• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 Abstract
 Introduction
 Review of related literature
 Methodology
 Results
 Summary, conclusions, and...
 Appendices
 References
 Biographical sketch














Group Title: Doctor Fox effect
Title: The Doctor Fox effect
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097461/00001
 Material Information
Title: The Doctor Fox effect a paired lecture comparison of lecturer expressiveness and lecture content
Physical Description: vii, 138 leaves : ill. ; 28 cm.
Language: English
Creator: Ramagli, Howard J., 1950-
Publication Date: 1979
Copyright Date: 1979
 Subjects
Subject: Lecture method in teaching -- Evaluation   ( lcsh )
Teachers -- Rating of   ( lcsh )
Student evaluation of teachers   ( lcsh )
Foundations of Education thesis Ph. D   ( lcsh )
Dissertations, Academic -- Foundations of Education -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 131-136.
Additional Physical Form: Also available on World Wide Web
Statement of Responsibility: by Howard J. Ramagli, Jr.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097461
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000089254
oclc - 05711122
notis - AAK4635

Downloads

This item has the following downloads:

doctorfoxeffectp00ramarich ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    Abstract
        Page v
        Page vi
        Page vii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
    Review of related literature
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
    Methodology
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
    Results
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
    Summary, conclusions, and recommendations
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
    Appendices
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
    References
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
    Biographical sketch
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
Full Text















THE DOCTOR FOX EFFECT:
A P'AiPED LECTL!URE COMPARISON OF LECTU:L.EF.
EXPRESSIVENESS AND LECTURE CONTENT









HEOWAR.D J. R.'IA'GLI, TP..


A. DISSEF.TATICN :RESENTED TO TH'E PRADUATI CCUN: :CI- CF
THE UNIVEF SIT' OVF FLORIDA
IN PAP.'-LL FTULF;LL..IENT OP THE .EQUIREMLETS Ff THE
DECREE OF DOCTOR OF PHILOSOPHY








NIVERSITY- OF FLORIDA


I Ic.



















Ackno-.ledgements should be brief and to the point. Therefore,

I would like to thank the members of my committee, both collective., ,

and indLvidually, for their understanding and know,.leage *..-ich helped

me through my tenure as a doctoral student-

To Dr. Harry Sisler, nmy thanks for his probing questions and

skepticism v.'hich always kept me on mi" to, s.

To Dr. James Algina, my thanks for his insightful approach to

the analysis of my data, and for taking the time to help me understand

what I was doing.

To Dr. John Newell, my thanks for his time and understanding

:n i.elping to see a problem or question with a new awareness of the

possibilities we confront each day.

I also extend my thanks to Jim Flavin in the Office of Instruc-

t:onal Resources and Lee Neal in the Educational Media Center at the

L'Un:ersity of Florida for their technical expertise and assistance in

th, implementation of this study.

Lastly, I thank Dr. Gordon Greenwood for having the patience of

Job with a doctoral student who burns the canidle on three ends. WAith-

out him this study would never have taken place.


A~i:~O!' L D ~. ~TzT














TABLE OF CONTENTS

ACKNOW LEDGEM.ENTS .................. ii

ABSTRACT . . . . . . . . . . . . .

Chapter

i. INTRODUCTION . .. . . .. ... .... ... 1

Purpose of the Study. ... . . .. . . .
Significance of the Study .. . . . . ... 3

ii. REVIEW OF RELATED LITERATURE .. . ... 6

The Use of Student Ratinos . . . . . .
The Validity of Student Ratines . . . . . 1
The Doctor Fox Literature . . . . . .

iii. M.ETHODOLOGY . . . . . . . . 19

Research Questions and Hy]potheses . . . .. 1
Design of the Study . . . . . . . 22
The Lectures . . . . . . . . .
The Content Dimension . . . . . .
The Expressiveness Dimension . . .. 27
Subjects . . . . . . . . . . 28
irstrunmentation . . . . . . . . 33
The Student Rating In:trument .. . . . 54
The Lecture Choice Insruet . . . 37
The Lecturer Adjectiv Checklist . . . 37
Statistical Analyses ... . . .... . 3
The, Lecnrre A A Analyses . . . . .. 38
Th e L --ecure b Ar.alyses . . . . . . -;
The Lecture Choice V'ariable Analsi . . 44
The Adjecr.i'.e Checklist Analss . . . 4-
Lir.-itati':ns of the Study . . . . . ... 4

.IV, R:SU- LTS . . . . . . .. 46

The Lectur A- Analysis Results . . .. . 46
The Lecture B Analysis Results . . .. . 9








TABLE OF CONTENTS iCcnt'd.)

The Lecture Choice Variable Anaiysis . ... 6
The Adjective Checklis: Analysis . . . .
Sun-mnary of the resultss ... . . .. .. 73

V. SUi:.lMAY, CONCLUSIONS, AND P.ECOLI'.LENDAT!CIOS 5

SunTm ar-r . . . . . . . . . .. . 7
Conclusions . . . . . . . . . .7
F.ecommendations . . . . . . ... 32

Appendices

A. ADMINISTRATION DIRECTIONS . . ..... . 3

B. IUCH CONTENT LECTUP.RE SCRIPT . . . . 4-

C. LOW CONTENT LECTURE SCRIPT . .. ... 9

D. LECTURE EVALUATION BOOKLET . . . .. .102

E. ADJECTIVE FREQUENCY DISTRIBUTION FOR
LECTURE CHOICE VARIABLE .. . . .. 110

F. ADJECTIVE FREQUENCY DISTRIBUTION FOR
RATING CHOICE VARIABLE . . . .. 121

REFERE;NCES ...... .... .......... .. 131

BIOGRAPHICAL SKETCH .... ........ ..... 137








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



THE DOCTOR FOX EFFECT: A PAIRED LECTURE COMPARISON
OF LECTURER EXPRESSI'.'ENESS AND LECTURE CONTE:-

By

Howard J. Ramagli, Jr.

June l'T7?

Cha..rnman. Gordon E. Green'xood
M.lajor Departmient: Foundations of Education

The purpose io this study '.'as to examine the Doctor Fcx effect

'.whLch :i. the primary influence of expressive teaching behaviors on

student ratings of instruction when students '..ere giv.'en the opoort.i-

nit'y to coLl.pa re lectures of v.arving le'.els of content and expressi.veness.

.Also, it -vas to examine what students say are instructor charactcr-

istics 'which cause them to rate one lecture presentation as better than

another, and thus, yie!ding insight into specific c.arr-cterist-:cs o. -.he

expressi-.enes s dimension of the Doctor Fox effect.

For this study copies of tapes used in the original Doctor Fo:x

scudie:" were used. The first five minutes of four cf the lectures Ihigh

ccntert-iich exoress-i'.eness, high content-low e:pres.siv'enes.s, lov.

cor..ten: -high expressiv-.eness, and Lc.'. co:tent-lov.. expre ssiv.er.ess) on

T h e Bicch-ltr istry of 'lemorv. -..'ere used. The high content ;serin-ents

'cocr i:ied six teaching rolnts; low coniteint segments contained no

rcacthiru points. i he four segJnents '.vere paired to L.id all 6 orde- ed-

p.air c'-irbinaticns, and thus, 16 treatment groups.







The 273 subjects were randomly assigned within intact classes

to :he 16 treatment conditions. The'' vie''.'ed Lecture A of the pair,

rated it, vie'.'ed Lecture B of the pair, rated it, selected .vhich of the

rw.o lecture presentations was the better presentation, and finally

selected adjectives, characteristic of the instructor, w.'hich influenced

their choice of the better presentation. The dependent measures v.'ere

a total student rating score ithe same rating instrument as in the ori-

ginal Doctor Fox studies, a lecture choice variable, and a choice/no

choice on each ci the adjectives on the check-lit.

Separate analys.es on the dependent .-ariables for Lecture A,

Lecture B, the lecture choice variable, and for the adjecti-es '.'vere

conducted. The 2 .x 2 content :.: express'venessi analysis for Lecture

A yielded the following results II) a significant interaction bet.v'een

content and expressiveness for total score on the student rating instru-

ment; (2 a significant difference in higher ratings of high content

lectures versus low. content lectures '.''hen expressiv.eness was high;

and 131 a significant difference in higher ratings of high expressiv-eness

versus lo:-. uxDressi'eness lectures for both levels of content. The

2 x x 2 x 2 content x expressiveness of Lecture A x contn.t x

expressiveness of Lecture B) analysis of Lecture B yielded the follow-

;ng results: (1) a significant interaction between content and expressive-

ness for Lecture B; (2) a significant difference in higher ratings of

high expressiveness versus low expressiveness lectures for both levels

of content for Lecture B; and (3) the high and low expressiveness








preconditions of Lecture A produced a significant difference in ratings

of Lect-re B for both levels of content for Lecture A. The chi-square

analysis of the lecture choice variable showed a significant frequency

distribution between choice of Lecture A versus Lecture B. The chi-

square analysis of the adiecti--es vyelded 21 adjectives which had sitni-

ficant frequency, distributions; IS adjectives showed a trend toward

relationship with the high epressivene.s lectures.

Because of the abo-ve results, the following conclusions were

drawn. i!; Hiih expressiveness of lecturing style has a major influ-

ence on student ratings of instruction. (21 After viewing two lecture

presentations, student ratings of the second lecture presentation were

influenced by the expressiveness of the first lecture presentation.

(3) The following adjectives are most descript.ive of the Doctor o:,-

express"-ve teaching stle- dramatiticnereti, inspiring, expressive,

:'-rcefu, interesting, stimulating, interested, warm, motiv-.' ing,

enthusias:tc, showmanship, enjoyable, hu.rrorous, sparkling: dynamic,

witty, and persuasive. 41i The Doctor Fox effect holds up when

students are g.':en the opportunity to compare r..o lectures varying in

c-'r.tenr. and expressi-.eness.














Chapter 1

INTRODUCTION


The Dc'ctor Fox effect originall: appeared in the literature in

1970 (Ware et al., !970 a and b). The itda for the Doctor Fox effect

stemmed frcn-m the .'.ork of Er-:.in Goffman, specificall,- his notion that

expressti.'e behav'.or nma, influence an audience as much as or mrre

than substance '.hen there is little time or reason for the aud.enrce to

evaluate the presentation (Goffmn-an, 1 '9. There arec..o oasic prer-

iseS b:nir.i the Doctor Fox effect. One premise is that highly expres-

s:'.'e bcha-.tior o~' 3 teacher ..ill have primary influencee on ard be a

determining "actor cf student ratinet of ir.strrictior. The .co.-.d pren--

ise i z an ...utt rot.vth c the first premi rn It s.tate. that stuce'nt-s '.'ill

be unable to .J'Liit uish betr'een lectures of different le'.'els of content

:'..en cx cs to a lecturer '.. ith a highly exprc sst:.'e ectLrin,_ style.


Purpose of the Study

The purpose of this study is to examine students: ability to "see

through" the Doctor Fox effect when given the opportunity to compare

lectures of varying levels of content and expressiveness. Also, it is

to c::arr.ine what students say are the factors which cause them to rate

one lecture presentation as better than another. Specifically, the

study will yield an insight into the specific characteristics of the







expressiveness dimension of the Doctor Fox effect. Also, the purpose

is to retest a portion of the originally stated Doctor Fox effect, that

being 'hat students are unable to distinguish levels of content -.vhen the

content is accompanied by a highly expressive lecture presentation.

Specifically, the purpose of this study '.'.as to examine the follo.v-

ing research questions.

I. Do levels of lecturer expressiveness and levels of lecture

content interact '..'ith each other in affecting student evaluation .:,f lec-

turer effec:i'eness?

2. After .-iev.inc a second lecture presentation v.arying in con-

tent and,,or expressiveness from the first lecture presentation, do the

level : of content and levels of expressiveness of this second lecture

interact v'.itn the le'.v1e of content ano levels o3 expressiveness of the

first lecture presentation in affecting the student evaluation of lecturer

e!iectiv.enes s

Are there differences in student evaluations of lecturer

effect:i.eneess for lectures .v.hich are h:gh in expressiveriess as opposed

t :thoze ..hizh a:-e lo.-.. in e:-p-es si.enes s

4. Doe- the level of content in a lecture presentation affect

student evaluation of lecturer effectiveness

5. Are there differences in student choices of adjectives, char-

acteristic of the instructor, to describe what they feel is a better

lecture pre eni.at.on?







Significance of the Studc.y

V'hile WVilliams and ''are (19771 have atte.npted to e '.a lish the

Doctor Fox effect over t'.vo consecutive leet-res, th-', hai not te.tz:d

the effect by: providing students '.'ith an opportunity to cmnp.are lec-

turer expressiveness and lecture content. If students are able to iden-

tify the express.iveness element of a lecturer and yet are unable to

identify content when given the opportunity to compare t.'.'o lectures., her

Lt may be that the Doctor Fox effect can be sho,.n to be a strong deter-

miner of student ratings. If this is the case, then the highly e:.-pre.-

si-'.e behaviors might be taught to teachers to improve their lecturing

technique.

On this, there are t'.,'o points of significance 1. which relate speci-

fically to the college setting. The first point'is that there is a strong

rn-ov.mer.r in higher education institutions to set up and to develop pro-

grams for improving the quality of college teaching (Boyd and Schie-

tinger, 1976; Gaff, 1975; Seldin, 1975). Among the techniques being

used for faculty development programs, systematic observation

attempts to identify specifically the teaching behaviors of the college

professor. This study continues the efforts to identify what are suc-

cessful teaching behaviors. As such. the findings of this study have

implications for the continued efforts to improve college teaching and

to provide effective faculty development programs.

