in individual conference sessions

by the classroom instructors.)

Assessment of Learning Styles and Psychological Type

During the fourth week of classes, 172 of the 188

students enrolled in the ten sections of DVST 1012, Learning

Skills, were assessed to determine their learning styles and

psychological types. The assessments took the first hour of

each two-hour class. When the class met for the second hour

later in the week, feedback and score interpretations were

provided to the class as a whole. The instruments used were

the Kolb-McCarthy Learning Style Inventory (LSI) and the










74

Myers-Briggs Type Indicator (MBTI, Form G). Additionally, a

self-report on the accuracy and relevance of the assessments

was asked of each student. (See Appendix B, Accuracy and

Relevance of Style and Type.) All assessment and feedback

sessions for the LSI and MBTI were provided by the

investigator.

Kolb-McCarthy Learning Style Inventory

The nine-item, forced-choice LSI was administered to

students by the investigator in each of the ten sections

after a brief explanation of the purpose of the assessment

was given and standard instructions were read. Students were

asked to rank order the four completion responses given for

each of the nine stem items, for a total of 36 responses.

The assessment was not timed, but students typically finished

in 15 minutes or less. Students were asked to record their

name, age, sex, date, and section number on the answer sheet.

Anonymity was assured by telling students that names were to

be used only to match against other information and that only

aggregate data were to be reported. The answer sheets were

later hand scored, and individual feedback reports were

prepared.

At the group feedback session (the next class period

later in the week) students were told of the underlying

theory of the LSI, and they each received a graph identifying

their perceiving/processing preferences and the resulting

learning style. The characteristics of each style were

explained to help students better understand themselves in

various learning situations, to stimulate their thinking












about a choice of major, and to better understand others.

They also received a handout explaining all four styles.

Myers-Briggs Type Indicator

The 126-item MBTI (Form G) was administered to students

by the investigator in each of the 10 sections of the course

after a brief explanation of the purpose of the assessment

was given and standard instructions were read. The

assessment was not timed, but students typically finished in

30 minutes or less. Students were asked to complete the

standard Form G answer sheet which asks for their date of

birth, the date of testing, sex, student status (yes or no),

preferred subject, occupation type (if any), and job

satisfaction (if employed). Anonymity was assured. The

answer sheets were later hand scored, and individual feedback

reports were prepared.

At the group feedback session, the underlying theory of

psychological type was explained, and students were given

individual type reports and descriptions of their type as

handouts. Type strengths were emphasized, and students were

advised to think of type as their unique way of interacting

with the world around them.

Self-Report of Accuracy and Relevance

At the feedback sessions, and as the results were being

explained, students were asked to rate the accuracy and

relevance of both the LSI and MBTI. Specifically, students

were asked to:

1. Guess their style and type after hearing a general

description of both but prior to receiving their individual

reports;












2. Assess the accuracy of their style and type after

hearing and reading a description of them;

3. State if there was a style or type they would rather

be; and

4. Answer relevance questions about the assessment and

feedback process pertaining to personal insight, application

to college work, application to personal life, understanding

of others, and thinking about career choices. Appendix B,

Accuracy and Relevance of Style and Type, provides a

description of the assessment and the results obtained.

Selection of the Population to Study

The problem statement in Chapter One, Introduction,

concerns the lack of success of entering college freshmen.

Thus only entering freshmen without prior college experience

were selected for analysis from the 172 students who received

the initial assessment. Students were chosen regardless of

their age as long as they met the entering-freshman criteria.

There were 128 students from the larger pool who met that

criteria. All but five of the students had graduated from

high school the previous spring (1983). As stated earlier,

the population was not sampled from the entire entering-

freshman class, and no inference of population representation

was drawn.

Selection of Independent and Dependent Variables

As stated in Chapter One, Introduction, college freshmen

often do not succeed in college even though high school grade

point averages and the results of standardized assessment

indicate they should. Such students demonstrate poor grades










77

in core-curriculum college coursework, low or failing grade

point averages, and a resulting academic status of probation,

suspension, or withdrawal. The selection of independent and

dependent variables was guided by the major research

question: How may advisors and others better predict success

or failure in core-curriculum coursework, freshman-year grade

point average, and resulting academic status by examining

factors in addition to high school grade point average and

standardized (ACT) assessment?

Independent variables

Five major independent variables were chosen for the

study:

High school grade point average (HSGPA)

American College Testing (ACT) scores

Kolb-McCarthy Learning Styles Inventory (LSI)

Myers-Briggs Type Indicator (MBTI)

Learning Skills Assessment (LSA)

The high school grade point averages were computed by

examining high school transcripts for the students.

Transcripts were available for 125 (97.7%) of the 128

students. In computing the HSGPA, all work performed by the

students for which they received a grade was used, since

overall high school grade point average was used at that time

for admissions and prediction purposes at the University.

American College Testing (ACT) scores were available for

116 (90.6%) of the 128 students. ACT scores were expected

for admission to the University when the study was initiated,

but students were not denied admission if the scores were not












available. (Beginning in fall, 1985 entering freshmen

without ACT scores will be denied admission.) In one

case, a Scholastic Aptitude Test (SAT) score was provided by

the student, and translated scores were used for ACT English,

Mathematics, and Composite. A translation formula was not

used for the Natural Sciences and Social Studies portions of

the ACT, since the SAT does not test these areas. (See

Langston and Watkins (1976) for an explanation of SAT to ACT

translation.)