The second point of significance for this study concerns the

increased use of student ratings and other teacher effectiveness data









in personnel decisions, such as promotion, tenure, and salary Ln-

creases, r-.ade bL college administrators about teaching faculty

(Greenwvood, 1977; Greenw'.ood and Renner, 1,75i. Since ::he expres-

sive behaviors of the Doctor Fox effect have been shown to he related

to both student rating- and to student achievement, the results of this

study, particularly those results concerned w:-th these expresl'sve be-

haviors, could have strong implications for researchers studying the

validity of student ratings and for college a.ni inis trators concerning

the validit, of the use of student r3.ting data sources for personnel

decision-making.

Moreover, this study has important implications for the educa-

tional psychologist. The educational psychologist is interested in the

nature of teaching, the effect of teaching upon the student in the class-

room, and the effect of teaching on learning in the classroom. This

exp:erui ent sheds some light on part of the effect of teaching style and

presentation in a lecture situation. It is designed to retest that portion

of the Doctor Fox effect dealing w.itn the impact of expressi-.eness cr.

students' perceptions of teacher performance. If this experiment

shows that the observations of the effects cf expressiveness on student

ratings a--e "-lid, then the next step -..'ould be to examine the validity

of the effects of express siveness on learnt.-i and achie"-ement. This

is of tremendous importance for teacher training and faculty develop-

ment at the college level.







Thus, this study deals .vith the effect of expressiveness and

content level on student perceptions of teacher presentation as meas-

ured by' a student rating instrument. Also, it deals '..ith '.vether or

not the student can identify those elements of expressiveness in the

Doctor Fox effect. in suntrnary, this study: is designed to fill a gap

in the Doctor Fox literature, since no studies providing for an imme-

diate comparison bet.veen lecture presentations have been done to

date: to provide a possible confirmation of the Doctor Fox effect,

that is, the finding that high expressiveness of lecturing style has a

primary influence on student ratings and that high expressiveness

tends to mask content level of a lecture presentation; to establish

empirically the high expressive behaviors these beha.viors identified

should be of use in improving the lecture style of profee sors'; and

to add further to the research recardLng the validity of student ratings

of college professors, particularly as they, are used in administrative. e

decision-making (e.a. promotion, tenure, and salary increases).














Chapter II

REVIEW OF RELATED LITERATURE


In this chapter the literature related to the Dcctor Fox effect,

as .ell as the Doctor Fox literature is re,.-ie'.ed. Specifically, the

literature on the use of student ratings in colleges and universities,

the literature on the -validity of student ratings, and the Doctor Fox

literature are examined. These three bodies of literature present a

set of interrelated issues. The increased use of student ratings of

faculty b'y college administrators for personnel decisions is of inter-

est because of the inconclusive results in the validation of student

ratings. This inconclusive' eness of the validity of student ratings can

be examined, and partially can be explained by the findings on the

Doctor Fox effect, '.vhich conclude that student ratin2s. of instruction

are primarily influenced by a high degree of lecturer express.ieness

in lecturing style and, secondarily, that this high expressiveness

tends to mask student ability to identify lecture content.


The Use of Student Ratings

One issue raised in relation to student ratings of instruction is

their continued, increased use in administrative decision-making,

spectfii:llv regarding faculty promotion, tenure, and salary increases.







This issue is paradoxical since the use of student ratings of instrucr'un

often is not combined with a program of faculty development and instruc-

tional improvement iSmith, 1?76). T.vo surveys serve to point out this

increase in use for administrative decision-making.

Seldin (1975) surveyed colleges and universities across the

countr-. as to whetherr or not they used student ratings and .econdiy,

how' they used student ratings -.hen the,. used them. More than one-

half of the liberal arts colleges state that they use student ratings, in

some form or another, to evaluate faculty performance. Further,

more than 601, of all colleges and univ. rsities state that the primary

use :or their student ratings .vas to evalua:.t faculty teaching perform-

ance for use in administrative decision-making.

Eoyd and Schietinger i!'76) surveyed colleges and un:-.ersit
in the southern region of the U. S. for the Southern Regional Education

Board. Their results confirmed the trends cited by Seldin (1975) and

showed that, at least in southern colleges and universities, the trends

were becoming more widespread. They found that 62% of doctoral

institutions use student ratings of instruction for making administra-

tive decisions about the faculty. In addition, they found that large uni-

versities have little or no concern for the use of student ratings of

instruction for faculty development programs. Still further, they

found that the primary users of student ratings for decision-making

purposes are academic deans and department chairmen. Finally, they

found that 88% of all the institutions surveyed use student ratings of

instructions for decision-making.









This use of student ratings for decision-making has prompted

comments from man.- individuals and researchers. For example,

M.cKeachie il '69 i raised the question of the use of student ratings for

personnel decision-making and contended that students are qualified

to evaluate teaching effectiveness. Centra i 1'7> review'.ed trie prob-

lem of the use of student ratings for decision-mnaking with rega:-d to

retention, promotion, and tenure. He reported a survey in .%which

students stated that the'.- believe that trieir ratings of instruction

should be the basis, at least in part, for decisions relating to promo-

tion. Greenw.ood and Renner 1 '75, rev.ie'..ed the literature on the use

of student ratings and point out that the lack of vaildity of student

ratings places their use in question. Further, they; called attention

to the fact that there is some agreement that some measure of student

outcome, i.e. student achievement data in some form, should be the

basis of validation for student ratr.gs of instruction. Yet, the; con-

cluded that the amb ialence of the results of validation research

raises questions of the reasonableness for the use of student ratings

in adcn-inistrative decision-making.

Another problem w.vith the validity of student ratings is created

when students are informed of the use to be made of their ratings of

the instructor. Aleamoni and Hexner (1973) found that when students

were told that their ratings were to be used for promotion and tenure

decisions, these students gave significantly different ratings of their

instri.ctors -..hen these ratings were compared to students who were








not told of the use of the ratings. Abrami et al. (1976) found that

students rate their instructors more pcsitv.'elv .'hen the',' are told

that the faculty association, as opposed to the stud-ent association,

is sponsoring the e-.'aluations of the instructors. Lcnne',' I 'PT73' con-

cluded that seniors at a small liberal arts college '.-'.,uld prefer that

any evaluations of faculty, that they. did, be cued for feedback in aiding

an instructor to improve the course and his teaching; they did not '..'ant

their evaluations used for adrn-inistrat,.:e deci-sion-n',akinL Alexander

i19 '3) found that, if students thought their rating couldl d affect the sal-

ar.' of the instructor, 47'r, of the students satisfied '.'ith an instructor

'.'.ould change their rating of the instructor, and 9.?. S', of these would d

make it higher. Of the 29. rS'r '.vho '..'ere dissatisfied, '34. 4'2r '.vould

lo.'.'er their rating of the instructor. Alexander concluded that students

'.ill punish or re'.'.ard an instructor by a change in ratings, if the stu-

dents know the ratings will affect that instructor's salary.

There are two important points which have been raised thus far.

One point is that there seems to be a consensus of opinion that there

is an increased use of student ratings of instruction as a primary part

of administrative decision-making with regard to faculty promotion,

tenure, and salary increases. The second point is that there is a

growing concern over the validity of student ratings, particularly in

light ci how the ratings are affected by students' perception of how

these ratings will be used. This second concern, the validity of stu-

dent ratings, is considered next.








The '.'aliditv of Student Ratings

As indicated above, there is an ambivalence Ln the results of

validity research conducted on student ratings of instruction. How-

ever, what seems clear is that the most accepted method for validating

student ratings is to compare then- w'.ith student achievement scores.

Rosenshine '1971) and Cohen et al. 1i-731 :'eview'ed v.alidation studies

of student ratings against achievement data. While they pointed out

the inconclus1.iveness of the validation results to date, the' found that

the primary validation instrument or measure is somrre measure of

student final outcome, either an achievement test, a final examination,

or a final erade in the course. T'v.ent. of the studies reviev'.ed used

achievement gain as the validation outcome measure. Cook and Neville

i11 .71 advocated the use of both achievement outcome ias a direct

measure and student ratings of instruction las an indirect outcome) to

combat some of the problems that relate to the valld:r.- of student

ratings.

Costin et al. I 1'71 revie'.v.ed 119 publications that dealt .''ith the

validity of st-dent ratings. The.' found that single achievement mneas-

ures. primarily student grades, are correlated w'.ith student ratings

and genc rally yield correlations :n the neighborhood of 35. They con-

cluded that student ratings do not provide a complete assessment of

teaching effectiveness. In another review of five studies, McKeachie

et al. (1971) found both positive and negative correlations between stu-

dent ratings and achievement. They concluded that validation should be

dcne .--ith student learning gains in relationship to student ratings.








Some studies of the relationship between student ratings and

achievement gain yielded mixed results. Rodin and Rodin 11'721

reported a -. 75 correlation between student ratings of teaching assis-

tants and achievement gain. Gessner 11973) iound a high positl'e cor-

relation bet'.een student ratings and achie.vemrent gain on a national

achievement test and a low positive correlation bet.veen student ratings

and a departmental examination. In addition, Gessr.er found that stu-

dent ratin-s are related to students' previous knowledge of the subject.

Greenwood et al. 11976) found no correlation between four types of

student ratir. instruments and achievement gain in mathematics.

Turner and Turner (1974) found negative correlations between student

ratings and student achievement. McGuigan (1974), in correlating

achievement (achievement gains, final examination score, and course

grades) with student ratings, concluded that how much a student learns

is a preferable way (as opposed to student ratings) to evaluate faculty.

A different approach to the validity of student ratings was to

compare other types of faculty evaluation ratings with achievement

and student ratings. Fcllman and Merica (1973) compared student

ratings, supervisor ratings, peer ratings, and self ratings with student

achievement and found correlations in the magnitude of .35. They con-

cluded that student ratings are as good as the other types of ratings, in

terms of validation against student achievement, in evaluating teaching

effectiveness. Blackburn and Clark (1975) found that student ratings

are as effective as peer, administrator (department chairman), and








self ratings in evaluating teaching effectiveness, w.'hen all the abc.'e

ratings are compared to student achievement.

Touq et al. 11973) tried still another approach to validating stu-

dent ratings. They de,'eloped a student rating instrument .vhch paral-

leled the Flanders Interactional Analysis System. The subject teachers

w'..ere observed systematically by expert observers and these results

.vere compared to the student ratings of the same instructors. They

found that student ratings agreed *.vith the systematic observation on

four out of the nine interactional categories on the Flanders system.

A different t'.vist to the above approach was used by a number of

researchers. In their research the, attempt to validate certain aspects

of student ratings .vhch seem to be more effective in assessing faculty

performance. Frey et al. (1Q75) found that pre' ious student accom-

plishment, presentation clarity,, and organizational planning correlate

highly :''it! a final examination score. They concluded, along with

Sullivan and Skanes (1'741 and Centra (I177i that organizational and

presentation skills are a more valid indicator of teaching effectiveness

than measures of student/teacher interaction. Centra (I '77) reported

that global ratings of teachers, ratings of lectures, ratings of courses,

and ratings of course objectives and organization are the student rating

types most highly related to student achievement. WAhitel-, and Doyle

(174l !Q76) and Lev.enthal et al. 119761 concluded that student ratings

of teaching effectiveness are affected by previous student experience of

instructors and the selection of teachers based on the teacher's ability

and.'or reputatlor, respectively.








From- the above review, the foilo..'ing summary statements about

the literature can be made. (I There is an increased use of studentt

ratings or instruction by ccileae administrators for personnel decision-

making affecting faculty promotion, tenure, and salary increases. (21

The student ratings of instructors are influenced by the student percep-

tion of how these ratings '.ill be used. 131 A number of approaches to

validating student ratings have been used. As of this ratingng, the results

of these validation studies are mi ed and inconclusiv-e. 14- This incon-

clusivr.eness of validation results has increased the concern over the

validity of the use of student ratings by administrators for personnel

decision-making. This question of the -.alidity of student ratings also

has been examined by the direct, experimental manipulation of teaching

style. A rev.iewv. of this literature is presented next.


The Doctor Fox.: Literature

The Doctor Fox effect originally appeared in the literature in

'1970 iWare et al., 1970 a, b). The idea for the Doctor Fox effect stem-

med from the work of Erving Goffman, specifically his notion that

expressive behavior may influence an audience as much as or more

than substance when there is little time or reason for the audience to

evaluate the presentation (Goffman, 1959). Expressiveness, or seduc-

c:.*.ness, has been identified with lecturer characteristics such as

enthusiasm, humor, friendliness, charisma, and personality in various

correlational. factor analytic, and experimental studies of teacher be-

havior as related to effective teacher behaviors (Rosenshine and Furst,








'971: Ccffman, 1954: Isaacson et al., 1964; Solomon, 1966. Coats and

Smidchens, i966; Harry and Goldner, 1972; Finkbeiner et al., 1i731.

Ware ui'?74, defined expressiveness as follows

a number of different terms have been used,
all of which appear to refer to the seduction-
expressiveness aspect of teaching effectiveness,
including: "charisma ', "dynamism", "emotional
appeal", 'friendliness", "interest", "personality'',
"seduction", "stimulation' and "warmth".
Seduction-expressiveness is operationally defined
as the extent to which these mannerisms are
present in the lecture presentation. (p. 41

Some research has been conducted which sheds light on the effect

of expressiveness in teaching. Carpenter and Hadden (1964) and Fish-

bein i1972) found that students are unable to detect lack of substance in

seductive lectures. Ero'.ne and Anderson (1974) concluded that there is

a lack of student sensiti'itrv to variations in instructional style, specifi-

cally that students could not detect variations in communication structure

as related to their degree of knowledge acquisition. Martin l1963j, who

studied hieh school students, found that students who heard enthusiastic

teachers learn more and have more favorable course attitudes than

students of less enthusiastic teachers. Coats and Smidchens (1966),

.'ho used no outcome measures, concluded that dynamic lectures result

in higher student achievement than less dynamic lectures. And finally,

Zelbvy I1974), who took a different perspective on the idea of expressive-

ness in lectures, concluded that faculty can teach so as to obtain more

favorable student ratings for particular instructor and course character-

:stics enthusiasmm.







The first of the Doctor Fox studies '.'as reported b' Naftuitn et al.

l1973). The conclusion from this study wa.s that three groups of experi-

enced educators were satisfied %with the amount the., learned in a lecture

presented by Dr. Fo::, despite the fact that the lecture itself w'.as lo..- in

content coverage land high in expressiveness). Kaplan 1974) criticized

this original study on two fronts:

1. that it '.-as not possible to determine if the Doctor Fox

effect was due to the style of the lecture presentation

or due to the impressive credentials given to Dr. Fox,

and

2. that there was no item on the lecture evaluation question-

naire which asked about learning, that is to say. that there

was no way to establish whether the subjects had the

illusion of having learned.

The Doctor Fox studies .were an extension of the above research

and criticism; these studies took an experimental approach to the prob-

lem of the effect of the expressiveness of teachers. They' did this b'

varying the amount of expres-siveness with amount of content coverage

in a lecture. Specifically, the design, w.vhich involved the use of a pro-

fessional actor as Dr. Fox, '.w.-as a 3 x 2 factorial design with three

levels of content coverage Ihich, medium, and lo..) and r.vo levels of

expressiveness high and low) for the lect-res ,'.are, 19741. The lec-

tures were videotaped and shown to six groups of students.

Ware and Williams (1975a) reported that student ratings of an

instructor are not sensitive to the amount of information covered in








lectures, ever. though student achievement is affected. Specifically they

found that, under high seductive iexpressivenessi conditions, the stu-

dents could not distinguish between a high content coverage lecture and

a low content coverage lecture. Further, they found that difference in

lecture content coverage and lecturer seductiveness appeared to affect

both student ratings and student test performance. In another reporting

of this effect, Ware and Williams '197 Sb and l'977) found that differences

in faculty information-gi-.-ing are easier to detect using student rating

scored when faculty members are less enr.thusiastic (low in seduction)

in their lecture presentations. Also, they, reported a statistical confir-

mation (using discriminant function analysis) of all the above results,

that is the majority of error in detecting differences in lecture content

is a result of bias Ln student ratings due to seductiveness of presentation.

Wilha.ns and Ware reported t'.,o efforts in attempting to "see

through" the Doctor Fox effect. In the first (19761, they pro'.ndeo stu-

dents '.'.ith varying amounts of monetary incentive for learning in the

lectures. They concluded that sensitizing students to the content of the

lectures did not improve student accuracy :n the rating of content cover-

age. In the second case (1977), they tried exposing students to two con-

secutive lectures of the same type in content coverage and seductiveness.

They found that accuracy in rating of content coverage did not improve

when students were exposed to a second lecture under the same condi-

tions, i. e. providing students with additional exposure to the lecturer

doe-_ not enhance the students' ability to "see through" the Doctor Fox

effect.




17


However, throughout all the above studies, one result remained

constant. Lecturer expressi'.eness, or seduction, has a major impact

on student ratings of instruction, and in everry stud,' reported, there

was a significantt difference in the rating of high e::pressiveness lectures

as opposed to low expressiveness lectures. This result is the primary,

premise of the Doctor Fox effect.

Two recent studies confirmed this primary premise (Perry, et al.,

1979; MIeier and Feldhusen, in press). However, Frey (1979) has

questioned these results b,, contending that bias in student ratings, due

to .his primary. premise, can be reduced by, altering student ratings

instruments so that they contain more items geared to'..ard learning

outcomes. W'are and W'illiams t1979) responded to this criticism by

pointing out that Frey's criticism, based on his focus that specific items

relating to content coverage be included in Doctor Fox research, is not

valid. The,, pointed out specific items (knew subject matter .'.ell,

increased student knowledge of the subject, inspired confidence in

knowledge of the subject, and organized and presented the subject

natter well) on their rating instrument which show Frey's criticism to

be in error. Further, they contended that their results and the results

of Meier and Feldhusen (in press) showed that students are unable to

aist.nguish content level under varying experimental conditions. Finally,

they reaffirmed their original contention that the Doctor Fox effect

dealt primarily with the influence of expressiveness on student ratings,

and that content recognition by students was a secondary issue.







From the above review of the Doctor Fox literature, it can be

seen that, to date, there ha:.e been no studies providing students '.ith

an opportunity to compare immediately lectures var':ing in content and

in expressiveness. Thus, by designing a paired lecture comparison

study of lectures varying in content and in expressiveness, an additional

testing of the Doctor Fox effect, both the influence of expressiveness on

student ratings and the masking of content by expressiveness, is pos-

sible. In addition, the literature, as re.iew.ed, does not provide any

specific experimental information as to .vhat are the high expressive

behaviors of the Doctor Fox effect. By giving students an opportunity

to select adjectives, characteristic of the instructor, that are descrip-

tive of a better lecture presentation, a descriptive analysis of the hign

expressive beha'.iors is possible. Finally, any study examining student

ratings -.vill add to the validity literature, particularly a study examin-

ing the influence of lecturer expressiveness on student ratings.













Chapter III

.. METHODOLOGY


This chapter contains a restatement of the questions addressed bv

this study, a discussion of the overall experimental design .ith attend-

ant hypotheses, a specific discussion of the independent "ariables of

content and of expressiveness, a description of the subjects and of the

research instrumentation, a discussion of the statistical analyses for

this study, and a statement of the limitations of the study.



Research Questions and H.ootheses

The followi-ng are the research questions addressed by this stud,..

1. Do levels of lecturer effectiveness and levels of lecture con-

tent :nteract '.vith each other in affecting student evaluation of lecturer

efiect'ive.nn : s

2. After viewing a second lecture presentation varying in content

and/or expressiveness from the first lecture presentation, do levels of

content and expressiveness of this second lecture interact with levels of

content and expressiveness of the first lecture presentation in affecting

the student evaluation of lecturer effectiveness?

3. Are there differences in student evaluations of lecturer

effectiveness for lectures which are high in expressiveness as opposed

to those which are low in expressiveness?








4. Does the level of content in a lecture presentation affect

student evaluation of lecturer effectiveness"

:. Are there differences in student choices of adjectives,

characteristic of the instructor, to describe 'chat the'-, feel is a better

lecture ores entation

As a result of the above research questions, the iollov.in major

null hypotheses were developed to be tested by this study.

For the first lecture presentation (Lecture Al,

1. There -'.ill be no differences on the dependent variables

bet.veen the four treatment conditions, t.v'o levels of content and t,.vo

levels of expressiv.eness. iNo content by e:.:pressiv.eness interaction.

2. There will be no differences on the dependent v-.ariables

between the two conditions of content. (No main effect for content.

3. There will be no differences on the dependent '.,ariables

between the two conditions of expressiveness. (No main effect for

expressiveness.)

For the second lecture presentation I Lecture B),

1. There will be no differences on the dependent variables

between the sixteen treatment conditions, two levels of content and t'.vo

levels of expressiveness for Lecture A and two levels of content and

two levels of expressiveness for Lecture B. (No content (A) by

expressiveness (A) by content (B) by expressiveness (B) interaction.

2. There will be no differences on the dependent variables

ber..-een tl.- treat-iment conditions for Lecture A, two levels of content








and two levels of expressiveness, and the treatment conditions for Lec-

ture B, t.vo levels of content and tw-o le,'els of expressiveness.

,a. No contentiA) by expressive:iess AIA by contentlBl interaction.

b. No contentlAl by expressl.-eness(Ai. by expressi'-enesstB)

interaction.

c. No contentlAl bv content(B) by expressivenesstB) interaction.

d. No expressi'.eness(A) b-. content(B) by cxpressi.venessiBl

interaction.

e. No contentl.a) b, contentrB: interaction.

No content(A) by expresoiv.'eness )l interaction.

g. No expressiv'eness(A) by contentlB) interaction.

h. No expressi.'eness IA) by ex. pressi.eness(B interaction.

3. There .vill be no differences on the dependent variable be-

tw.eer. the four treatment conditions for Lecture A, two levels of content

and tw.o levels of expressiveness. (No contenttA) by expressivenessiA)

interaction. )

4. There will be no differences on the dependent variables be-

rtw.en the two conditions of content for Lecture A. (No rain effect

for content(A). I

5. There .will be no differences on the dependent variables be-

twveen the two conditions of expressiveness for Lecture A. iNo main

effect for expressiveness(A).)

b. There will be no differences on the dependent variables be-

tween the four treatment conditions for Lecture B, two levels of content








and t'..'o levels :f expressiveness. (No contentlBi by expressi.enessiB;

ir.teraction.

7. There '.'.'ill be no differences on the dependent variables be-

t.v.een the t-..o conditions of content for Lecture B. iNo main effect for

content BL.

8. There will be no differences on dependent variables be-

tween the two conditions of expressiveness for Lecture B. (No main

effect for expressi'.enessiB )


Destin of the Study

In the design for this study, students .vere presented a pair of

videotaped lecture presentations, asked to rate each of the lecture pre-

sentations, asked to choose which of the two lecture presentations was

the better presentation, and asked to select those adjectives which

were influential in their choice of which of the two lecture presentations

was the better presentation. The various components of the data to be

collected from this procedure gave rise to various design considerations

for this study.

The design for this study contains two separate designs, one for

the first lecture presentation and one for the second lecture presenta-

tion. There are two reasons for the separate designs for the two lec-

ture presentations. First, the research questions presented above can

best be answered by this separate design tactic. Second, the first lec-

ture presentation acts as a treatment precondition for the second lec-

ture presentation, thus making it possible to see if there is an effect







for the independent variables from the first lecture on the second lecture

independent .-ar:ables.

The design for the iirst. lecture presentation is presented in

Fi-ure 1. This is a 2 :: 2 design 'vith r'.vo levels of content (high and low)

and tw'.'o levels of express.iveness (high and lo.l.v. This design is a repli-

cation of the original Doctor Fox study'. (Ware, 1974) without the medium

content lev.el. A specific discussion of the lecture presentations and of

the independent variables of content and expressiv.eness is found in the

next section in this chapter.

The design for the second lecture presentation is presented in

Figure 2. This is a 2 x 2 x 2 x 2 design with two levels of content and

two levels of expressiveness for the second lecture presentation, as

well as two levels of content and two levels of expressiveness corres-

ponding to the first lecture presentation and acting as the treatment pre-

conditions for the second lecture presentation.

The data collection for these designs went as follows. (Note: see

Appendix A for the administration directions for the following sequence.)

The first lecture was presented on videotape and then rated by the

students. The second lecture was presented on videotape and then rated

by' the students. Upon completion of the second rating, the students

were asked to choose which of the lecture presentations was better.

After making the choice, they were asked to indicate why this presenta-

tion was better by completing a checklist of teacher behaviors and

characteristics. Thus, there were four dependent variables in this




24


Content


High




EF


Lov.


High


Low


Figure 1.


Content
(Lecture B)


High




Content
SLecture A-




Lov.,


High


Lov.'


Expressiveness Exprressiveness
lecture BLecture B ectre


High


High


n -




'0

n -





'[




__ __


Figure 2.


I


Lo,."








design- two student ratings of lectures, a choice variable, lecture A

or lecture B, and a reason variable, consisting of the choices made on

the checklist.


The Lectures

The lectures used in this study are the original Dr. Fox lectures

as used by Ware (1974). These lectures are on i.-deotape and are cop-

ies of the original lectures. However, instead of using the entire

tent, minute videotaped lecture as did Ware, onl, the first five min-

utes of a given lecture .were used. To prevent '.new.er fatigue, and v'et

to allow enough vie'.ving time for comparison, the same five minute seg-

ment frcm each of the lectures w'as used.

Each lecture segment used had a specific level of content and a

specific level of expressiveness, both content and expressiveness could

be either high or low. The combinations of content levels and expres-

siveness levels produced four lecture segments. These were, including

their average length of time in parentheses, as follows:

High content High expressiveness (5 minutes, 6.4:5 seconds

High content Low expressiveness 15 minutes, 23.29 seconds

Low content High expressiveness (5 minutes, 8. 68 seconds)

Low content Low expressiveness (5 minutes, 25. 66 seconds)

These lecture segments were then paired together in all possible

combinations, including each of the four lectures followed by itself, and

were edited together on the videotape. Each lecture segment in the pair

was preceded on the videotape by a fifteen second segment which








displayed a sign on the screen identifying the lecture segment as either

"Lecture A" or "Lecture B". The total number of combinations of these

four lecture segments w.as sixteen. The combinations are displayed

belo'.


Lecture A


followed by


Lecture B


1. High content-High expressiveness


High content -Lo'.'

Low. content-High

Lo'.' content-Lo'.'

High content-High

High content-Low'

Lo'.' content-High

Lo'.v content- Low

High content-High

Low' content-High

High content-Low.v

Lo'.v content-Lo'.v


expressiv.eness

ex:pres si.eness

expressi'.eness

express si'.enes s

expressiveness:

express i.enes

expressivenes s

expre s i.eness


ex:pres si .enes s

express sivenes s

expressiveness


13. High content-High expressiveness

14. Lo'.. content-Low' express v'.eness

15. High content-Lov. expressiveness

16. Lo'. content-Hih expressiveness


High content-High expressi-.eness

High content-Lo.'. expressiv.eness

Lo.'.: content-High expressiveness

Lo'.' cc ntent-Lo'..' expressiveness

High content-Low. expressiveness

High content-High expressiveness

Lo' content-Low e:xpressi'eness_

Lo'.'. content-High expressiveness

Lo'.'. content-High expressi-.eness

High content-High expressiveness

Lowv content-Low expressiveness

High content-Lo-. expressiveness

Lo'.v content-Lo'. express .ieness

High conntent-High ex.presst iveness

Low.. content-High expressiv.eness

High content-Lowv expressiv.eness


It should be noted that the above combinations also take into account that

the order of presentation of the lecture segments must be controlled for,

and thus e:-ch pair of lecture segments has its reversed order pair

directly, follo.'.ing it in the list, beginning with pair number 5 and 6.




7 -



Videotaped lecture segrr.ents '..ere used for several reasons.