Five sections of the LSA relate to goals of the Learning

Skills course (DVST 1012) and were chosen to comprise LSA

independent variables for the regression equations. The

sections chosen were:

A. Reading Rate and Comprehension

B. Textbook Reading

C. Listening Skills

D. Note Taking

E. Study Habits

Section A contains two scores, Reading Rate, in words

per minute, and Comprehension, expressed as a percentage, and

both were used in the analysis. Section B, Textbook Reading,

contains questions in three areas: subject matter, main

idea, and detail. The scores in this section are expressed

as percentages, and the three percentages were averaged into

one to yield a single textbook reading score for the study.

Section C, Listening Skills, has a single percentage score

and was used as is for the study. Section D, Note Taking, is

comprised of five main ideas and two details per idea, and










79

students' scores in this section were combined by multiplying

the number of main ideas they were able to identify times the

number of details identified to yield a single score which

could range from 0 to 50. Section E, Study Habits, is

comprised of 50 true-false questions relating to study habits

and attitudes, and these questions are subdivided into seven

major study skills areas: concentration, memory, organizing

time, studying textbooks, listening and notetaking, taking

tests, and motivation. Students' scores from the seven areas

were added to yield a study skills score which could range

from 0 to 50.

The Kolb-McCarthy Learning Style Inventory (LSI) was

used as an independent variable in the regression equations

by examining each of the scale scores (CE, RO, AC, AE) and

two combination scores (ACCE and AERO). Since the

combination scores are computed by subtraction, the resulting

scores ranged from -24 to +24, depending upon the relative

strength of the scale scores. Negative scores resulted when

CE was larger than AC and/or RO was larger than AE.

Additionally, the two combination scores, when "mapped,"

yield a specific learning style, and style was also examined

as an independent variable in an exploratory analysis of

cumulative grade point average, but not as a part of the

regression equations.

The MBTI yields an identification of psychological type,

e.g., ESTJ, INFP, etc. Raw-point scores are also available

on each of the eight single-letter attitudes or functions (E,

S, T, J, I, N, F, and P), and computed-point scores are












available on each of the four MBTI dimensions (El, SN, TF,

and JP). The single-letter raw-point scores and the two-

letter computed-point dimension scores were examined as

independent variables in the regression equations. Since

only one unsigned score is available from the MBTI on each of

the computed two-letter dimensions, a method of determining

direction as well as strength was employed. A dimension

score was considered to be negative if the student's

preference on that dimension was found to be E, S, T, or J.

A dimension score was considered to be positive if the

preference on that dimension was found to be I, N, F, or P.

The MBTI types, as well as certain two-letter combinations,

e.g., ES--, IN--, -NT-, etc., as recommended by McCaulley

(1977), were used in exploratory analyses of cumulative grade

point averages, but not in the regression equations.

Dependent Variables

Five major dependent variables were chosen for the

research consistent with the problem statement and research

question. These variables were considered to be a hierarchy

of academic outcomes, ranging from the most general to the

most specific:

1. Academic status at the end of the freshman year.

2. Cumulative grade point average at the end of the

freshman year.

3. Success or failure in the core-curriculum

disciplines.

4. Success or failure in the core-curriculum

departments.












5. Success or failure in the core-curriculum courses.

Academic Status. Students academic status at the

end of the freshman year was used as the dependent variable

in this analysis. Academic status was coded as either "1"

(good standing or returned to good standing) or "0" (placed

on probation, continued on probation, suspended, or

withdrawn). A step-wise regression analysis was run using

LOGIST (developed by Frank E. Harrell, Duke University

Medical Center) under the Statistical Analysis System (SAS)

at the University Computing Center.

Cumulative Grade Point Average. The cumulative

grade point average for all work attempted, computed at the

end of the freshman year, was used as a dependent variable in

this analysis. A standard step-wise regression analysis was

run using SAS at the University Computing Center.

Core curriculum. Success or failure in the core

curriculum was determined at three levels: discipline,

department, and course. At the highest level, four core

disciplines were chosen for analysis: Fine Arts, Humanities,

Natural Sciences, and Social Studies. (These disciplines are

the four major groupings of coursework in the Fulbright

College of Arts and Sciences at the University of Arkansas.)

At the next level, 11 departments were chosen from the four

disciplines. Departments were chosen only if 20 or more

enrollments from the study population were in that

department. (In many cases, students enrolled for more than

one course in a given discipline or department.) At the

lowest level, 14 courses were chosen for analysis. Courses












were chosen only if 20 or more students in the study were

enrolled in the course. Table 3-2, Disciplines, Departments,

and Courses chosen for Analysis, is a complete listing of the

core-curriculum dependent variables, and the number of

enrollments from the study population is shown for all three

levels. It can thus be determined from the table that the

128 students in the study represented 913 enrollments in the

4 selected disciplines, 732 enrollments in the 11 selected

departments, and 570 enrollments in the 14 selected courses.

(Refer to Appendix C, Titles of Courses Chosen for Analysis,

for the complete titles of the chosen courses.) The

dependent variables in the discipline, department, and course

analyses were coded either "1" for successful (A, B, or C) or

"0" for unsuccessful (D, F, I, or W) completion. A step-wise

logistic regression analysis using LOGIST was run under SAS.

Because of the large number of independent variables (30

in all), and to avoid losing large degrees of freedom during

the analyses, small sections of the independent variables

were entered at each computer run, and only if a variable

contributed significantly to the regression model was it

retained for the next run. No more than four new independent

variables were entered on each run, and nine runs were

required for each dependent analysis. The variables and the

order of entry into regression runs is shown in Table 3-3,

Order of Entry for Independent Variables. The regression

models were designed so that an independent variable (i.e., a

possible regressor) needed to be a significant correlate (p =

.1) with the dependent variable to enter the model for