First, this procedure allov..ed for the control of the various, essential

conditions and treatment levels in the experimental design. Second, the

original Doctor Fox studies '.':ere done '.vith videotapes, and any replica-

tion of the Doctor Fox effect wouldl d be biased if videotape w.'ere not used.

in addition, the use of the original tapes from the Doctor Fox studies

made it possible for implications from any replication of the Doctor Fox

effect. Third, it is difficult to control a live presentation. No amount

of rehearsal and practice can guarantee the same qualities in the pre-

sentations every, time. Thus, the -.ideotape '..as used for control of

unwanted variability .

The Content Dimension

The content dimension .'as defined b:y Ware i19741 as contatr.ing

a specified number of teaching points for the conditions of the content

dimension. A high content lecture contained Iii the entire lecture '.vere

used) 26 teaching points. in this study the high content lecture segment

contained 6 teaching points. The lowv.. content lecture contained Iif the

entire lecture '.ere used 4 teaching points. In th:s study the lov'. con-

tent lec-ure segment contained 0 teaching points.

Verbatim scripts for the high content and low'. content lectures can

be found in Appe.tdices B and C, respectively. The six teaching points

for the high content lecture are underlined in Appendix B.

The Expressiveness Dimension

Expressiveness was defined by Ware (1974) as follows:

S. a number of different terms have been used,
all of which appear to refer to the seduction-







expressiveness aspect of teaching effectiveness,
including "charisma", "dynamism", "emotional
appeal", "friendliness", "interest", "personality",
"seduction", "stimulation", and "warmth". Seduction-
expressiveness is operationally, defined as the
extent to whichic h these mannerisms are present in the
lect-jre presentation. ip. 4i

Thus, the high expresstr.e lecture contained ail or most of the elements

of the above definition of expressiveness. The lo'w expressive lecture

contained fe-.., if any., of the above elements.

Note that in Appendices B and C that there are phrases in the

lecture texts which h are in parentheses. These phrases are omitted

from the lov'. expressive lectures for the respective content levels. In

addition to the above information on the expressiveness dimension,

Ware 1 '974, pp. 61 64) gives a detailed description of the training of

the actor to display, the expressiveness elements cited abo-e.


Subjects

The subjects for this study ..'ere obtained from those students

registered in undergraduate courses of the Foundations of Education

Department at the Universit'- of Florida. The undergraduate courses

..'ere Human Growth and Development 18 sections, Social Four.ndations

of Education 14 sections), Educational Psychology 12 sections), and

History. o'. Education in the United States (2 sections Each of the six-

teen classes -."as randomly assigned to one of the sixteen lecture pairs

stared abo-e. Thus, this -'as a random assignment of intact classes.

Ind,,-idual ass5o= .-unent for randcinizaticn -.va. not done and therefore

certain orecautions r.-'.'st be taken before conducting the analyses of

the data.








These precautions are to insure that the partitioninE of the be-

tween groups '.ariance is not confounded by the classes nested withinin

treatments. If the v.ariance attributable to these nested classes is large

enough to be significant i.hen tested against the variance for the subjects

withinn the groups, the interpretation of results 'vill be confounded.

Th:s can be tested b, conducting a 2 x 2 analysis of '.arnance .with the

classes nested for Lecture A. This w'.as done and an F value of .T7

at 12 and 262 degrees of freedom was obtained. This --alue is not sig-

nificant and the classes nested within groups does not need to be parti-

tioned out of the total -.ariance for the model tested in the Lecture A

anal.-sis.

In all there were 278 subjects in this study with a mean age of

21.4 years. The median age was 19.8 years. Of the 278 subjects,

33. 8% were females and 16. 2% were males; their median grade point

average was the equivalent of a B+. Juniors comprised 54.7% of the

sample; seniors, 25. 2% of the sample; and sophomores, 17. 3% of the

sample.

The primary colleges of registration in the sample were Educa-

tion (32. 4%), University College (28.8%), Arts and Sciences (14.7%),

and Physical Education (13. 3%).

The principal academic majors of the sample were as follows:

Elementary education (14.7%), nursing (11.2%), speech pathology

(10. 1%), physical education (7. 9%), and special education (7. 2%). A

more detailed breakdown of these data is presented in Table 1.




30



Table 1

Sample Population Frequency Distributiors for Age, Sex,
Grade Point Average, Year in College, College of Reg-s-
tration, and Academic -.taor


Frequenc-, Cum Freq Percent Cum Percent

Age
Age


13
56
155
216
23 '
247
249
256
259
261
263
265
267
26?
270

273
274
276
277
2 7 0


4. 676
15.468
35. 612
21. 942
S. 273
2. 87S
0. 719
2. 518
1. 079
0. 71Q
0. 71 'Q
0.719
0. 719
0. 7 19
0. 360
0. 360

0. 360
0. 71 9
0. 360
0. 360


4. 676
20. 144
55. 755
77. 698
85. 971
88. 849
89. 568
92. 086
33. 165
'?3. 885
*4. 604
95. 324
a6. 043
96. 763
97. 122
97. 432
8. 201
38. 561
'-'. 2'81
99. 640
100. 000


S e:,:
Sex

Female 233 233 33.813 3. 813
.iale 45 2-3 16. 187 100. 000








Table I ICont'd. )


Frequency Cum Freq Percent Curi Percent

Grade Point A-.era~e

GPA

3.6 4.0 34 34 12.230 2. 230
3.1 3. 5 9S 132 35.252 47.482
2. 6 3. 0 99 231 35. 612 3. 094
2. 1 2.5 44 275 15. S27 9 921
1.0 2.0 3 27S 1.079 100. 000


Year In Colleee

Year

Freshr.-ian 1 1 0. 360 0. 360
Sophomore 48 49 17.266 17.626
Junior 152 2 1 54.676 72. 302
Senior 70 271 25. 10 97.482
Po-t Graduate 7 273 2. 518 100. 000

College of Registration

College

Agriculture 6 6 2. 1 5S 2. 153
Architecture & 8 14 2. S73 5. 036
Fine Arts
Arts & Sciences 41 55 14.74S 19.734
Business Admrin. 2 57 0.719 20. 504
Education 90 147 32.374 52.373
Health Related 2 149 0. 719 53. 597
Professions
Journalism 5 154 1.799 55. 3c6
Physical Educ. 37 191 13. 309 68.705
University Ccl. 80 271 28. 777 97.482
Undecided or 7 278 2. 518 100.000
unknown








Table 1 tCont'd.)


Frequency Curn Freq Percent Cum Percent

Acadmrc .Major

Major

Early Childhood 6 6 2. 158 2. 158
Education
Elementary Edu- 4i 47 14.743 16.906
cation
Foundations 5 50 1.079 17. 986
Secondary Educ. 20 70 7. 194 25. 180
Spec-ial Educ. 24 4 8. 633 33. 813
English 5 99 1.799 35. 612
Nursing 31 130 11. 151 46. 763
History 3 133 1.07' 47.842
Mathemnatics 4 137 1. 439 49. 21
Art Education 7 144 2. 518 51.799
Music Education 5 149 1.799 53. 598
Ps.ycholog 10 159 3. 597 57. 195
Speech Pathology 28 187 10. 072 67. 267
Agricultural Ex- 4 191 1.439 6. 706
tension Educ.
Physical Educ. 22 213 7.914 76.620
Recreation 15 228 5.396 82. 016
Health Educ. 10 238 3. 597 85. 613
Other 29 267 10.430 96.043
Unknc'wn or 11 278 3. 957 100. 000
Undecided








By way of comparison, the original sample of the Doctor Fox

studies iWare, 1974; Ware and Williams, 1975) contained 207 students.

That sample contained 67'~) females and 33'r. males. The median age

,..as '9. 9. There '.v.ere 21"', freshmen, 30') sophomores, 2S5'~ Juniors,

185'. seniors, and 3~, graduate students. The primar- colleges of reg-

istration of the students w.'ere liberal arts and sciences 142'.), education

Il l' ), engineering (11 '), and business 17'i1. Thus, the present study

has, in comparison, a higher percentage of females, of upper classmen,

and of individuals registered in the college of education. Other than

these, the samples are comparable. Research rev-e'.'.s of student

characteristics and student ratings (Rosenshine and Furst, 1971;

Cohen et al., 19731 sho'.ved that any differences in student ratings due

to student characteristics '.'.ere produced by undergraduate versus grad-

uate student ratings. The rtwo samples discussed here have a similar

percentage of graduate students. There were no differences, found in

the reviews, attributed to the other characteristics used to describe

the sample of the present study.


Instrumentation

The data collected in this study were student ratings of two lec-

ture presentations, a choice of which of the two presentations was the

better presentation, and a list of adjectives describing the lecture each

student chose as the better lecture. The data were collected in a lec-

ture evaluation booklet. This booklet is presented in Appendix D. The

booklet in Appendix D contains the student rating instrument, the








lecture choice instrument, and :he adjective checklist. These are

discussed separately below.

The Student RP.ing Instrument

The student rating instrument is the same instrument used in the

original Doctor Fox studies (Ware, 1974). It is a modified student

evaluation instrument as dev'eloned by Pohlman !19751, with an internal

consistency of total score rating of 9?, using the KR-20 formula.

This rating instrument, adapted by the original Doctor Fox researchers

from a ore-existing instrument, is designed to measure the character-

istics of the expressiveness and content dimensions of the Doctor Fox

effect.

The reliability analysis for the student rating instrument for this

study is presented in Table 2. The total reliability presented is a

cooled KR-)0 reliability across treatment groups. This pooling is

done to avoid the between group differences, caused by any treatment

effect, 'which might affect the reliability of the instrument. The reli-

ability analysis for this study w.vas done using the Statistical Package

for the Social Sciences (SPSS) (Nie and Hull, 1977i.

The pooled reliability for the student rating ir.strument on the

Lecture A presentation is .9425. Tr.e pooled reliability for the student

rating instrument on the Lecture B presentation is .9530. These reli-

abilities are relatively high, and account for 89' and 915 of the vari-

ance in the total rating scores on the student rating instrument These

compare favorably with .'hat is reported by Ware (1974; a. reliability







Table 2

Student Rating Instrument KR-20 .eliabil it'
Indi.vdual Treatment Groups


tor


Treatment SS di SS df ra
Group between subjects residual

Lecture A


1 219.6144 1o 119.091 272 .96 1
2 173.0377 27 250.4270 45 .9149
3 80. 1482 11 75. 1852 187 .9456
4 94. 667 12 110.7179 204 .9312
5 108. 1241 10 133. 8170 272 9245
6 93.2353 16 157. 0000 272 .9010
7 219. o07S 16 155.8039 272 .9583
8 116. 5238 13 109.4762 221 .447
S302.9 924 21 231.8712 357 .9550
10 127. 5000 15 174. 2500 55 196
11 7 1354 15142. 3021 255 .3 07
12 136.9889 19 147. o111 323 .9366
13 103. 1620 11 8.9213 17 .9436
14 31.0404 10 73.8o87 170 .9464
15 239. 6732 1o 167. 5033 272 589
16 255. 090 28 213. 1379 476 .9509

Total 2421. 5193 262 2365. 943 4454 .9425








Table 2 iCont'd. I


Treatment SS df SS di ra
Group bet.veen subjects residual

Lecture B


1 136. 24 3 16 111.9869 272 .9517
2 04. 228 27 225.3790 459 .9349
3 102. 111 11 66. 3889 17 .9618
4 137. 291 12 102. 3243 204 .9563
172. 3856 16 119. 1433 272 .9593
6 37.9020 16 92.8039 272 .9379
7 115.7190 16 126.7516 272 .9356
S105.0794 13 99 3492 221 9444
9 265.7702 21 184. 6339 357 .9591
10 103. 7743 15 105. 9132 255 9400
11 135. 7222 15 17. 777 25 .9490
12 229.7194 19 135. 0306 323 654
13 31.4444 11 90. 33 9 1S7 .9344
14 115.6465 10) 55. 2626 170 .9719
15 149. S 16 135.9608 272
16 303.4023 23 138. 1494 476 9641

Total 2451.7859 262 1958.2503 4454 .9530


al.R -2
br =


L-r*. = '1032







of 96 .vhich accounted for 92'r) of the variance in the total rating

scores on the same Lnstrunment.

The Lecture Choice Instrument

The lecture choice instrument is a simple, dichotorrous forced

choice question. The students '.'ere asked to select .vhich of the t'.vo

lecture presentations the',' sa-.v .;.as the better presentation. They, 'vere

asked to check a box corresponding to Lecture A or a box corresponding

to Lecture B. They were asked to choose only one of the two lectures.

Of the 273 subjects, 277 responded as directed to this question. The

analysis of this question is presented in Chapter 4.

The Lecturer Adjective Checklist

The lecturer adjecti.'e checklist consists of forty-se. en adjec-

tives. These adjectives are drawn from the literature on character-

istics of effective teachers IEridges, et al., 1971. Hildebrand et al.,

1971, and WVotruba and V/right, '975).

The characteristics identified *.'ere compiled into one list and

all multiple word descriptions '.vere reduced to single '.ord adjectives.

This list '..as then compared to the descriptions in V.'are lr ?74, pp. 4,

16-18, and 62-63) of the expressive or seductive lecturer. All adjec-

tives which were not in the Ware sources were rejected as inappropri-

ate. The remaining adjectives were then given a random order and

the checklist was complete.

The intent of comparing these effective teacher characteristics,

identified in various sources, with those of Ware was to identify the








teaching behaviors associated with the Doctor Fox effect. This was

done in the desire to expand the information on the teaching behaviors

in the Doctor Fox effect.

Table 3 presents the adjective checklist and its sources, both

from the literature and from Ware. It should be noted that one adjec-

tive, seductive, recei-ed no choices in the study, and -.'ill not be discus-

sed hereafter.


Statistical Analyses

The statistical analyses for this study are presented in this sec-

tion. First, there is a consideration of the analyses appropriate for

the hypotheses related to Lecture A. Second, the analyses appropri-

ate for the hypotheses related to Lecture B are discussed. Thirdly,

there is a brief discussion of the analyses appropriate for the lecture

choice variable. Fourthly, the analyses appropriate for the adiecti.ve

checklist are presented. It should be noted that all statistical analyses

were done with the Statistical Analysis System (SAS) (Earr et al.,

1976).

The Lecture A Analyses

The design and hypotheses for Lecture A require that the analysis

be in the form of a factorial analysis of variance. However, because

the number of subjects per cell was not equal and because this unequal

n was the result of natural groups of unequal size, an unweighted

means solution is appropriate to analyze the data. (Timm and Carlson,

1975).








Table 3

The Lecturer Adjecti'.e Checklist and its Sources


Adjective Ware (974): Bridges et a!. Hildebrand W'otruba and
(197 1 et a. 19? ) W right i1975)


Dramatic LC

Knov.'ledgeable C X X X

Thorough D X

Coherent D X X

Well-prepared D X X

Energetic L X

Lo2ical C X

Explanatory D X

Inspiring LC

Understandable D X X

Well-paced D X

Expresii,.'e LC X

Forceful LC X

Inte resting LC X X

Variety L X X

Stimulating L X

Interested LC X X X

Warm LC

Scholarly C X

Orderly C X X




40



Table 3 iCont'd. i


Adjective Ware l'?74).- Bridges et al. Hildebrand V.'ctruba and
I l'?-71 et al. i '?T11 W right i 1975



Stylized LC X

olotiv.ating LC X X

Masterful D X

Enthusiastic LC X X X

Showmanship LC X

Vital LC X

Concise D X

Lucid D

Enjoyable C X X

Clear D X X X

Invent ve LC X

Cultured C X

Confident C X X

Poised C X

Humorous C X X X

Sparkling LC X

Comoetent C X X

Charismatic LC

Conscientious D X

Dynamic LC X








Table 3 iCont'd. )


Adjective Ware i1974' Bridges et al. Hildebrand Wotruba and
1971 et al. (1971) Wright ( 197 i



Well-groomed C X X

Organized C X X X

Careful D X

Witty,' C X X

Unique LC X

Persuasive LC X


: In the column for Ware, the L refers to his list o0n page 4, the C

refers tc his chart on pages 16-18, and the D refers to his

descriptions on paces 62-63.







The variance for the betv.veen groups is partitioned for the content

main effect, the expressiveness main effect, and for the content/expres-

siveness interaction. Should there be a significant interaction, the

between groups variance must be repartitioned for simple main effects

of content at both levels of expressiveness and expressiveness at both

levels of content. This analysis provides ans'.vers to the follo'.vin2 three

null hypotheses-

1. For the first lecture presentation I Lecture A), there '.ill be

no differences on the dependent .ariables betv.'een the four treatment

conditions, t-.o levels of content and tv.'o levels of expressiveness.

2. There 'ill be no differences on the dependent variables for

the tw.o conditions of content.

3. There .v'ill be no differences on the dependent variables for

the t'.vo conditions of expressiveness.

The Lecture B Analy.ses

The Lecture B analyses are similar to those for Lecture A. The

unw.eighted means solution is the appropriate analytical method to be

employed. The analysis is more complicated because the design con-

tains four independent variables, content and expressiveness for Lec-

ture A, and content and expressiveness for Lecture B. The partition-

ing of the variance is complicated by the number of possible interac-

tions between the four independent variables. However, any significant

interactions call for a repartitioning of the variance for the variables

involved in that interaction. The approach to this partitioning remains

the same as that used for Lecture A.







This analysis provides answers to the following null hypotheses

1. There '.'ill be no differences on the dependent variables be-

t.veen the sixteen treatment conditions, rto levels of content and tw.-o

levels of expressiveness for Lecture A and r.vo levels of content and

rwo levels of expressiveness for Lecture B.

2. There will be no differences on the dependent variables be-

r.veen the treatment conditions for Lecture A, tvo levels of content and

t.o levels of expressiveness, and the treatment conditions for Lecture

B, two levels of content and t-.vo levels of exoressiveness.

3. There will be no differences on the dependent variables be-

tr'.een the treatment conditions for Lecture A, r.to levels of content and

two levels of expressiveness.

4. There .'ll be no differences on the dependent variables for

the r.o conditions of content for Lecture A.

5. There will be no differences on the dependent variables for

the r.o conditions of expressiveness for Lecture A.

6. There will be no differences on the dependent variables be-

r.veen the four treatment conditions for Lecture B, two levels of content

and two levelS of expressivenes .

7There w'..ill be no differences on the dependent variables for

the two conditions of content for Lecture B.

8. There will be no differences on the dependent variables for

the two conditions of expressiveness for Lecture B.








The Lecture Choice Variable Anal sis

The lecture choice variable is a dichotomous variable inv'.olving

the choice of either Lecture A or Lecture B as the better lecture. There

are tw.o types of analyses that are performed on this data.

The first analysis is to compare the lecture choice selection of

each subject with the lecture that each subject rated higher. This is

done by using a Spearman rank order correlation coefficient. This

reveals the relationship betv.'een the lecture choice and the higher rated

lecture. It also provides insight into the accuracy of the subjects'

lecture choice versus the rating choice.

The second analysis appropriate for this data is a chi-square

analysis which .would look at the research hypotheses on content main

effects and expressiveness main effects. This analysis furthers the

information on content and expressiveness main effects obtained from

the analyses of the student rating instruments.

The Adiective Checklist Analysis

The statistical analysis appropriate for the adjective checklist is

chi-square analysis. The data are sorted on which lecture the subject

chose as the better lecture and then a chi-square analysis is conducted

on each adjective for the lecture chosen. A chi-square that is not sig-

nificant .'vouid mean that the frequency of choice of a particular adiec-

ti.ve is random and is not affected by the various combinations of con-

tent level and expressiveness level.








An additional note on the analysis of the lecture choice -.ariable

and adjecti-.e checklist is appropriate here. In addition to the analyses

mention for both, analyses of the rating choice .ariable I','.hich lecture

.vas rated higher) are performed. These analyses are identical to

those performed %.vith the lecture choice :ariable and the adjective check-

list. These analyses provide additional information on the .alidit-. of

the lecture choice variable and the adjective checklist.


Limitations of the Study

1. The interaction of Dr. Fox s..'ith the sample is IL1nted.

Because his expressiveness is a fixed quality (on videotape), he has

no '.vay, of interacting .vith the sample classes.

2. The dependent measures (the student rating instrument and

the choice/checklist instrument) are not pretested in'this population.

Therefore, reliability and validity are not previously established

(although the rating instrument does have these measures from the

previous Doctor Fox studies).

3. The generalizability from this sample to a larger population

is hampered by the above factors.

4. The generalizability of the results of this study is limited

to lecturing behaviors of college teachers.














Chapter I-

RESULTS


In this chapter, a presentation of the results of this stud;,' is made.

First, there is a presentation of the results of the analyses for Lecture

A. Second, the Lecture B analyses '.'ill be presented. Third, there is

a consideration of the results of the analyses related to the lecture

choice variable. Fourth, the adjective checklist analyses are reported.


The Lecture A Anal.ysis Results

The analyses for Lect. re A are presented in Table 3. These

results are discussed below in the order that the statistical hypotheses

were presented in Chapter 3.

1. There will be no differences on the dependent variables in

this case, the student rating instrument tctal scored between the four

treatment conditions, two levels of content and two levels of e:-:pres sive-

ness. (No content by expressiveness interaction. AI .s can be seen in

Table 3, the overall between groups .urm of squares has a significant F

ratio at the .05 level (F = 22.80 for 3 and 274 degrees of freedom). In

addition, the F ratio for the content by expressiveness interaction is

also significant at the 05 level IF = 6.71 for 1 and 274 degrees of

freedom). Because of the two results above, the first null hypothes is








Table 3

2 x 2 Anal'ysis of Variance for Student P.ating Instrument
Total Score on Lecture A


Source di Sum of Squares M[ean Square F

Anova for 2 :: 2 Design


Between Groups 3 11268.0675 3756.)225 2'2. 80
il.odel)
Content 1 904. 9677 904. ?677 5. 4 '-
Expressive- 1 8778. 2679 3773. 267? 33. 23
ness
Content x Ex- 1 1104.3126 1104.3126 6.71
pressiveness
Within Groups 274 45142. 4S29 164.7536
IError)

Total 277 56410. 5504

Simple Main Effects


Expressiveness 1 9234.2145 9234.2145 56.
at High Content
Expressiveness 1 1620.5162 1620.5162 9.73***
at Low Content
Content at High 1 1924.0999 1924.0999 11. 69***
Expressiveness
Content at Low 1 5.1983 5.1938 .03
Expressiveness


ip<.05, F(3,274,.05) = 2.64)

** p<. 05, F(1,274, .05) = 3.88)

*** (p<.01, F(1,274,.01) = 6.74)








is rejected and instead, the alternative hypothesis that there are differ-

ences on the dependent variables for the four treatment conditions is

accepted.

Ho.we'.er, these results require that to examine the remaining t.vo

hypotheses, a simple main effects analysis must be conducted on the

data (Kirk, 1968). This v'.as done and the results are reported in

Table 3.

2. There will be no difference; on the dependent variables be-

tr.'een the two conditions of content. (No main effect for content. ) To

examine this hypothesis, content must be analyzed at both levels of

expressiveness. This '.w'as done. When expressiveness '.'as high, the

content dimension produced a significant F ratio at the 01 level (F -

11.6'? for 1 and 274 degrees of freedom). When expressiveesss '.'as

low, the content dimension did not produce a significant F ratio. Thus

the null hypothesis for content is partially accepted and partially rejec-

ted. The hypothesis is accepted '.'hen content is examined at a low

expressiveness level. The hypothesis is rejected '.'hen content is exam-

ined at a high level of expressiveness, and the alternative hypothesis,

that there are differences on the dependent variable for the conditions

of content '.'hen expressiveness is high, is accepted.

3. There .v.ill be no differences on the dependent variables be-

tv'.een the two conditions of expressiveness. iNo main effect for expres-

siveness. ) To examine this hypothesis, expressiveness must be ana-

lyzed at both le.ve:s of content. This was done. When content was high,








the expressiveness dimension produced a significant F ratio at the .01

level (F = 56.05 for 1 and 274 degrees of freedom. When content was

low, the expressiveness dimension produced a significant F ratio at the

.01 level (F = 9.73 for I and 274 degrees of freedom,. Thus the null

hypothesis for expressiveness is rejected completely, and the alterna.-

tive hypothesis, that there are differences on the dependent v.'ariables

for the two conditions of expressiv.eness, is accepted.

Table 4 presents the rating instrument means for the four treat-

ment conditions for Lecture A. Figures 3a and 3b portray ; graphically,

the results mentioned above. Figure 3a shows that there is a difference

between the means for the high expressiveness groups of the content

dimension, while there is ver, little, ii any,, difference be.veen the low.

expressiveness groups for the content dimension. Figure 3b shows the

differences between the means for both content levels for the expres-

siveness dimension.

Sammarizing the results for the Lecture A analy,'es, there are

significant differences bet'.veen high expressiveness and low. e::pressive-

ness for both content lev'.els, and there is a significant difference be-

t-ween the content groups if the expressi.'eness w as high.


The Lecture B Analyses Results

The analyses for Lecture B are presented in Tables 5 and 6.

Th-ese results are discussed below in the order that the statistical

hypotheses were presented in Chapter 3.




5I



Table 4

Student Rating Instrument Mean Scores for
Lecture A Treatment Groups


Standard
N .lMean Deviation

High Content
High Expressiveness 68 56. 5735 14. 2466
Low Expressiveness 90 41. 1333 11.0293
Low Content
High Expressiveness 62 48.3710) 14. 1160
Low Ex-pressiveness 53 41. 5172 12. 2564


60 -


Mean
Rating
Scores


50 --


40 4-


Low Content
High Content


High


Low


Expressiveness Groups


Figure 3a


i




























High Content


-Lo'w Content


High Low
Content Groups


Figure 3b


Mean
Rating
Scores


60 -

55-

50-

45 -

40 -




52


1. There '.'.ill be no differences on the dependent variables (in

this case, the student rating instrument total score) bet.veen the sixteen

treatment conditions, t.vo levels of content and t..'o levels of exDressive-

ness for Lecture A and t-vo le'.els of content and t,.vo levels of expres-

siveness for Lecture B. (No content IAl by. expressiveness (A) by con-

tent IB) by expressiveness (B) interaction. ) As can be seen Ln Table 5,

the overall between groups sum of squares has a significant F ratio at

the .05 level (F = 11.25 for 15 and 262 degrees of freedom). In addition,

the F ratio for the content (A) by expressiveness (A) by content IB) by

expressiveness (B) is not significant at the .05 level iF = 2. 13 for I and

262 degrees of freedom). Because of the second result above, the first

null hypothesis, that is there are no differences between the sixteen

treatment conditions of the experiment (No four-way interaction, is

accepted.

2. There will be no differences on the dependent variables be-

tween the treatment conditions for Lecture A, two levels of content and

two levels of expressiveness, and the treatment conditions for Lecture

B, two levels of content and two levels of expressiveness. In order to

make a decision on the acceptance or rejection of this hypothesis, each

of the possible combinations of interaction between the treatment condi-

tions for Lecture A and the treatment conditions for Lecture B must be

examined.

a. Content (A) by expressiveness (A) by content (B) interaction:

This was not significant at the 05 level (F = 2. 98 for 1 and 262 degrees

of freedom).








Table 5

2 x 2 : 2 : 2 Analysis of Variance for
Student Rating Ir.strument Total Score on Lecture B


Source df Sum of Squares .lean Square F


Beetr.een Groups 1 5
I .lodell
Content 1
(Lecture A)
Expressiveness 1
(Lecture A'
Content 1
SLecture B)
Expressiveness 1
(Lecture B'
Content (A) x Expres- J
siveness IAl
Content (A) Con- 1
tent (B)
Content (A) x Expres- I
siveness (BE
Expressi.veness IA) : 1
Content IB)
Expressiveness (A) x 1
Expressiveness IB)
Content IBI x Expres- 1
siveness lB)
Content IAl x Exores- 1
siveness IA) x
Content (B)
Content (A) x Expres- 1
siveness (A) x Ex-
pressiveness (B)
Expressiveness (A) x 1
Content (B) x Ex-
pressiveness (B)
Content (A) x Content 1
(B) x Expressive-
ness (B)
Content (A) x Expres- 1
siveness (A) x Con-
tent (B) x Expres-
siveness (B)
Within Groups (Error) 262

Total 277


28434. 1371

10. 7918

2261. 6510

5'-. 55S3

19277. 8 53

26. '317

586.7555

16. 5 51J

262.252S

140. 0010

1096. 59?49

502.4264


343.3416


44.0272



450.7489


358.5234




44133.5751
72567.7122


1895. 6091

10. 7?918

2261. 6510

59. 5583

19 277. 8'953

26. '3 17

586. 7555

1 6. 35'51

262.2528

140. 0010

1096. 5?4?

502. 4264


343. 3416


44. 0272



450.7489


358.5234




168.4488


*(p <.05, Pus, 262, .05) = 1.71 05) = 3.88)


11. 25:

.06

13. 43'-

.35

1 14. 44;:

.16

3. 48

.10

1.56

.83

6. 51:

2. 98


2.04


.26


2. 68


2.13


. 05) = 3.88)


.(P <.05, F(15,262,.05) = 1.71








Table 6

Student Rating instrument Mlean Scores for
Lecture B Treatment Groups


Standard
Group N MIean Deviation


High Content LLecture BI
High Expressiveness Lecture B)
High Content Lecture A)
High Expressiveness ILec-
ture A)
Lov. Expressiveness Lec-
ture AI
Lo,.v Content i Lecture A)
High Expressiveness Lec-
ture A I
Lo'., Expressiveness (Lec-
ture A)
Lo'. Expressiveness lLecture BI
High Content (Lecture AI
High Expressiveness
SLecture A)
Low'.' Expressiveness
ILecture A)
Low Content (Lecture A)
High Expressiveness
(Lecture A)
Low Expressiveness
(Lecture A)
Low Content (Lecture B)
High Expressiveness (Lecture B)
High Content (Lecture A)
High Expressiveness (Lec-
ture A)
Low Expressiveness (Lec-
ture A)
Low Content (Lecture A)
High Expressiveness (Lec-
ture A)
Low Expressiveness (Lec-
ture A)
Low Expressiveness (Lecture B)
High Content (Lecture A)
High Expressiveness (Lec-
ture A)


17 60r. 1765

17 71. 5294



16 c6o.4375

11 6i4. 181




17 40.0412

28 41.6786



17 44.1675

20 49.9500





22 57.7727

29 64.5172



12 53. 0000

14 66.7143




12 49.0000


12.3806

9. 9443



11. 15 3

14. 427Q




13. 0260

11. 6684



12. 9819

14. 7522





15. 0931

14.0804



12. 9263

12. 0667




11. 5444




55



Table 6 (Cont'd.


Group


Lo'.v Expressiveness (Lec-
ture A)
Lo'. Content I Lecture A.
High Expressiveness ILec-
ture A)
Low.v Expressiveness i Lec-
ture A)


N XMeap

15 51. 2 _.


40. 6)5?3


Standard
Dev.iation

12. 7619



11. 4098

1 4. 3786,







b. Content (A) by expressiveness (A) by expressiveness IB) inter-

action: This .as not significant at the 05 level IF = 2. 04 for 1 and 262

degrees of freedom).

c. Content (Ai by content (B) by expressiveness ',B interaction:

This was not significant at the .05 level iF = .26 for I and 262 degrees

of freedom).

d. Expressiveness (A) by content iB) by expressiveness (B) inter-

action: This wvas not significant at the 05 level (F = 2. 63 for 1 and 262

degrees of freedom).

e. Content (A) by content IB) interactions: This was not signifi-

cant at the .05 level 'F = 3.43 for 1 and 262 degrees of freedom).

f. Content I(A by expressiveness (B) interaction: This wvas not

significant at the 05 level (F = 10 for I and 262 degrees of freedom).

g. Expressiveness (A) by content (B) interaction: This was not

significant at the 05 level (F = 1. 56 for 1 and 262 degrees of freedom).

h. Expressiveness '(A by expressiveness IB) interaction: This

w'as not significant at the .05 level (F = .33 for 1 and 262 degrees of

freedom). Since ncne of the above results are significant, the second

null hypothesis, that there are no differences betv'.een the treatment

conditions for Lecture A and the treatment conditions for Lecture B,

is accepted.

3. There will be no differences on the dependent variables be-

tween the four treatment conditions for Lecture A, two levels of con-

tent and two levels of expressiveness. (No content (A) by expressive-








ness (A) interaction. ) The F ratio for this interaction is not significant

at the 03 level (F = 16 for I and 262 degrees of freedom Therefore,

the above null hypothesis is accepted.

4. There will be no differences on the dependent variables be-

tween the two conditions of content for Lecture A. (No main effect for

content (A)l. The F ratio for the content (Al main effect is not signifi-

cant at the .05 level (F = .06 for I and 262 degrees of freedom). There-

fore, the above null hypothesis is accepted.

5. There will be no differences on the dependent variables be-

t-.veen the r.tvo conditions of expressiveness for Lecture A. (No main

effect for expresiv.eness(A). ) The F ratio for the expressiveness (A)

main effect was significant at the .05 level (F = 13.43 for I and 262

decrees of freedom). Because of the above result, this null hypothesis

is rejected and the alternate hypothesis, that there is a difference be-

tween the expressiveness treatment conditions for Lecture A is accepted.

The results for 4 and 5 above also are shown by an examination of

Table 7 and Figures 4a and 5a. Figure 4a shows graphically that there

is little or no difference between the means for the two content condi-

tions across expressiveness conditions. Figure 4b shows graphically

the differences between the means for the two conditions of expressive-

ness across content conditions. In addition, the low expressiveness

condition received higher ratings than the high expressiveness condition

for the ratings of Lecture B when examining the precondition of expres-

siveness for Lecture A.




58



Table 7

Student Ratine Instrument .lean Scores on Lecture B
Collapsed into Lecture A Precondition Treatment Groups


Group N .lean


High Content
High Z; ressi'.-eness 68 51. 9726
Low Ex:pressi-eness 90 57. 2438
Low Content
High Expressiv.eness 62 50.91 59
Low Expressi'.eness 58 57. 480


60

55

50-

45-


Low Content
High Content


High Low
Expressiveness Groups


Figure 4a




























S- - Lo', Content


Hieh. Content


High


Low


Content Groups


Figure 4b


Mean
Rating
Score


60 -

55-

5') -

4- -







6. There will be no differences on the dependent variables be-

tween the four treatment conditions of Lecture B, tw.o levels of content

and t'.vo levels of expressiveness. iNo content IB) by expressiveness fB)

interaction. i From Table 5, the F ratio for the content (BI oy expres-

siveness tBl is found to be significant at the 05 level (F = 6. 51 for 1

and 262 degrees of freedom). Because of the significant F ratio, the

null hypothesis is rejected, and the alternative hypothesis, that there

is a significant difference between the four treat;-ment conditions of

Lecture B, is accepted.

However, this result requires that to examine the remaining .vo

hypotheses, a simple main effects must be conducted for the above

interaction tKirk, 1968). This w.as done, and the results are reported

in Table 8.

7. There will be no differences on the dependent variables be-

tween the tw.o conditions of content for Lecture B. (No main effect for

content IBI. To examine this hypothesis, content IBI must be analyzed

at both levels of expressiveness IB). This was done. When expressive-

ness kB) *was high, the content tB) dimension did not produce a signifi-

cant F ratio at the .01 level (F = 5.21 for 1 and 262 degrees of freedom).

When expressiveness (B) was low, the content dimension did not produce

a significant F ratio at the 01 level (F = 2. 02 for 1 and 262 degrees of

freedom). Thus the null hypothesis, that there are no differences be-

twv.een the two conditions of content (B), is accepted.








Table 8

2 x 2 x 2 x 2 A analysis of '.'ariance for Student Ratin, Instrument
Total Score on Lecture B. Simple M.ain Effects :or Content (B) x
Expressiveness (B) Interaction


Source di Sum of Squares M.Iean Square F


Content at 1 S73.421S 87S. 4213 5.21
High Expressiveness
Content at 1 339.6761 339.6761 2.02
Loew Expressiveness
Expressiveness a 1 16014. 5010 1 6014. 5010 95. 07
Hiuh Content
Expressiveness at 1 5724. 6224 5724. 6224 33. 9:
Low Content
Within Groups 262 441 33. 5751 16S. 44SS
(Error)


:.p< .01, F 1,262,.01) = 6.74)







8. There '.ill be no differences on the dependent variables be-

tw.een the t.vo conditions of expressiveness for Lecture B. i No main

effect for expressiveness iB). ) To examine this hypothesis, expressive-

ness (B) must be analyzed at both levels of content IB). This .vas done.

'When content IB) .vas high, the expressiveness (B) dimension produced

a significant F ratio at the .01 level iF = 95.07 for 1 and 262 degrees

of freedom). When content (B) .vas lo'.v, the expressiveness (B) dimen-

sion produced a significant F ratio at the 01 level iF = 33. 98 for 1 and

262 degrees of freedom). Because of the above results, the null hypoth-

esis is rejected and the alternative hypothesis, that there is a difference

between the t-.o le.'els of expressiveness iB), is accepted.

The results for 7 and 8 above are sho'.vn further by an examina-

tion of Table 0 and Figures 5a and 5b. Table 9 presents the means for

the four treatment conditions examined for Lecture B. Figure .a

sho.vs graphically the lack of difference between the content (B) condi-

tions across expressiveness (B) conditions. Figure Sb shows graphi-

call,, the differences bet.veen the expressiveness lB) conditions across

the content iB) conditions.

Sumrrnarizina the results from the Lecture B analysis, there is a

significant difference bet.veen the precondition expressiveness (A)

groups on the ratings of Lecture B 'vith the lo'v expressiveness iAl

dimension being rated higher than the high expressiveness iA) dimen-

sion. Also, there is a significant difference bet.veen the expressiv.e-

ness IB) groups across content for both levels of content.




63



Table 9

Student Rating Instrument M.ean Scores on Lecture B
Collapsed into Lecture B Treatment Groups


Group N M.ean


High Content
High Expressiveness 61 65. 5313
Lo.w Expressi.eness 77 44. 1343
Low. Content
High F-xpressiveness 82 60. 5011
Lo'.v E_

Low Content
High Content


45-

40 -


High Low

Expressiveness Groups


Figure 5a


70 --

65--

60-

55-

50-


Mean
Rating
Score






















High Content






Lo-.v Content








High Lov.'
Content Groups


Figure 5b








The Lecture Choice Variable Analysis

As was stated in the methodology chapter, two types of analyses

are appropriate for the lecture choice data, in addition to descriptive

information. One analysis is a Spearman rank-order correlation be-

r.veen lecture choice and rating choice Iv.hich of the r.'o lectures was

rated higher by the student). The second analysis is to conduct a ch.-

square analysis on the lecture choice variable and the rating choice

variable, separately. The above analyses '.,.'ere done and the results

are reported below.

Descriptively', Table 10 presents a frequency distribution break-

down of the lecture choice and rating choice variables. A-s car be seen

in Table 10, 73 students 126. 26'i) selected Lecture A, 204 students

173. 38'rj0 selected Lecture B. one person did not select a lecture. For

the rating choice variable, '?4 students (33.81' S rated Lecture A higher;

170 students 161. 1'.,) rated Lecture B higher, 14 students (S. 4'r) rated

the two lectures identically.

The Spearman rank-order correlation between lecture choice and

rating choice yielded a value of 618. 'While this value is significantly

different from zero (p <. 01), it is not of practical significance. The

. 618 correlation indicates some lack of agreement between how students

rated a lecture and which lecture they chose as better, and that there

are reasons for this lack of agreement between these two variables.

An examination of the frequencies for both the lecture choice and rating

choice variable in Table 10 yields the following trend. The lecture








Table 10


Sample Population Frequency Distribution on
Lecture Choice and Rating Choice Variables


Curr Cum
Freq Freq Percent Percent
Lecture Choice Variable


No Choice I 1 0. 360 0. 360
Lecture 73 74 26. 259 26.61'
Lecture E 204 27e 73. 381 1 00, 000


Rating Choice Variable


No Choice 14 14 5.036 5.036
Lecture A 94 1 0 33. i 3 33. 84'?
Lecture B 170 273 61. 1 51 100. 000








choice variable has a greater frequency of choice of Lecture B than

does the rating choice variable. This may indicate a .veakness in the

use of a global, overall selection variable, or it ma'. be that the factors

of content and expressiv.eness for Le'rture B or a cumulative effect over

both lectures may be in-luencing the lecture choice. This problem w.vill

be addressed in the conclusion section of the next chapter.

The chi-square analysis for the lecture choice variable and for

the rating choice variable is presented in Table 11. The lecture choice

variable yielded a X- = 9. 542; at three degrees of freedom, this is sig-

nificant at the 05 level. The rating choice variable produced a

C- 3. 043, which is not significant at the 05 le.el '.'.ith three degrees

of freedom. In looking at the frequencies reported for both variables,

there is a tendency to select Lecture B more often than Lecture A, and

a tendency to select the high expressiv.eness lectures rather than the

low expressiveness lectures. There is no apparent trend for the content

dimension.

The Adjective Checklist Analysis

The analysis for the adjective checklist was conducted by exam-

ining which of the four lecture combinations (high content-high expres-

siveness, high content-low expressiveness, low content-high expressive-

ness, or low content-low expressiveness) a student chose (or rated

higher) as the better lecture in relation to whether or not that student

chose a particular adjective. The results of this frequency analysis

for the lecture choice variable are reported in Appendix E, and for the

rating choice variable in Appendix F.








Table 11

Sample Population Frequency Distribution on Lecture Choice
and Rating Choice V.ariables across Type of Lecture



Lecture Type'-

Lecture Choice HC-HE HC-LE LC-HE LC-LE


Lecture A 34 10 23 6
freq. L 12.3 3. 6 3.3 2.2
Lecture B 58 44 69 33
freq. 20.9 5.9 24. 9 1. 9


Rating Choice


Lecture A 38 17 26 13
freq. 14.4 6. 4 9. 9 4. 9
Lecture B 51 37 57 25
freq. '% 19. 3 14.0 21.6 9. 5


SHC = High Content

LC = Lo'.v Content

HE = High Expressiv.eness

LE = Lo'.v Expressi.-eness







In order to conduct the chi-square analysis for each adjective, the

significance level was set at 01 because of the number of adjectives

being considered and because of the significant chi-square for the lec-

ture choice variable anal',sis. The results of the chi-square analysis

for the adjectives -with lecture choice is presented in Table 12. The

results of the chi-square analysis for the adjectives with rating choice

is presented in Table 13. A comparison of the adjectives with signifi-

cant chi-squares for the lecture choice and the rating choice variables

is presented in Table 14. It should be noted that with three exceptions

(logical, showmanship, and persuasive are absent from the rating choice

variable list), the tv.o lists are identical.

An examination of the frequency tables in Appendices E and F for

these adjectives with significant chi-squares yields two trends in the

frequency distributions for both the lecture choice and the rating choice

variable. The first trend is that, with three exceptions knowledgeable,

orderly, and logical), for both the lecture choice and the rating choice

variable, all the adjectives tend to have higher frequencies related to

the high expressiveness lectures. The second trend is that, 'or the

exceptions cited above only knowledgeable and orderly in the case of

the rating choice variable), all tend to have frequencies associated to

the high content lectures, but more specifically to the high content-high

expressiveness lecture.







Table 12

Chi-Square values s for Adjectives on
Lecture Choice Variable


Chi-Square


Adjective


Chi -Square


Dramatic
Knowledgeable
Thorough
Coherent
Well-prepared
Energetic
Logical
Explanatory
Inspiring
Understandable
Well-paced
Expressive
Forceful
Interesting
Variety
Stimulating
Interested
Warm
Scholarly
Orderly
Stylized
Motivating
Masterful


* p<. 01, df= 3


47. 378
14. 774'
4. 521
5. 477
4. 026
94.10q*
13. 3783
7.286
16. 973*
4. 164
3. 993
48.246*
31. 729*
22.463*
1.986
27. 030*
13. 093*
14.862*
4. 404
19.220*
5.910
27. 598*
1. 192


Enthusiastic
Showmanship
Vital
Concise
Lucid
Enjoyable
Clear
Inventive
Cultured
Confident
Poised
Humorous
Sparkling
Competent
Charismatic
Conscientious
Dynamic
Well-groomed
Organized
Careful
Witty
Unique
Persuasive


Adjecti.ve


96. 341-
12. 017
4. 894
5. 630
1. 058
20. 780':
1. 900
3. 977
. 77
6. 118
4. 398
49. 172:
14. 145!
5. 134
8. 122
1. 089
26. 206'
0. 180
I. ? 35
6. 671
24. 496::
5. 627
1 1. 354'




71




Table 13

Chi-Square Values for Adjectives on


Rating Choice


Variable


Chi-Sauare


Adjective


Chi-Square


Dramatic
Knowledgeable
Thorough
Coherent
Well-prepared
Energetic
Logical
Explanatory
Ir.spi ring
Understandable
WVell-paced
Expressive
Forceful
Interesting
Va ri et,
Stimulating
Interested
VWarm
Scholarly1
Orderly,
Stylized
iotiv.ating
M ias terful


36. 77 5-
12. ?14 ;
4. 621
2. 106
0. 996
74. 896
7. 542
7. 198
12. 542:
2. 568
3. 46
44. 011
26. 046 :-
22. 284 *
4. 606
25. 41S ':
13. 727-
15. 851
8.8?9
11. 461
7. C63
26. 201
0. 482


Enthusiastic
Showmanship
Vital
Concise
Lucid
Enjoyable
Clear
Invent'ive
Cultured
Confident
Poised
Humorous
S parking
Competent
Charismatic
Conscientious
Dynamic
Well-groomed
Organized
Careful
Witty
Unique
Persuasive


Sp< .01, di = 3


Adjective


81. 640;
S. 605
3. 185
2. 365
0. 845
17. 5433
3. 360
7. 474
7. 697
4.212
4. 08 2
49. 01 5-'
14. 863-6
4. 505
3. 531
1. 638
24. 851-
0. 551
2. 023
4. 440
22. -?25::
3. 044
'. ?03




72




Table 14

Comparison of Lecturer Characteristics Significant for
Lecture Choice Variable and for Rating Choice Variable


Lecture
Choice Variable


Dramatic
Knowledgeable
Energetic
Logical
Inspiring
Expressive
Forceful
Interesting
Stimulating
Interested
Warm
Orderly
Motivating
Enthusiastic
Showmanship
Enjoyable
Humorous
Sparkling
Dynamic
Witty
Persuasive


Rating
Choice Variable


Dramatic
Knowledgeable
Energetic

Inspiring
Expressive
Forceful
Interesting
Stimulating
Interested
Warm
Orderly
Motivating
Enthusiastic

Enjoyable
Humorous
Sparkling
Dynamic
Witty




73


Summary of the Results

The summary of the results will be presented in tw.o parts. A

summary of the results of the analyses of the student rating instrument

for both Lecture A and Lecture B .v'll be presented first. Then, a sum-

mary of the lecture choice '.ariable and the adJecti-.e checklist results

will be considered.

The results of the student rating instrument analysis yielded trvo

consistent factors- l) a significant interaction bet.veen content and

expressiveness for the lecture under consideration, and (2) after con-

ducting a simple main effects analysis for each lecture, a significant

difference between high expressiveness and low expressiveness lectures

was found for both levels of content for each lecture considered. In

addition, there were two additional findings. For Lecture A, the simple

main effects analysis found a significant difference bet.veen high content

and low content lectures when expressiveness was high, but not w.'hen

expressiveness w.-as low. For the Lecture B, the precondition of ex-

pressiv.eness for Lecture A produced a significant difference betw..een

high and low expressiveness levels for both levels of content. The

above differences were the result of differences on the total score of

the student rating instrument.

For the lecture choice variable analysis, a comparison with the

rating choice variable yielded a Spearman rank-order correlation of

.618, which is significantly different from zero, but is not practically

significant. The chi-square analysis of the lecture choice variable








yielded a significant chi-square value. A chi-square analysis of the

rating choice .-ariable did not produce a significant chi-square value.

In an examination of both analyses for trends, the results yielded a ten-

dency to choose Lecture B over Lecture A and to select high expressive-

ness lectures over lo'.v expressiveness lectures.

The results of the adjective checklist '.vere as follo'.vs:

1. A chi- square analysis '.vith lecture choice yielded 21 adjectives

'.vith significant chi-square values.

2. A chi-square analysis '.vith rating choice yielded 13 adjecti-.es

'.vith significant chi-square values.

3. A comparison of the lists of adjectives found significant

showed that the lists were identical with the exception of three adjectives

not on the rating choice list.

4. In the cases of 18 of the 21 adjectives from above, an exam-

ination of the frequencies showed a trend toward higher frequencies for

the particular adjective associated with the high expressiveness lectures.

5. In the cases of three of the adjectives excepted above, an

examination of the frequencies showed a trend toward higher frequencies

for the adjectiv-es associated with high content lectures, specifically,

high content-high expressiveness lectures.














Chapter V

SUMMARY, CONCLUSIONS, AND RECONINMMENDATIONS


S umnn a r,

The purpose of this study '.vas to examn-ine the influence of the

Doctor Fox effect i.e. expressive teaching behaviors) on student

ratings of instruction when students '.ere given the opportunity to com-

pare a pair of lectures varying in levels of content and expressiveness.

Also, the study'. as to examine students' identification of lecturer char-

acteristics as these characteristics were related to expressive teaching

behaviors.

Several reasons for conducting this study were cited. From a

re.-iew of the literature on the use of student ratings and the v-.alidity of

student ratings, the following trends were enumerated as important for

this study. (1) There is an increased use of student ratings of instruc-

tion by college administrators for personnel decision-making affecting

faculty promotion, tenure, and salary increases. (2) The student ratings

of instructors are influenced by the student perception of how these

ratings will be used. (3) A number of approaches to validating student

ratings have been used; the results of these validation studies are mixed

and inconclusive. (4) The inconclusiveness of validation results has

increased the concern over the validity of the use of student ratings by








college administrators for personnel decision-making. From a review.'

of the literature on the Doctor Fox effect, the following points, relevant

to this study', were made. (1) There have been no studies of the Doctor

Fox effect providing students with an opportunity to compare inr.mediatelv

lectures varying in content and in expressiveness. 12) The literature

did not pro-.ide any' specific experimental information or insight as to

what are the high expressive behaviors of the Doctor Fox effect. (3)

Any, additional studies of the Doctor Fox effect .vill add to the literature

on the validity of student ratings, since such studies examine the influ-

ence of lecturer expressiveness on student ratings of instruction.

T,..-o hundred and seventy-eight (278P subjects .vere assigned ran-

domlv -'ithin intact classes or groups to 16 treatment conditions. The

sixteen treatment conditions consisted of all ordered pair combinations

of the four lecture segments (high content-high expressiveness, high

content-lo.o expressiveness, lov.' content-high expressiveness, ana lo".

content-low' ex:pressivenessl, w.vhich w.'ere approximately, five minutes in

duration each, used for this studio The subjects .iewed one lecture seg-

rnent (Lecture Ai of the pair, rated it, viewed the second lecture seg-

ment I Lecture B? of the pair, rated it, selected w'.'hich of the tw.o lecture

preser.'ations ''..-s the better presentation, and finally, selec-ed adjec-

tiv.es, characteristic of the instructor, h'. ich influenced their choice cf

the better lecture presentation. Thus the dependent measures for this

study were (1) total student rating instrument score (the rating instru-

ment used was the same instrument used by Ware (1974), (2) a lecture




I I



choice variable and (3) a choice,/no choice on each of the adiecti.ves on

the adjective checklist.

Separate analyses on the student rating instrument total score for

Lecture A and for Lecture B, on the lecture choice arablel, and on the

adiecti,.'es .were conducted. The 2 x 2 Icontent x expressilvenees I anal-

ysis on the student rating instrument total score for Lecture A yielded

the following results: II) a significant interaction ber.veen content aind

expressiveness, (2) a significant difference to'.'.ard higher ratings of the

high content versus the low content lectures when exoressi. eness was

high, and (31 a significant difference toward higher ratings of high

expressiv-eness versus low expressiveness lectures for both le'-els of

content. The 2 x 2 x 2 x 2 (content of Lecture A x express.-ieness of

Lecture A x content of Lecture B x expressiveness of Lecture El of the

student rating instrument total score yielded the following results: (1)

a significant interaction between content and expressiveness for Lecture

B, (2) a significant difference toward higher ratings of high expressive-

ness versus low expressiveness lectures for both levels of content for

Lecture B, and (3) the high and low expressiveness preconditions for

Lecture A produced a significant difference in the ratings of Lecture B

for both precondition content levels. The chi-square analysis of the lec-

ture choice variable yielded the following results: (1) a significant dif-

ference in the frequency of choice of Lecture B over Lecture A and (2)

a trend toward the selection of high expressiveness lectures over low

expressiveness lectures. The chi-square analysis of the adjectives on








the adjective checklist yielded the following results- 1l 21 of the 47 ad-

jectives had sIgnificant difference in choice versus no choice and 2Z) 18

adjecti.-es (dramatic, energetic, inspiring, expressive, forceful, inter-

esting, stimulating, interested, w'.arm, motivating, enthusiastic, show-

manslup, enjox.yable, humorous, sparklihn, dynamic, witty, ana persua-

s:-'.el showed a strong relationship with the iugh expressiveness lectures.


Conclusions

Overall, the results of this study seem to point in one direction,

confirmation of the primary premise of the Doctor Fox effect. In other

.words, the results sho.vine differences on the student ratings of the lec-

tures and on the lecture choice variable due to the expressi-eness dimen-

sion for both levels of content confirms the Doctor Fox effect research

previously mentioned. Further, it appears that the expressiv.enicss di-

mension does have a primary tniluence on the student ratings of instruc-

tion.

The results of this study seem to indicate that previous experience,

or the opportunity to compare lecturing styles, also may influence eval-

uation of the teaching process by students. A careful examination must

be made of the result of the significant difference between the ratings of

high and low. expressiveness lectures produced by the expressiv'.eness

precondition of toe iirst Lect-ire. This result showed that low.' express-

iv.eness of the first lecture i.-ifluenced higher student rating of the sec-

ond lecture and that high expressiveness of the first lecture influencedd

Lo.\er studen: ratings rf the second lecture. This result ,.vns true for




7-


both precondition content levels. Since there w.-ere no significant inter-

actions between the preconditions of the first lecture and the treatment

conditions of the second lecture, it can be concluded that the expressiv.e-

ness precondition of the first lecture presentation has an independent

influence on the student ratings of the second lecture presentation. This

precondition expressiveness influence is independent of the influence

that the content and expressiveness levels of the second lecture presen-

tation have on the student ratings of that second lecture presentation.

The results of Ware and Williams (1975a) and Williams and Ware

(1977) indicated a secondary premise, or corollary, to the Doctor Fox

effect. The conclut.ion, drawn from their results, '.-as that under a con-

dition of high expressiveness, lecture content w..ould have no effect on

student ratings: under low expressiveness conditions, lecture content

would have an effect on student ratings. The results of this study dc not

support entirely this conclusion. In fact, for the analysis of the first

lecture, the exact opposite of this conclusion was found. For the first

lecture, the results showed no difference in student ratings for content

at low expressiveness; however, at high expressiveness, there was a

significant difference in the rating of the high content versus the low con-

tent lectures. However, this significant difference disappeared for the

analysis of the second lecture.

The explanation of this reversal of the content dimension results

may lie in the shortness of the lecture presentation. It may be that five

minutes is not enough time to ascertain fully the content present in the







lecture segment. Rather, the rating difference may be an artifact of the

content by expressi-eness interaction. This explanation is supported bv

the fact that the result disappears T'.hen the second lecture results are

examined. Thus, this shortness of time may. focus the attention of the

students on the more easily recognizable quality in lecturing, style; that

of expressiveness.

This focusing on the express.ieness of the lecture by stud.ncs is

supported additionally by the lecture choice and adjecti-.e checklist re-

sults. Students had a tendency to select high expressiveness over low.

expressiveness lectures. Also, 15 out of the 21 adjectives -.with signifi-

cant frequency distributions shoved a tendency to be related to high

expressiveness lectures. These same 1; : adjectives -.ere descript:'.e of

the expressive lecturer. In fact, 15 of the 13 adjectcies came directly

from Ware's (1974) definition of expressiveness. These factors seem .o

be, at least, a confirmation that students can identify hi1h expressive

beha-.-ors '.,.hen selecting adjectives -which influence their choice of a

best or better lect-re presentation. The conclusion from the facts

above is that students' selection of -.which lecture -.was better is based

primarily on high expressive behaviors.

As a means of sun-rrar.--.in the conclusions of this study, an ex.am-

ination of each of the or:-inal research questions follo.'.'s

1. Do levels of lecturer effectiveness and levels of lecture content

interact with each other in affecting student evaluation of lecturer effec-

tiveness? Yes, they do. This is a confirmation of the original Doctor

Fox effect experiment.







2. After viewing a second lecture presentation varying in content

and/or expressiveness from the first lecture presentation, do levels of

content and levels of expressiveness of this second lecture interact wi'th

levels of content and levels of expressiveness of the first lecture pre-

sentation in affecting the student evaluation of lecturer effectiveness?

No, not directly. The results showed that the expressiveness level

from the first lecture (a precondition) has an independent effect on :he

student ratings of the second lecture.

3. Are there differences in student evaluations of lecturer effec-

tiveness for lectures which are high in expressiveness as opposed to

those which are lo,. in expressiveness Yes, for both lecture anal., jes

and for the lecture choice and adjective checklist analyses.

4. Does level of content in a lecture presentation affect student

evaluation of lecturer effectiveness? The answ'.er to this question is

amb-ivalent. Differences were found for the first lecture analysis oni,'

if expressiveness was high. No differences were found for the analysis

of the second lecture. The adjective checklist analysis yielded three

adjectives Ikno,.wledgeable, orderly, and logical) .'-hich ..'ere partially

related to hich content lectures.

5. Are there differences in student choices of adjectives, char-

acteristic of the instructor, to describe what they feel is a better lec-

ture presentation? Yes. These differences in choices showed a trend

toward a relationship with high expressiveness lectures.








Recommendations

Two types of recommendations and comments are 'warranted as a

result of this study. The first are specific research recommendations

for further study. The second type are implications, recommendations,

and comments related to issues raised in this stud,,.

The following are the research recommendations for further study.

1. A different analysis of this data using changes in student rating

instrument scores from Lecture A to Lecture B to look at order effects

of lecture presentation on expressiveness and content and to examine

the independent, expressi-.eness precondition effect on student ratings.

2. A follow-up study, varying the time of lecture presentation to

determine if time exposure is a factor influencing student ratings.

3. The development of a systematic observation instr'.,irent frcm

the adjectives identified as significant and shown to be related to hiah

expressive beha-,iors and then testing the reliability and '.ali.Ji;- of such

an instrument. The development of such an instrument w..ould enable

researchers to look at the Doctor Fox effect in a real classroom setting

over an entire quarter. Further, it v.would aid in faculty developn-ent

program by, providing a tool to assess faculty lecturing st.-!e and then

by providing feedback to faculty to atd them in the improvement of :hc:r

lecturing style.)

The implications of this study and the attendant recommendations

and comments relate to r,.vo issues raised previously. Those issues

are the increased use of student ratings of instruction by college







administrators for personnel decision-making and the inconclusi-.e

results of validity research on student ratings of instruction. Since the

use of student ratings hinges on the validit-y question, the issue of the

validity of student ratings is considered first.

The results of this study imply that the primary Influence on

student ratings of instruction is expressi.-e lecturing style and behaviors.

In addition, this study implies that students, '.-hen compar-ng lecture

presentations varying in content and expressiveness, are not able to

Identify differences in content level, or at the very least, become un-

aware of content le"el in a lecture presentation. These implications

raise serious questions about the validity of student ratings. Supposedly

student ratings pro.-ide an assessment of the student's perception of a

faculty merr.ber's ability, to teach content. Yet, the conclusions of this

study are that student perception is influenced prin-arily. by, not h'. at

the professor presents, but how." he presents it. Thus, student ratings

are providing an assessment only of student perception of teaching stylc,

and this is only rar: of 1hie picture.

The validity of student rat!nps can not be assessed completely .f

student reporting cf faculty eaching is biased because of expressive

teaching style. Student ratings do not provide all of the information

necessary for administrators to assess faculty teaching performance,

let alone for administrators to make personnel decisions about faculty

promotion, tenure, and salary increases.

The increased use of student ratings is questionable in the light of

the above implications from the results of this study. If it is in fact the







case that faculty performance is being judged solely on the basis of stu-

dent ratings, then faculty members with a ncnexpressi'.-e teaching style

are being punished in cases where student ratings are used primaril'l

for personnel decisions. Further, students can be fooled or seduced

into believing a faculty member is a good teacher solely, by his exhibi-

tion of a highly expressi-.-e teaching style, regardless of the amount of

information that the faculty,, member may. be imparting to students. Thus,

it may be possible to reward a faculty, member, who teaches his students

little or no information, because he has a highl,, expressive teaching

style and high student ratings as a result of that teaching style.

The above implications lead to the recommendation of a restruc-

turing of faculty evaluation and reward procedures. Cook and Neville

.1,'7.': ha.-.e suge e sted that some form of a multiple data approach tc fac-

ulty evaluation is necessary. The results of this study, support thar con-

tention. To begin .vith, since student ratings of instruction can be influ-

enced b. expressive lecture style, student ratings can pro-.ide an accu-

rate assessment of the expressiveness of teaching style. This is not to

negate the value of student ratings. Rather, it says that student ratings

of lecture style are important. Ware and Williams (1975a) and Williams

and Ware (1977) have shown that high expressive teaching style and be-

haviors have a positive influence on student achievement. This alone is

reason enough to include student ratings as part of a multiple data ap-

proach to faculty evaluation.

However, student ratings do not assess accurately actual impact

of teaching style on student learning growth. While teaching style does








have an effect on learning growth, the measurement of learning growth

by student ratings s not appropriate because of the above mentioned

influence of expressiveness on student perceptions of faculty teaching

ability.-. Thus, the direct measurement of student learning growth

should be considered as a second part of a multiple data approach to

faculty, evaluation. While final course examinations, final course

grades, and achievement tests are all measures of student achievement,

the', fal to control for any, of the student's pre'.'ous experiences with

course material or related course material. It is more appropriate, in

measuring teacher impact on student learning growth, to use a pre -

post achievement gain score. This gain score controls for the precon-

ditions of any previous student learning, and thus is a better measure

of the impact of faculty teaching ability on student learning growth.

So far, two parts of a multiple data approach have been identified;

they are student ratings of instruction and achievement gain scores. A

final piece of information should be considered in order to complete the

picture of faculty teaching performance evaluation. This information

should illuminate the student/teacher interaction in the classroom. One

of the better ways to assess this interaction is through systematic obser-

vation techniques. This study has identified a number of lecturer char-

acteri. tics related to expressive teaching style. These characteristics

can be converted into a systematic observation instrument to obtain in-

formation on student/teacher interaction in the classroom. The syste-

matic observation instrument, when used by trained observers, would








provide an unbiased view of the student/teacher int-traction, as '.vell as

an objective assessment of teaching style. With the information of sys-

tematic observation, student ratings, and achicve:T.ent gains, a college

administrator -.%ould have the appropriate information about faculty

teaching performance upon which to base decisions affecting faculty pro-

motion, tenure, and salary increases. Any, program of faculty evalua-

tion that en-.ploys the sole use of student ratings very likely is not pro-

viding adequate information on faculty teaching performance and may

lead to erroneous personnel decisions.

One last recommendation is appropriate for the results of this

study. The adjectives identified in this study provide a clearer picture

of the high expressive teaching style. This style has been shown to

influence positively student ratings and student achievement. Thus,

faculty development programs could develop specific training prog ramns

in teaching style for college professors based on the adjectives identi-

fied in this study.

These adjectives are lecturer characteristics, and if converted

into the systematic observation instrument mentioned above, would pro-

vide an excellent source of feedback to aid college professors to improve

their teaching style. The behaviors associated w'.ith these adjectives

might be taught to professors, and thus have the potential for improving

their classroom presentations. This is important since these highly

expressive teaching behaviors identified by the adjectives have an im-

pact on student learning grow.th. Since part of the goal of teaching is to








have an impact on student learning, these behaviors ire likely to aid

professors in their attempts to foster student learning.

In summary, the tv.'o practical recommendations grov.'ing out of

the results of this study are the follo'.ving:

1. A restructuring of faculty evaluation and re.vard procedures

into a multiple data approach using systematic observation procedures,

student ratings of instruction, and student achievement gain scores as

the basis for e',aluation of faculty teaching performance.

2. An addition to faculty development programs of a program to

teach the high expressive teaching behaviors associated ..ith the Doctor

Fox effect to college professors, so that the faculty, of colleges and

universities can improve their teaching style and have a greater impact

on student learning.






























APPENDIX A

Administration Directions













ADMINISTRATION DIRECTIONS

Good morning. NIy name is Howard Ramagli. This morning I would

like to ask you to help me. In an effort to develop new ways of provid-

ing learning experiences for students, we are f-eld testing some v.deo-

taped lectures as possible supplementary material for introductory

classes in the area of psychology. The specific lectures we are testing

are on the biochemistry of memory. What we are asking ,ou to do to-

day is to help us evaluate the quality of lecturing in these videotaped

lectures. You will be shown t..'o segments of tapings of a lecture on

the biochemistry of memory. After you view each lecture segment,

you vill be asked to evaluate the quality of lecturing in that particular

segment. Let me repeat this. I will show you a segment of Lecture A

and then I will ask you to evaluate the quality of lecturing in that seg-

ment. I will then show. you a segment of Lecture B and :ou will be

asked to evaluate the quality of lecturing in that particular segment.

Are there an., questions so far? Please open to page 1 of your booklet.

The evaluations we are collecting from you will be kept confidential,

however, we do need some descriptive information on you as a group.

This information will also be kept confidential. Would you please fill

in the information requested on page 1. When you have finished, please

do not turn the page. PJerse look up at me to indicate you have finished.



89








PAGE 2

iFilling out of descriptive information)

Is everyone finished?




Now,. that all of .ou are finished, we .;'ill begin. We will start with a seg-

ment from Lecture A. ITURN ON VIDEOTAPE) This is Lecture A.




I View,'ing of Lecture Ai

Please turn to page 2 in your booklet. Th-s is the evaluation form we

would like for you to fill out for Lecture A. (P.EAD INSTPUCTIONS)

(Discuss items 1 and 8, explaining their five choices to circle) Please

respond to all of these items on pages 2 and 3. Please begin. When

you have finished, please do not turn the page. Please look up at me to

indicate you have finished.




(Rating of Lecture A)

Is everyone finished?




Now that you all are finished, we will continue. We '.,ill now,. see a seg-

ment from Lecture B. (TURN ON VIDEOTAPE) This is Lecture B.




(Viewing of Lecture B)

Please turn to page 4 in your booklet. This is the evaluation form for

Lecture B. It is the same form you filled out for Lecture A. Please







PAGE 3

complete this form. Please respond to all of these items. When you

have finished, please do not turn the page. Please look up at me to

indicate you have finished.




(Rating of Lecture BI

Is everyone finished?



Nov. that ,ou have finished, please turn to page 6. The instructions ask

you to select which of the r.to lecture segments you saw was the better

lecture presentation. Please place a check in the box below either Lec-

ture A or Lecture B. Please do not check them both. When you have

finished, please look up at me.




(Completing Choice Variable Page)

Please turn to page 7. On this page you will find a list of adjectives.

Please read the instructions carefully. They ask you to select those

adjectives, characteristic of the instructor, which were influential in

your choice of which lecture presentation was better. Please place a

check on the line to the left of each adjective, characteristic of the lec-

turer, which was influential in your choice of the lecture segment you

chose. You may select as many or as few adjectives as you choose.

When you have finished, please look up at me.




(Completing of Adjective Checklist)






PAGE 4

Now, that you have finished, please pass your booklets for'.ard. I '.vould

like to thank each of you for your help. VWe will be conducting this eval-

uation with a number of classes in the Ccllege of Education. In an ef-

fort to get the honest opinion of all of those individuals wve .vill be asking

to evaluate the .-deotapes, -ve ask you please, do not discuss what you

have done or seen .vith anyone outside of this class. This is most irr.-

portant if wve are to get an honest evaluation of the quality of these video-

tapes. I '.vill be glad to come back, if you are interested, toward the

end of the quarter and provide you with feedback as to the results of

these evaluations. Thank you for helping me.






























APPENDIX B

High Content Lecture Script




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs