University of Southern California Dissertations and Theses

The effects of teacher-student cognitive style similarity on performance evaluation, performance prediction, intelligence estimation and general preference

(USC Thesis Other)

The effects of teacher-student cognitive style similarity on performance evaluation, performance prediction, intelligence estimation and general preference

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE EFFECTS OF TEACHER-STUDENT COGNITIVE
STYLE SIMILARITY ON PERFORMANCE EVALUATION,
PERFORMANCE PREDICTION, INTELLIGENCE
ESTIMATION AND GENERAL PREFERENCE
by
William Arden Spindell
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Education)
August 1975
UMI Number: DP24172
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI DP24172
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346
Copyright © by
WILLIAM ARDEN SPINDELL
1975
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 9 0 0 0 7
This dissertation, w ritte n by
......... WiUi&m..A£d£xi.£jund&U..........
under the direction of hij&..... Dissertation C om
mittee, and approved by a ll its members, has
been presented to and accepted by The Graduate
School, in p a rtia l fu lfillm e n t of requirements of
the degree o f
D O C T O R O F P H I L O S O P H Y
Dean
DISSERTATION COMMITTEE
N fuhairm an
TABLE OF CONTENTS
LIST OF ILLUSTRATIONS..................... iv
LIST OF TABLES ...................... v
Chapter Page
I. THE PROBLEM. . 1
Introduction
Background of the Problem
Purpose of the Study
Statement of the Problem
Assumptions and Limitations
II. REVIEW OF THE LITERATURE................... 12
Early Expectation Studies
Early Interaction Studies
Interaction of Individual
Teacher-Student Factors
III. METHOD....................................... 401
Overview
Subjects
Reliability and Validity of the
Test Instruments
Pilot Study
Materials
Methodological Limitations
Procedure
Tabulation of GEFT Data
Tabulation of Ranking Response Data
Analysis of Ranking Response Data
IV. RESULTS....................................... 65
V. DISCUSSION 73
Chapter Page
VI.. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS . . 81
Summary I
Conclusions
Recommendations
APPENDIXES
A. GEFT Score Distribution...................... 88
B. Sample Page of GEF Test. .............. 90
C. Criteria for Selection of Children......... 92
D. Videotape Objectives and Instructions
to the Teacher............ 94
E. Concept and Videotape Word List............ 97
F. Effectiveness of Videotape Stimulus
Characteristics............. 100
/
G. Videotape Response Record............... 103
H. Contingency Tables for Ranking Responses . . 105
REFERENCES................. 112
LIST OF ILLUSTRATIONS
Figure Page
1. Experimental Matrix.......................... 64
iv
LIST OF TABLES
Table Page1
1. Expectancy Group Differences in Child
Behavior and Teacher-Child Interaction . . 16
2. Validity Coefficients.................... 44
3. Videotaped Direct Response Tabulation. . . . 54
4. Ranking Responses as Related to Research
Hypotheses ........................ 66
5. Chi-Square Comparisons Between Field-
Dependent and Field-Independent Subjects . 67
6. Cell Means for All Ranking Response
Categories............... 70
7. Frequency Comparison of Response Magnitudes
for All Six Response Categories Combined . 72
v
CHAPTER I
THE PROBLEM
Introduction
A difficult problem for any science occurs when
researchers of good repute develop contradictory theories
based upon a seemingly conscientious selection of research
results and other experimental evidence. This is not how
ever exclusively restricted to the behavioral sciences but
'even prevails in such quasi-precise fields as astrophysics,
where speculations are still polarized on the origins of the
universe and the evolution of our own solar system.
In the late 1960s and early 1970s educational psy
chology was confronted with some provocative and often con
tradictory evidence regarding teacher expectation and its
impact on educational processes and, more importantly, edu
cational results. This evidence, including that of Rosen
thal and Jacobson (1968), cited in the literature review
that follows dealt with a special process of expectation
dynamics, the self-fulfilling prophecy, and its implications
toward teacher-student relationships. This theory, expanded
later on, posited that expectations once generated tend to
fulfill themselves through a complex and interactive pro'ces sj
which tends to progressively narrow behavior latitudes. For!
.example one theory suggested that when a teacher initially i
forms an opinion about a student and is sufficiently rigid ,
in maintaining this opinion, he will treat that student
accordingly (Brophy & Good, 1974). They went on to say that
the, student, consciously or unconsciously aware of this
treatment, tends to behave in a way which he perceives to be
appropriate because he is successively reinforced in some
manner by the teacher for that appropriate behavior and in
the same manner ignored or even punished for contrary behav
ior. Several iterations of this process in the classroom j
serve to strengthen and shape emerging behavioral bonds
until the resulting behavior conforms very closely with the
expected behavior. In the literature of social and educa
tional psychology prior to 1968, expectation had been ex
plained as a mind state or mental set which served to pre
dict the behavior of others given very few substantive
stimuli.
The simplistic behavioral analysis of classroom pro
cesses drew empirical support from some of the early expec-
, : 1
3 I
i
tations studies and was strongly refuted by many of the
later ones (Thorndike, 1968? Claiborn, 1969; Fleming &
Anttonen, 1971). Fortunately, the ensuing controversy
spurred much quality research, and in the truest sense of
the word, this study is a beneficiary of the research
generated.
While the importance of expectation in teacher-
student relationships had perhaps always been suspected by
behavioral scientists, its impact on student performance has
only recently been seriously considered or systematically
measured.
I
Background of the Problem
This research was beneficial in identifying, occa
sionally by omission, some of the most important elements of
expectation dynamics as well as some contributing reasons
■for its selective effects. For example, in numerous studies
expectation was often shown to be communicated through some
of the most subtle classroom processes: calling on selected
students for particular answers, seating arrangements,
prompting of responses, selectively waiting for answers from
some students but not for others, and some other more global
treatment indices. Regarding selective factors, the tradi
tional ones of race, ethnicity, and sex were of course
identified in many studies, but other preference "systems"
were also identified. One of these Witkin, Dyk, Paterson, :
Goodenough, and Karp (1962) identified in a different but ;
i
related context as cognitive style similarity or dissimi
larity between the student and the teacher.
Witkin et al. (1962) defined cognitive style or
"personal style" as a person*s characteristic approach to
the perception and understanding of the world around him.
^They further elaborated cognitive style as a characteristic
i
approach to the solution of problems as confronted in almost
any aspect of his life. In short, the processes of psycho
logical differentiation during human development, among many
other genetic and environmental factors, tended to produce
people who differed along an interestingly relevant and mea
surable dimension of the cognitive style system-— that of
field dependence and field independence or a hypothetical
continuum which encompassed both. They suggested that field-
dependent people are typically more context-bound in their
perception, their approach to problems, and their general
behavior. Field-independent people, on the other hand, are
much less context-bound and approach problems with inter
nally defined solutions. They are more analytic and are
less cognizant of social cues. Again, this is not a simple
either-or dichotomy but is a distribution approximately nor-;
mal in shape (Witkin, Oltman, Raskin, & Karp, 1971).
Witkin et al. (1971) went on to conclude that teach-
!
ers are also distributed along this continuum but more often1
toward the field-dependent end, as a result perhaps of the
selection process which drew them toward this profession.
i
Students are distributed more or less normally, not yet hav
ing been subject to these choices and "natural selection"
pressures tending to support the contentions of Witkin et ali
(1971). Castaneda and Gray (1974) made the added claim that
these processes--bicognitive processes-— were of special
i
.importance in multicultural education since:
(a) public education tended to favor a teaching style
that is more appropriate to one cognitive style than
the other; (b) many children from ethnic minorities
come from homes which tend to teach in a style more
appropriate to the Other cognitive mode; and (c) equal
educational opportunity is denied when educational
policy and practice favor one cognitive style over the
; others. (p. 204) .
Witkin (1973) and Witkin and Moore (1974) presented
impressive empirical support for their contention that teach-
i
ers prefer students closer to their own cognitive styles than
those who are considerably divergent from it. They docu
mented this contention with evidence across a wide diversity
of outcome studies including, for example, reduced drop-out
rates,among cognitive style matched (vs. mismatched)
nursing students and their teachers, learning gains which j
were significantly greater among style matched teachers and j
i
tutorial students, and other indices of strong preferences j
!
among teachers and students of similar cognitive styles. !
In the context of this problem statement, the
results of a variety of teacher-student relationship stud
ies, expectation studies, and cognitive-style match vs. mis
match studies may be categorized in the following manner:
1. Classroom process studies dealing with teacher-
student relationships revealed that teachers
interact significantly, differently, and pre
dictably with different students. Further, the
geneses of these differences are found in the
personal characteristics of teachers on one
hand and students on the other.
2. Studies designed to assess performance differ
ences among groups of students for whom expec
tation messages were communicated, and in which j
the dependent variable was some outcome measure
(e.g., IQ gain, learning differential, or grade
improvement) as often as not showed no signifi
cant differences when compared with suitable
controls.
3. The credibility of the source communicating the
message and the content of the message were
important factors, particularly if evidence con
trary to the expectation message existed.
4. The strength of the treatment, that is, simple
suggestion of differences vs. strong factual
difference (e.g., cumulative card records) was
an important variable.
5. Significant differences often appeared in pro
cess variables (i.e., teacher-student verbal
7 !
i
contacts) while not in product or outcome
variables. !
t i
6. Compatibility between the cognitive style of the
teacher and that of the student provided for
measurably better educational outcomes in a
' diversity of settings from elementary schools
through professional schools and even in mili-; 1
tary cadre-student relationships. ;
i
From this evidence, it would appear that a poten- 1
I
tially valuable study would be one Which could evaluate I
j
expectation dynamics and cognitive style matching phenomena
|jointly. Such a study could discover an important set of
expectation mediational factors which in turn could help to
I
'explain the relatively contradictory results of many pre- ■
i . 1 '
viously attempted studies. Specifically, such a study could
determine the influence of style matching on expectation and
teacher-student preference systems as well. Its importance
to educational psychology is apparent: given the recent
evidence of the favorable results of teacher-student cogni
tive style match and clear-cut evidence that performance
expectation, intelligence estimation, preference, and per
formance evaluation may also show a similar bias, positive j
steps may be taken either to compensate for these biases or ■
to make teachers aware of their existence.
Purpose the Study
Accordingly, the present study attempted to take
■these expectation studies and cognitive style studies one
■more important and logical step: to combine the findings of
each group of studies in an experiment designed to deter-
.mine whether cognitive style similarity between teachers and
students, influenced expectation. Expectation was measured
by teachers' estimates of student IQ, rating of on-task per
formance, potential in specific subject areas, and the gen-
t
eral preference for the various cognitive styles depicted
among stimulus students.
By creating, validating, and presenting videotapes
of two microteaching sequences depicting children of the
field-dependent and field-independent cognitive styles, the
study attempted to determine whether a cognitive style bias
existed among subjects classified along the same cognitive
!style dimension. These videotapes, taught by a field-
dependent teacher, provided an opportunity for a boy and
girl of each cognitive style classification to demonstrate
their unique style characteristics in subject areas in which
they theoretically tended to do best: physical sciences for
the field independent and social science for the field
dependent (Witkin & Moore, 1974).
Statement of the Problem
What relationship exists between student teacher
perceptions of children who are field dependent and field
independent when the student teachers are themselves cate- j
gorized along these same dimensions? From the research data
cited, it would appear that a bias would be operative in
which field-dependent student teachers would perceive field-
dependent children as smarter, more likeable, more capable,
'and more likely to excel in their studies.
. \
Assumptions and Limitations
I v Conduct of such an experiment has some limitations
in the generalizability of results to a classroom environ- (
ment. An important experimental consideration observed by
Beez (1968) was to parallel the task closely with that of
the ordinary classroom activities. Beez used behavioral
"report" data. This experiment used simple on-task perfor
mance rankings, intelligence,ranking, and predicted perfor- ,
mance ranking with strong face-similarity to normal class
room ranking activities.
A further limitation on the generalizability was the
use of teacher education students who may lack the experi
ence required to make rapid evaluations based upon limited
exposures to a stimulus situation. Experimental evidence
10
indicates reasonably similar results on other ranking/rating,
tasks between in-service teachers and student teachers, how
ever. In the same sense, experience may have dictated to
in-service teachers that observations of children in certain
situations might not be generalizable to others. There is j
again experimental evidence that other factors such as j
teacher reactive style— the manner in which he or she reacts I
to the stimulus equalities of the students— may be a more
jimportant determiner of whether or not expectations, are
formed than experience (Feshbach, 1968).
Perhaps the singular technical limitation of the
i
study involved the use of videotaped stimulus students as
opposed to observing students in a classroom with subjects
.making on-line judgments. The primary advantage of video
tape in this regard is its reusability. Greater numbers of ,
subjects may be exposed to the stimuli in a given time
l
period, resulting in a moderate increase in statistical
power.
A statistical limitation relates to ranking as
opposed to rating the stimulus children. The primary con
sideration was the limited exposure opportunity afforded by
the videotapes. Despite the statistical restrictions, upon
considering the level of subject sophistication along with
11
the brief time factor, it was judged that rank ordering the
stimulus children had several advantages. In the context ofi
the experiment, the chief advantage was to gain as simple 1
i
I
an estimate as practical of a bias toward stimulus children j
of like cognitive styles.
CHAPTER II
REVIEW OF THE LITERATURE ;
This chapter is organized into three parts. The
first part, early expectation studies, discusses the re-
i
search findings and theoretical framework within which the 1
expectation phenomena was thought to occur in educational
settings. The expression "early" roughly denotes the period
from about 1968 to 1972 but, more importantly, refers to thei
'attempts to replicate the study by Rosenthal and Jacobson
(1968), using the bogus expectation paradigm (Thorndike,
1968). !
I
The second section deals with precise studies of
classroom interaction unobstrusively conducted in the class-
'room environment. These studies provided evidence of numer
ous teacher and student interactions including those of
racial, ethnic, and personality factors which could, in some
respects, predict the quality and frequency of teacher-
student interaction.
12
The final section treats the more recent work by
Witkin and Moore (1974) on the more specific factor of cog
nitive style and how it has been shown to influence the
quality of classroom and other educational interactions.
The section integrates the previous two and presents the
rationale underlying the research hypotheses.
Early Expectation Studies
, Rosenthal and Jacobson (196 8) published the results
^of their Bay Area elementary school study in a book called
I Pygmalion in the Classroom and, in so doing, started a vig-
f
Jorous controversy which persists today. In its simplest
i
statement, the theory posits that teacher expectation for
student achievement would function as self-fulfilling proph
esies. Accordingly, teachers given information about stu
dents in the form of "expectation messages" such as "late
ibloomer," "bright," and "dull" may, if they accept these
labels, tend to set up conditions, however inadvertently,
that will cause these expectations to come true. Expecta
tions are communicated by other means as well: racial ste
reotypes (Rist, 1970; Rubovits & Maehr, 1971) and other
salience characteristics including sex (Jackson & Laherderne
1967), classroom behavior (Kelley, 1950), and ethnicity
(Williams, Whitehead, & Miller, 1972). While not all these
14 |
expectations are, as Merton (1948) called them, self- j
fulfilling prophesies, they often precondition attitudes j
which may in turn make expectation messages seem more plau
sible.
The Rosenthal and Jacobson (1968) study used a methodj
' I
i ,
of attributing especially favorable performance expectations!
to randomly selected students under the guise of ability 1
testing. Teachers, so informed of the select status of
these students, rated them as high potential and as having
achieved highly during the ensuing semesters and those stu
dents labeled "late bloomers" made impressive gains in IQ
and other performance indicators.
The conclusions by Rosenthal and Jacobson (196 8)
that the acceptance of an expectation communication from an ;
authoritative source could result in differential classroom
treatment of the experimental students and that this treat- ■
i
ment could positively and significantly influence both
school achievement and IQ became a topic of controversy.
Attempts to replicate the original findings failed repeat
edly (Claiborn, 1969). Criticisms of the conclusions, of i
the data analysis technique, and the experimental methodol
ogy (Thorndike, 1968) raised other serious questions.
Many studies were conducted involving considerably
weaker treatments and methods with other serious methodolog
ical problems (Brophy & Good, 1974). These methodological
problems, for example, involved applying labels after the
teacher had had considerable contact with the student
(Claiborn, 196 9; Fielder, Cohen, & Feeney, 1971) and then
attempting to interpret the resulting non-significant-
differences (NSDs) as not supporting the claims of Rosenthal
and Jacobson (1968).
Table 1 summarizes some of the variables found in
group differences between students ranked as highs and lows
in an earlier Good and Brophy study (1970), high reflecting
a high performance ability estimate by the teacher and low
indicating the opposite.
In general, naturalistic studies made in classroom
settings tended to support the assertion that the teacher
i
responded qualitatively and quantitatively differently to
different students (Kranz & Tyo, 1973; Mackler, 1969; Rist,
1970). These studies used various methods of process mea
surement with the common aim of sampling classroom interac
tion (Flanders, 1970.) at periodic intervals and also of de
termining through measurement which students received what
type of contact with the teacher (Good & Brophy, 1970, 1972;
Table 1
Expectancy Group Differences in Child Behavior and Teacher-Child Interaction
Measures Lows Highs
Number of times called on to answer an open question 1.71 1.96
Number of times called on to answer a direct question 1.83 2.50
Number of times called on by teacher during reading groups 4.79 3.29*
Number of times child called out answer during reading group 2.96 3.54
Procedural contacts initiated by child 3.17 5.13**
Work-related contacts initiated by child 1.79 7.38***
Procedural contacts initiated by teacher 2.58 2.04
Work-related contacts initiated by teacher 6.00 3.79
Number of behavioral criticisms from teacher 4.92 2.04***
Total teacher-initiated response opportunities 10.96 10.29
Total child-initiated response opportunities 7.92 16.04***
Total dyadic contacts with teacher 33.67 35.17
Number of times child raises hand to seek response opportunity 8.88 16.67***
Number of times called on/number of times raises hand 0.20 0.12**
Total correct answers 6.67 8.92*
Total incomplete, incorrect and "don't know" answers 4.63 2.38***
Average number of reading problems per reading turn 4.67 2.23***
Percent of total contacts involving praise from the teacher 3.88 11.00***
Percent of total contacts involving criticism from the teacher 24.33 10.75***
*p < .10 **p < .05 ***p < .001
Source: T. Good and J. Brophy, Teacher-Student Relationships3 Causes and Consequences
(New York: Holt, Rinehart & Winston, 1974), p. 95.
17 i
i
Silberman, 1971). While significant differences appeared
between high and low performance ability students on 11 out
of 18 interaction measures (including procedural contacts, j
j
behavioral criticisms, and number of times called on/number 1
of times raises hand), naturalistic studies did not gen- j
f
I
erally attempt to document performance differences under- j
lying factors (Good & Brophy, 1970). t
Laboratory studies, however, tended more toward
i 1
(product measures of gain over time, for example, in IQ based
on the communication of some expectation message or label to
.teachers or to naive subjects and later the measurement of
I
some dependent performance variable. In general, the lab
oratory studies tended not to support the assertion that
increments in academic performance or IQ could arise merely j
from communicated labels or, as some called it, "bogus ex
pectation" (Fleming & Anttonen, 19 71) .
!
A notable exception was a study made by Beez (196 8)
which used the labeling method and then measured the number
of words taught to the differentially labeled students in a ■
microteaching experiment. The study showed significant dif-:
ferences in the number of words taught to the "brighter"
students of "good" home backgrounds.
Each group of studies had characteristic weaknesses ;
i
i
and strengths. The naturalistic studies pointed toward the j
potent verbal and behavioral messages communicated by seat
ing (Rist, 1970; Seaver, 1,971) , tracking, and, quite impor-
i
tantly, sex biases, while the experimental studies suffered
I
from the superficiality noted plus severe drawbacks associ- (
ated with absence of control groups, extremely brief expo- '
sure of subjects to treatments, and labels applied after
impressions had been firmly formed, which often appeared
ludicrous as a result. Further, as ambient knowledge on
expectation increased among student and teacher populations,:
studies using the "bogus expectation" model became increas
ingly difficult to conduct (Brophy & Good, 1974; Panda &
Guskin, 1970). Some epxeriments which did not result in
product gains nevertheless influenced process. Thus, while
Brown (1970) failed to get learning differentials among
i
tutored students, teaching behavior among the tutors dif
fered considerably. Similar results were reported by Kester
and Letchworth (1972).
For those studies dealing with outcome or product
measurements— IQ, GPA, SAT-score, or other achievement test
scores— the results were very ambiguous. Brophy and Good
(19 74) stated that "most studies using product measures
found no expectancy advantage . . . but the replication
studies . . . tended to involve weaker treatments" (p. 73). |
According to Good and Brophy (1972) the major weaknesses in j
several experimental paradigms were (a) whether or not the |
teachers adopted the expectation; (b) whether or not they
had evidences to the contrary; and (c) whether or not the
expectation, although adopted, had any effect on perfor
mance outcome.
i
i |
Early Interaction Studies
In general, the limitation of the labeling and ex-
jpectation communication studies conducted prior to 1970 was
that few results could be generalized from the experimental '
situation to the general school case. Reflecting this view
point, a rather weak statement by Brophy and Good (1974),
typical of the state of the research at that time, was that
under certain circumstances, expectation effects did accrue.
Another major limitation was that expectation was attributed
mystical qualities because it did work sometimes and some
times did not. It began to become clear to many researchers
that the intervening variables required more intensive study
than the magnitude of the dependent variables.
As noted earlier, the strongest challenge by Clai-
born (1969) and Thorndike (1968) and others expressed as a
question was: why could not the Rosenthal and Jacobson
i
(1968) effects be replicated? Rosenthal and Rubin (i971) j
responded that failure to replicate often reflected failure j
to set up a study which could provide the minimal conditions'
needed to generate the effect. Clearly, an impasse existed
here, but the definitive work of Good and Brophy (1970) and
others helped to provide the next logical step in research.
Rather than addressing group means for treatment and
control groups, which could at best identify differential
group effects, the thrust of these later studies was more
toward identifying the interaction characteristics which
lead to such differences— those between the teacher and the
individuals in the class. Do teachers with certain identi
fiable characteristics interact with students with their own
identifiable characteristics is the more appropriate re
search question (Sperry, 1972; Witkin & Moore, 1974).
This new research direction hopefully represented
the end of simple expectation effects studies, although some
were still attempted. While it is both probable and logical
that expectation somehow influences teachers to interact with
students differently and that the many repetitions of these
interactions result in progressively reinforced behavior on
both sides, this elaboration is hardly an explanation of a
science..__________________ ______________________________________
21
1 A better elaboration of the behavioral dynamics of
expectation is found in the specifications for a model of
fered by Brophy and Good (1974). They state that only when
all the requisite steps are present will there be a valid
instance of self-fulfilling prophecy effect. For this cause
and effect relationship to occur and solidify through suc
cessive reinforcement, the following events are necessary.
The model below is especially useful in dispelling the mys
tical qualities of the self-fulfilling problem and thus in
providing an understanding of behavioral elements instead.
; A Model for Teacher Expectation
1. The teacher expects specific behavior and achieve
ment from particular students.
2. Because of these different expectations, the teacher
behaves differently toward the different students.
3. This teacher treatment tells each student what
behavior and achievement the teacher expects from
him and affects his self-concept, achievement moti
vation, and level of aspiration.
4. If this teacher treatment is consistent over time,
and if the student does not actively resist or
change it in some way, it will tend to shape his
achievement and behavior. High-expectation stu
dents will be led to achieve at high levels, while
the achievements of low-expectation students will
decline.
5. With time, the student's achievement and behavior
will conform more and more closely to that origi
nally expected of him. (Brophy & Good, 1974, pp.
39-40)
22 |
Based upon this understanding of expectation, Brophy!
I
and Good (1970) argued strongly for the pursuit of natural- j
istic expectation process studies. They observed that in
ference problems from laboratory to classroom are minimized.1
While these studies lack flexibility in manipulation of
experimental variables and are easily affected by unknown
extraneous influences, they serve to assess classroom inter
action better than laboratory studies. To implement these
studies, coding schemes like those of Flanders (1970), 1
Flanders and Amidon (1961), and Good and Brophy's (1970)
Didactic System were carefully devised, with attention given
i
to trend, causal factors, and observation of frequency.
Using variants of these classroom coding systems,
many such studies were completed on classroom interaction
phenomena during the post-1970 period. Some, such as Mack-
ler (1969) related the pervasive and sometimes unfair aspects
of teaching systems which are especially disturbing in some
of the European school systems (Burstall, 1968). By means
of plotting the frequency and evaluating the quality of
teacher-student contact, Good and Brophy (1972, 197 3) estab
lished firmly that significant classroom interaction dif
ferences prevailed, which was reaffirmed by Brophy and Good
in 1974. In addition, classification of global preference
23 I
systems of teachers toward their students was accomplished !
by identifying four categories through teacher ratings:
indifference, concern, attachment, and rejection (Jenkins,
1972; Silberman, 1971). Through Jenkins' studies, teacher- i
student treatment differences on a number of variables, such
as criticism, praise, and behavioral contacts were identi
fied. Later studies took these major teacher rating cate
gories and analyzed those student characteristics most fre
quently occurring in each category (Willis & Brophy, 1974). j
Unfortunately, the dynamic processes underlying the
.teachers' preference systems were not, however, studied in
i
the expectation context, aside from a few studies which ten
tatively identified such categories as sex preferences and ,
i
teachers preferences toward more orderly, cooperative stu- !
i
dents. The studies indicated that not only were these subtle
processes occurring (Silberman, 1971), but that certain
teacher-student interactions (Heil, Powell, & Feifer,.
1960) could be predicted from individual teacher and student’
characteristics not related at all to ability factors or
traditionally defined educational processes (Ostfeld & Katz,
1969).
It is this point which is perhaps the most crucial:
that for a variety of reasons, certain students were more
24
salient (a) in general and (b) even more so in the percep
tual field of certain teachers. This suggests the importance
'of match and mismatch between student and teacher as a par-
i
tial predictor of performance outcome (Silberman, 1971). i
Without any empirical justification, merely postu
lating match and mismatch between student and teacher is
also unworthy of a science. While it is a fair assumption
that a teacher's expectation or perhaps conditioned attitude
toward certain student classes may make him treat some stu-
!
dents more appropriately than others, it is the subsequent
, classroom interactions which determine the impact of the
i
teacher's preference systems (Rist, 1970; Rosenthal & Rubin,
19 71). For example, Brophy and Good (1974) stated: "The
teacher who waits patiently for a response form one student
but gives up easily on another . . . or calls on others for
easy ones . . . is communicating expectation in indirect
ways" (p. 117). In a similar way, teachers who demand quick
responses and are impatient while waiting for answers en
courage (condition) students to raise their hands only when
they are sure they know an answer and encourage (condition)
anxiety when they are not sure. It becomes more apparent
that teacher and student factors interact importantly. It
also becomes apparent that this interaction may be one of the
25 ;
reasons expectation effects occur in some studies and not in
others.
This is further compounded by the fact that some
teachers allow expectations and student differences to in- 1
fluence them and others do not (Feshbach, 1969). This was
elaborated by Brophy and Good (1974) in a detailed discus- ;
I
sion of several stable teacher characteristics which they j
called proactive, reactive, and overreactive. The pr:oaot'i:ve
teacher, they held, does not react to student differences
passively, but attempts to understand and compensate for
them so that interaction patterns are not arbitrarily stu-
I
dent-determined. The reaot'ive teacher, on the other hand,
does not form inappropriately rigid or extreme expectations
i
or engage in behavior tending to exacerbate student differ
ences but does little to compensate for them. The overreao-
tive. teacher is most susceptible to negative effects and is
i
often conditioned by student behaviors, in effect heightening
the differences from what they were originally. The conclu
sion which may be drawn from the literature review, in the writ
ing of Brophy and Good (1974), is: i
Pattern of teacher communication . . . is not uni
versal across teachers, although it appears in many.
This indicates that susceptibility to teacher ex
pectation effects is an individual difference variable
and [that] data are needed to identify the teacher
characteristics which make teachers more or less sus- j
ceptible to such effects. (p. 110) I
i
While the expectation studies point conclusively j
toward significant differences in teacher-student interac- i
tive processes attributable to both the personality factors
of each and the resultant classroom conditioning processes ]
i
which tend to influence teacher and student behavior in sub-!
sequent encounters, the studies do not resolve other issues.
Specifically, they do not account for why some students are
more salient to teachers and how the actions of students
interact with the personalities of teachers and how the
actions of students interact with the personalities of
teachers and, of course, the reverse— how the actions of
teachers interact with the personalities of students. Conse
quently, their value to educators is limited (Hunt & Sulli
van, 1974) .
1 Inter act-ion of Ind'Lv-idual
Teacher-Student Factors
Witkin and Moore (1974) , however, developed several
I
theories which, based upon earlier studies of psychological 1
differentiation (Witkin et al., 1962, 1972), suggested that
in addition to many of the traditional SES, IQ, race, and
sex characteristics, cognitive style similarity among the
field dependence-independence dimension is important in
27
determining the nature of the teacher-student relationship. |
Witkin (197 3) stated:
i ,
Especially important in its implications for how j
the teachers evaluate their students' abilities was the !
■ finding that teachers valued highly the intellects of
students close to themselves in cognitive style. Simi
larly, students viewed more favorably the personal 1
characteristics and cognitive competence of teachers
close to themselves in cognitive style, (p. 33)
i
Witkin (197 3) further suggested that such findings
made it far too simplistic to ta'lk about good students and
|bad students or good teachers and bad teachers even though
"by some criteria of competence such designations may be
justified" (p. 33).
Witkin et al. (1962, 1973, 1974) see these style
labels as clusters of characteristics, both cognitive and
.personal, justifying the designation of personal styles.
Summarizing the tendencies characteristic of the different
styles they acknowledge a separation between perceptual and
|
cognitive but nevertheless a consistency of behavior. Para-'
phrasing from Witkin et al. (1962, p. 80) this idea is de
fined in more detail in the following paragraphs.
A tendency toward an analytical or global way of
experiencing characterizes a person’s problem-solving activ
ities as well as his perception. The term "analytical field
approach" has been adopted for the style of functioning rep
resented in both the perceptual and intellectual behavior of
an individual, which involves the ready ability to overcome 1
j
1
an embedding context and to experience items as discrete !
from the field in which they are contained. The term j
"global field approach" has been suggested to describe the |
style of functioning that involves submission to the domi- !
nant organization of the field and the tendency to experi- !
ence items as "fused" with their background. Field-
dependence-independence represents the perceptual component ■
of this broader dimension. Individual performances are rep
resented continuously along the analytical-global dimension
of experiencing, rather than constituting distinct "types."
When we say that a person shows an analytical or global
field approach, we mean only that he falls above or below
i
the mean of his group on this dimension.
The pervasive self-consistency in cognitive func
tioning that has now been demonstrated suggests that the
classical division into the perceptual and intellectual is
of limited value in the study of cognitive styles.
i
The significant relation frequently reported between
measures of field dependence and total standard intelligence
test scores is "carried" largely by those portions of intel
ligence tests which require analytical functioning. In
i ■ 29~]
I
other words, the relation is based on the expression of a j
particular style of field approach in both. More will be I
said about the perceptual correspondences in a later con- j
text; however, one would expect identification by the 1
teacher and student with those style characteristics closest
to his own and the expressed preferences which are charac-
r
teristic of such identifications.
I
Witkin and Moore (1974) define cognitive style in
|several ways. Perhaps the simplest is "the observation of
self-consistency in an, individual's way of handling a wide
range of perceptual and intellectual tasks" (p. 1). Closer
to the point of classroom information processing, they state:
Whatever we call them, these styles may be con
ceived as our characteristic ways of processing infor-
( mation, regardless of whether the information had its
primary source in the world outside or within ourselves
and, when in the world outside, regardless of whether
the information is provided primarily by things or by
other people and their doings. (p. 2)
i
In general, perception of the field-dependent person
is guided by the organization of the field which he perceives
as a whole. His perception is global. Perception of the ;
field-independent person separates elements of the field
from their context. His perception is characteristically
analytical. Perception of the social field follows the same
patterns: field-dependent people are sensitive to people-
30 !
related situational elements, whereas field-independent per-j
i
sons are more impersonal and concerned more with abstrac- !
tions.
With Witkin's definitions in mind, it would appear
i
that this is an appropriate time to restate the behavioral
analysis of teacher-student relationships. The conclusions
based on the literature review of the previous sections
strongly suggest the existence and operation of a teacher-
to-student selection process based upon a fairly large num
ber of student stimuli but a much smaller number of teacher-
specific factors. Brophy and Good (1974) reported evidence
i
of such selection processes in such diverse areas as teach
ability (Thelen, 1967), where teachers selected students
.based upon suitability with their teaching aims; preference
for programs (on the part of the student teachers) which re
inforced their teaching styles (Rubin, 1971); and preference
for lecture vs. written format (James, 1962). Tobias (1972)
found significant differences in the performance of students
who were allowed to select instructional style over that of
students randomly selected.
Matching for style similarity also seems justifiable
on a logical, if not practical, level because similarity
produces mutual attraction (Lindzey & Byrne, 1968; Marlowe
31
& Gergen, 1969) . The reinforcing nature of this phenomena
has strong face validity and, at the very least, should lead
to affective gains.
i
Evidence exists on other than purely logical levels l
I
such as the interaction of leadership styles, ordinarily
difficult to assess, because of confounding with character- |
istic and culturally defined sex roles and biases. Never-
i ,
theless, in the educational areas, Feitler, Wiener, and ;
■Blumberg (1970) related personal needs to preferred class-
iroom leadership styles. Tuckman (1969) studied the inter
action between teaching styles and abstract independent vs.
concrete dependent personality in high school students and
concluded a general preference for nondirective teachers,
but also showed that concrete independent students tended j
not to differentiate in their preferences, indicating, as it
were, an insensitivity to teaching styles.
I
Preferences for thinking style, convergent and diver
gent, are also selection elements to be considered. Joyce
and Hudson (1968) concluded that divergent students did bet
ter with divergent professors in a medical statistics course. ,
Zussman and Pascal (197 3) found a preference bias by teachers
for students with styles similar to their own in terms of
attributed need for attention, need for achievement, and
32 |
I
achievement rating. While warmth in teachers produced gen
erally higher student achievement (Baird, 1973; Sears & Hil-
gard, 1964), certain studies show that for some students
warmth is especially important (Kleinfeld, 1972) , especially
among minority children attending "majority" schools. Mid- ;
dleman (1972) similarly found differential sensitivity to
facial and verbal instruction content among suburban middle
class and urban lower class children. For teachers' atti
tudes about children, cautious interpretation of Golden-
i
berg's (1971) findings relating to social desirability dif
ferences between middle class and lower class "head start"
teachers was suggested by Brophy and Good (1974).
Witkin and Moore (1973, 1974) argue that many of
i
these elements mentioned heretofore are representative of a j
more global characteristic frequently called cognitive style
(but more appropriately, personal style). They further con-'
tend that a variety of factors are influenced by this style i
but especially by the field-dependence-independence compo
nent. The choice of speciality area and especially the
choice of teaching as a profession reflects a strong field-
dependent orientation. The selection of mathematics over
teaching in the social sciences, however, is likely to re
flect a stronger field-independent orientation. Teaching
I
methodology may also vary according to orientation with the :
field-dependent teacher tending to favor the discussion-
i
discovery method (Ohnmacht, 1967). Wu (1967), examining j
dogmatism combined with high field dependence, found that it;
was associated with effectiveness in handling student logic
errors. Yando and Kagan (1968) found evidences of incorpo- j
i
ration of teachers' impulsive and reflective styles among ;
children taught for the reflective style. Witkin drew fur
ther -evidence from the counseling relationship in differen- ■
tial therapy styles for field-independent vs. field-
dependent patients (Greene, 1972).
, Witkin (197 3) suggested that the impact of social
reinforcement, in the form of praise or criticism, has a
.predictably greater effect on field-dependent children. j
Field-independent students tend to apply inner standards to
their performance.
DiStephano (1969), as reported in Witkin (1973),
provides some evidences on the consequences of style matching
and mismatching. Using extremely field-dependent and field-'
independent teachers, he had them describe their students
matched for the same characteristics. The results strikingly
supported the hypothesis for style preference between student
and teacher.
34 j
Lange (1972) studied the effects of style matching
among nursing education students, concluding that matching j
styles resulted in higher mean grade achievements, lowered |
withdrawal rates, more favorable perception of faculty, and I
more favorable perception by faculty. Impressively, Lange i
observed, the greater the number of style elements the
teachers and students had in common, the higher was the
probability of passing the nursing courses. Phillips and
Sinclair (197 3) found higher scores for involvement, auton-
i
omy, and positive responses among conceptual style matched
I
student teachers and faculty.
There appears to be clear-cut evidence that people
matched in cognitive styles get along better, see each
I
other in a more favorable light, and generally favor one j
another. We can speculate about the behavioral dynamics of
this preference, that similar behavior reinforces one's own
belief system. Finally, we can also conclude that field-
dependent people would be more sensitive to external social
cues as reinforcers than their field-independent counter
parts. They even differ in terms of use of interpersonal
space. Justice's (1969) findings show field-dependent people
seeking closer contact than field-independent.
35 I
Studies by Luborsky as reported by Witkin (1973) j
support the fact that field-dependent people make fewer \
I
self-references in their speech and field-independent per
sons use the personal pronoun most often (see also Doob,
1958).
What can be concluded from the literature reviewed
herein is that the differential classroom interaction fac- I
tors and differential expectation effects arise from the
personal traits that the teacher and student bring with them
i
to the situation. This coalescence of research strongly
^suggests further analysis of these combined forces.
i
This specific study investigated the cognitive style
element of both teacher and student to determine the influ
ence on educational outcomes. In this manner, the study can(
make a significant contribution to education by defining
some of the selection dynamics underlying those teacher
selection factors described in the work of Silberman (1971)
and Jenkins (1972) as attraction, concern, indifference, and
rejection for different students. The determination of stu
dent personal factors which may serve as stimuli for differ
ential treatment becomes a very important, almost a critical,
issue. Evidence accumulates that such personal traits as
sex, race, dress style, speech patterns, scholastic perfor-
mance, and personality are all contributing stimuli for dif
ferential treatment and the search continues for even more j
global determinants. Witkin et al. (1962, 1972, 1973, 1974)J
suggest that the cognitive style factor may be one of these I
global differentiating factors and support this contention |
with the growing body of experimental data cited herein. >
(
\
The particular cognitive style dimension of field dependence-
independence seems to be a prime candidate since it defines
perceptual and cognitive components. As indicated previ
ously, Brophy and Good (1974) reinforce this notion and hold
that susceptibility to teacher expectation effects appears
i
to be an individual difference variable and that data are
needed to identify those teacher characteristics that make
them more susceptible to such effects.
Among the large number of characteristics for field
dependence and field independence, Witkin and Moore (1974)
I
suggested that the following distinguished best between the
two major systems found in students:
Field independent students are characterized by
ability to rapidly restructure problems, reasoning
deductively, ability to dissemble salient problem ele
ments, ability to ignore extraneous distracting evi
dences/facts, non-reliance on peer group defined strate
gies, freedom from socially defined pressures, behavior
modeling from within, more attuned to factual material
than to social, choice of careers dictated by factual
content, rigidity in thinking strategies, sex role inde
pendent, receptive to lecture vs. discovery learning,
relatively uninfluenced by social reinforcers and is
less concerned with close physical proximity. . . .
i
$
The characteristics of field dependent students are:
slow to restructure/solve reasoning problems, context
bound by problem situation, easily distracted by ex
traneous social stimuli, influenced by peer group pres
sures, models behavior based on prevailing activities,
high social/low factual orientation, career choice in
social context, ease of strategic thinking changes, be
havior determined by defined sex roles, positive reac
tions to interaction method of learning, high influence j
i
of social reinforcers and prefers closeness to people,
(pp. 26-29)
The logic of this experiment was to determine whether
i
teachers, when confronted with students showing these charac
teristics, tend to favor students with cognitive styles
closer to their own cognitive style. That is, would teach
ers of a specific cognitive style favor students of that
same style? The experiment proposed to expose field-
independent and field-dependent students to teachers of the
same cognitive styles in order to determine whether such
biases exist. In order to implement this exposure, an ex-
\
perimental situation was devised which made these character
istics more salient. Dependent variable measurements were
collected to determine the numerical magnitude of these
biases as well as the areas in which they occur.
In several respects, this experiment was an impor
tant and logical step in analyzing the dynamics and effects
of teacher-student interactions. It also provided a partial
38
explanation for the equivocal results found in the body of
i
expectation literature. Most importantly, the experiment
established a strong case for style matching or developing
teacher sensitivity through training, to their preference
patterns and classroom biases.
Accordingly, restating the above in the form of a j
research question, this study sought an answer to the ques- 1
tions as to whether the cognitive styles of teacher and stu-!
dent interact in such a way as to produce favorable teacher
expectation effects when these styles are similar and (less/
or) unfavorable expectation effects when they are dissimilar.
Cognitive style is as expressed by the personal dimension
of field dependence or field independence.
Subordinate questions deal with the dependent vari- 1
!
ables:
1. Intelligence estimation
2. Performance assessment
3. School success prediction
4. Preference for style-similar students given a
choice.
The experimental design also permitted the examination of
boy-girl salience factors crossing style preference lines
although the literature does not suggest that such factors
will prevail in this type of experimental setting.
In the null, the above assumed no difference between
field-independent and field-dependent groups or any of the I
above four dependent variable measures. Specifically, the j
null hypothesis is that student teachers classified along I
the field-dependent, field-independent dimension of cogni- .
tive style showed no measurable biases toward students sim
ilarly classified in any of the dependent variable measures:
intelligence estimation, performance assessment, school
(success prediction, or preference for style-similar students
given a choice.
CHAPTER III
METHOD
Overview
The experimenter proposed to investigate the effect
of cognitive style similarity along the field-dependent/
field-independent dimension between teachers and their stu
dents. Of particular interest was detection of the presence
of a bias toward similar cognitive styles. The experimenter
developed and pretested videotapes which emphasized differ
ences in the independent variables, depicting students of
two divergent cognitive styles during microteaching lessons
of very different content. Student teacher subjects, them
selves classifed along the field-dependent/field-independent
;dimension, viewed these videotapes and ranked the students
they saw on their performance during the lessons, their
expectations for subsequent performance in school, their
IQs, and finally, their overall preferences for the students
on the tapes. Dependent variables were responses to a rank
ing instrument which required that the subject rank order
40
the students on a number of elements. Data were tabulated
and analyzed using the chi-square statistic and was further
averaged over discrete categories. ’
Subjects
Subjects comprised two intact classes with a total
of 56 third- and fourth-year teacher education students |
enrolled in two late afternoon classes at the University of
Southern California during the Spring 197 5 semester. Actual
test and stimulus administration occurred in March 197 5. In
each of the two classes all attendees participated (29 in
one and 27 in the other) in every aspect of the classifica- '
tion and experiment. Procedures for assignment to experi
mental groups are described below, these resulting in the
elimination of 10 of the original 56 students, for a total
46 subjects.
Based on Witkin and Moore's (1974) normative data
for the distribution of completion time scores, appropriate
cutoff values were used for the classification of subjects
into the two categories: field dependent for quartiles 1
and 2 and field independent for quartiles 3 and 4. Witkin
(1973), and Witkin et al. (1971, 1974) have observed that
these scores are distributed approximately normally; there
fore, dichotomizing about the mean for classification pur
42
poses was considered to be inherently problematical.
Accordingly, responses attributable to those subjects with
scores falling between the mean and plus or minus approxi
mately .25 standard deviation units were not used in the
subsequent experimental treatment since these scores could
not adequately distinguish between the classifications of
interest. As indicated, this resulted in the elimination of
10 cases that fell within these approximate limits. All
responses were nevertheless collected (see Appendix A). The
procedure provided the pool of 46 subjects for subsequent
assignment to experimental treatments yielding a ratio of
i
16:30 men to women and 23:23 for field dependent and^field
independent, respectively.
Reliability and Validity of
the Test Instruments
The group administration form of the Embedded Fig
ures Test (EFT) was administered to all students prior to
presentation of the stimulus videotapes. This test is copy
righted as the Group Embedded Figures Test (GEFT) and was
developed by Witkin, Oltman, Raskin, and Karp (1971) . This
test requires subjects to find 8 simple geometric forms
embedded within 25 progressively more complex figures and
then to trace the outlines of these forms within specified
43
i
time limits. The test was developed in order to overcome |
I
the impracticalities of the individual administration of the;
EFT when it is desirable to test and classify larger groups.
Regarding the split half reliability for the GEFT, Witkin
et al. (1971) observed:
Since the GEFT is a speed test, an appropriate
method of estimating reliability is the correlation
between parallel forms with identical time limits. }
Correlations between the 9-item First Section scores
and the 9-item Second Section scores were computed and
corrected by the Spearman-Brown prophecy formula, pro-
j ducing a reliability estimate of .82 for both males
(N = 80) and females (N = 97) . These reliability esti
mates compare favorably with those of the EFT. (p. 26)
i
With regard to validity, Witkin et al. (1971) based
their estimates on the more extensively tested EFT and
another conceptual assessment instrument, the Rod and Frame ,
,Test (RFT) and its portable version, the PRFT. They state: ,
There are several ways of assessing the validity of
the GEFT. Since the test is intended as a group form of
the EFT, the most direct criterion measure is the "par
ent" form of the test, namely the EFT. In one study,
subjects were administered the Second Section in its
group form and the Third Section as an individually-
administered test using the items in their original
colored form. Another group was given the Second Sec
tion individually and the Third Section as a group
test. (p. 28)
The correlations of the above, corrected for reduced test
length and combined for the two groups, are reported in
Table 2.
Table 2
Validity Coefficients
Population N Criterion Variable r with GEFT Score*
Male undergraduates 73 Individual EFT, solution time
i
•
00
to
Female undergraduates 68 Individual EFT, solution time -.63
Male undergraduates 55 PRFT, error -.39
Female undergraduates 68 PRFT, error -.34
Male undergraduates 55 ABC, degree of body articulation .71
Female undergraduates 68 ABC, degree of body articulation .55
*r's with the EFT or the PRFT should be negative because the tests are scored in reverse
fashion.
Source: H. Witkin, P. Oltman, E. Raskin, S. Karp, Manual for the Embedded Figures Test
(Palo Alto, California: Consulting Psychologists Press, 1971), p. 29.
4^
4*
45
Witkin et al. (1971) also related that the GEFT
could be measured for validity against other measures of j
psychological differentiation; for example, the degree of . ;
articulation of the body concept which is assessed by means
of scales applied to human figure drawings. One of these is
scale "ABC" (Witkin et al., 1962).
Finally, Witkin et al. (1971) provide an important 1
rationale for the use of these perceptual tests as valid
( measures of cognitive functioning in a succinct statement
found in the introduction to the test user's manual:
The EFT is a perceptual test. The subject's task
on each trial is to locate a previously seen simple fig
ure within a larger complex figure which has been so
organized as to obscure or embed the sought-after simple
figure. In the strictest interpretation, therefore,
scores on the EFT reflect extent of competence at per
ceptual disembedding. Individual differences in EFT .
performance, however, appear to relate to more than dif
ferences in perceptual functioning. Numerous studies
. . . , have shown that the ability to "keep things sep-,
arate in experience" in the EFT, signifying in effect
greater differentiation in perceptual functioning, mani
fests itself in congruent form in other areas of the
person's psychological activity, signifying greater dif
ferentiation in these other areas as well.
The concept that perceptual and intellectual (that ,
is, cognitive) tasks may serve in the assessment of 1
broad, salient dimensions of personal functioning is a
long-standing one in the history of psychological test
ing. Ever since the introduction of intelligence tests,
patterns of abilities revealed in test performance have
been used in clinical appraisals of ego functioning.
Similarly, perceptual tests like the Rorschach, in use
for many years in clinical assessment, have also fol
lowed the rationale that from the way in which an indi-
46
vidual perceives particular stimuli, inferences may be j
drawn about his personality. |
The specific rationale for using the EFT to assess j
broad dimensions of personal functioning comes from
cognitive-style theory and the evidence accumulated in
the course of its extensive research application. In
brief, cognitive styles are the characteristic, self-
consistent modes of functioning which individuals show ,
in their perceptual and intellectual activities.
(pp. 2-3) I
j
Ptlot Study '
i
The use of a videotape as a stimulus presented some
methodological problems. Accordingly, a pilot study was
conducted several weeks in advance, in order to investigate
stimulus saliency, videotape presentation order effects, and'
i
male-female response characteristics. Initially, a list of
student characteristics derived from the work of Witkin
(1973) and Witkin and Moore (1974) and that of Castaneda and;
Gray (1974) was compiled. Eight students of superior intel-'
i
ligence, all in the sixth grade, were identified with the
aid of two sixth-grade teachers who used the listed charac
teristics as a guide. These teachers were thoroughly famil
iar with the students, having been their official teachers !
during the previous semester (Fall 1974). 1
The eight children, in two groups of two boys and
two girls each, were videotaped under permissive discussion
like conditions for over 40 minutes. A checklist was
47
i
with Likert scales designated for each of seven behaviors,
characteristic of field-dependent and field-independent stu
dents. The videotapes were administered to eight in-service!
teachers who rated the students they saw on these seven seg-!
ments of behavior or apparent student orientation. The rat
ings were compiled and those students scoring highest in the
desired categories, characteristic of the field-dependent :
or field-independent orientations were identified. Subse
quently, all the students were tested using the EFT, with
those having the most distinguishable scores on this instru-
♦ F
'ment compared with the teacher ratings. Four students, two
'boys and two girls, each of the field-dependent and field-
independent orientations were chosen to participate in the
subsequent pilot study which entailed acting to two scripts.;
I
The scripts depicted interactive lessons in physical
science and social science areas and employed an in-service
I
teacher who also acted according to designated lines and
responses. Response opportunities for the students were
carefully meted and responses were restricted to the script
format; however, each of the students had the opportunity to
act to his or her character in whatever way they chose. The
responses of the field-dependent students were made more
salient in the social science script, while those of the
48
i
field-independent students were more salient in the physical
■science script. This was done in order to increase the
stimulus value of the children within the brief exposure
time afforded by the videotape. It was further based on
Witkin and Moore's (1974) prediction regarding performance
superiority by the children of the different cognitive ;
styles.
The tapes were transposed--in one case the physical
science tape was first and in the other the social science
.tape was first. They were administered to a class of 4 0
Icommunity college students. These students had previously
l
been classified using the Group Embedded Figures Test (GEFT)
with the resulting distribution of 23:17 for field dependent
and field independent, respectively (see Appendix B for j
sample GEFT). Analysis of results revealed that there were
no effects attributable to order of videotape presentation.
i
Further, analysis indicated no pattern of male-female bias
toward the videotape students.
Subsequently, pilot administration of these video
tapes to a class of teacher education students revealed that
the acting superficiality overshadowed other considerations.
Based upon this response, new tapes with new students were
developed, using procedures similar to those described for
49 ;
the pilot study. The principal exception was that responses
were spontaneous, although the new students were directed to,
respond in a manner which would increase the stimulus value j
of those characteristics which they actually possessed. The
videotapes were also increased in length in order to provide
a greater amount of time for the subjects to view and hear
the two lessons although the content, specifically the learnf
ing objectives and lesson material, were identical. ■
Mater'lals
Seven sixth-grade advanced track students were pre-
iselected, using the same teacher selection strategy as in the
pilot study. The GEFT was administered with the values for
each student compared to the list of selection criteria used
.by the teacher. Four students were again selected, two boys
and two girls, with GEFT scores well within the first (0 to
.9 problems solved) and fourth (15 to 18 problems solved)
quartiles described by Witkin et ai. (1971). (Appendix C
contains selection criteria and GEFT scores.) A teacher was
selected and given the lesson objectives which she translated
into teaching strategies. The teacher was further directed
to take deliberate steps to elicit student interactions with
the subject matter through any or all normal classroom meth
ods. She was given a set of instructions (Appendix D) which
, 5 X ) _
I
identified the objectives of the study and which enabled her
to create the strategy which would best accomplish the 1
I
l
objectives; specifically, to teach in a manner which would
i
bring out each student's unique characteristics. I
The students were briefed on the desired behavior
and examples of desired interactions were given without ref-'
i
erence to the lesson content. For example, an instruction ;
to a field-independent boy, Gregg, was to show his factual
orientation, which he possessed (and knew he possessed), as
much as possible in the physical science tape. Similar in-
structions were given to the other students. The students
.were then permitted to see themselves on tape in the identi
cal setting as would be used in the actual taping, this to
minimize any on-camera trepidations. However, none were
apparent.
In order to permit maximum resolution of facial
'expression during the lessons, the four students and teacher
were positioned around three sides of a rectangular work/
conference table. The fourth side of the table was vacant,
facilitating maximum TV view of the participants with mini
mum camera slew. Nameplates were positioned in front of
each participant, with the first name of each student and
the full name of the teacher, Miss S. This was done to
51 !
facilitate name recall during the ranking of students which ,
would follow the tape presentation. A variable focal length
lens was used on the camera, which permitted close (zoom)
focus during individual monol'ogues and during the question-
and-answer portion of the tapes.
A word list was used for each of the two concepts
i
i
discussed, supplemented by a diagram for the physical sci- 1
i
,ences material. These lists are found in Appendix E. The
■first of the two concepts was introduced by the teacher after
a brief personal introduction by all participants, during
which name, grade, and personal interests were described.
i
The first concept was Aerodynamics of Flight, a closed ended
lesson on the elementary physics of flight. The objective
of the lesson was to show that in order for an airplane to ;
fly, two forces— friction (drag) and gravity— had to be over
come by the other two forces— thrust and lift. The elements
l
of power output and airflow were explained and, based upon a1
brief evaluation conducted later, were completely understood
by all participating students. The second lesson was based
on Crime and Punishment, during which the nature of punish
ment was discussed in terms of its deterrent values, its
public protection aspects, and its inability to solve social
ills. While no closure was obvious, general agreement was
52
reached on the purpose of imprisonment, the role of rehabil-j
itation, and the reasons for some crime. !
The first lesson lasted 19.5 minutes and the second ;
approximately 20 minutes. A short break was scheduled be
tween the two (approximately 8 minutes), but was edited from '
the tape, giving the appearance that only 30 seconds elapsed
between the two videotapes. !
Design of the lessons was specifically directed
toward enhancement of response opportunity for the two cog-
i
nitive styles represented. In the first (physical science)
videotape, the desired behavioral elements were, in general,
i
the attraction of field-independent students to the more
factual domains. Their incisive and analytic abilites were
.of importance, as was their ability to put to use strategies;
developed in other unrelated contexts. In this respect,
drag, the frictional component of flight, was seen by the
i
two field-independent students as analogous to moving a
heavy, broad-based object across the floor. Similarly, these
students saw immediately and spontaneously that drag could
be reduced by reducing the leading surface area of the wing i
of the plane, during the teacher's skillful illustrative
example of a piece of paper flying through the air. While
the field-dependent students also contributed, their input
53
4
was minimal compared with that of the field-independent
students.
I
l
The second lesson, relating to the social sciences, j
was similarly designed to enable the field-dependent student!
i
to excel in his characteristic areas. The ability to see j
quickly the contextual implications of a problem, specifi- |
cally, that one crime is not equal to another crime and that
punishment without rehabilitation is futile, was considered
most important. In a similar manner, the field-dependent
I
students quickly saw that deterrence and protection of soci-
ety by imprisonment of criminals had little to do with re
habilitation per se. Their sensitivity to these problems
was apparent and was so indicated by the subsequent evalu
ations provided by observers.
Response opportunities were of importance: if one
student had more responses (particularly correct responses)
than another, the results could be confounded. That is,
frequency could perhaps be confounded with salience or, more
importantly, with the stimulus value of a particular studentL
Responses were tabulated and appear in Table 3. As noted,
they are approximately equal over the two tapes.
To validate the effectiveness of the stimulus char
acteristics, the videotapes were shown to two groups of four
Table 3
Videotaped Direct Response Tabulation
Student Aerodynamics
Crime &
Punishment
i
i
1
Total
i
FieId-Independent
Female
14 4 18
Field-Independent
Male
15 6 21
Field-Dependent
Female
5 14 19
Field-Dependent
Male
6 12
1
18
\
55 ;
i
teachers, each with instructions to attend especially to the|
characteristics of importance. It was further requested
that the teacher watch and listen for these characteristics
(so that they could later rate the four students according
to the degree to which they were manifested.
After the viewing, a questionnaire similar to that j
used during the pilot study but revised to reflect three I
additional characteristics, was administered, using the same.
Likert scale for each response. Appendix F shows this scale
and the results of the tabulation of responses. These de
pict the average of eight ratings for the ten characteris-
i
tics and show a strong correspondence with the character
istic cognitive style-related elements predicted by Witkin
(197 3). These stimulus elements were effectively communi
cated through the videotapes as shown.
Methodological Limitations
Somewhat limiting the scope of the investigation was
the use of a female teacher in each of the videotaped micro
teaching sequences. This decision was made with the elemen
tary schools' traditional preponderance of female teachers
in mind. Similarly, the choice of a teacher with warm and
outgoing personality characteristics was an implicit value
judgment but also was strongly based upon the prevalence of
such teachers in the elementary grades. (
i
To simulate actual teacher-student interaction in an
!
environment which would permit the children of the differ
ent cognitive styles to become salient— that is, to excel in
his or her grasp of a particular subject— two very different
content-topics were chosen: Aerodynmaics and Crime and
Punishment. In this manner, each set (e.g., field-dependent
boy and girl) of children mastered a particular area and was
very apparent in doing so in the videotape. While this
technique facilitated the conduct of the experiment and pro
vided the data for subsequent analysis of style-similarity
bias regardless of performance on-task, the idea of a third
"neutral" subject matter was considered and discarded. The
reason for discarding this idea was that if cognitive-style ,
biases were apparent in the two extreme subject areas chosen
and if these biases prevailed in the face of superior stu
dent performance by style dissimilar students, they would
almost certainly appear in some intermediate subject matter
as well.
Finally, the selection of one boy and one girl each
of the desired cognitive style (field dependent and field
independent) for the videotaped microteaching lesson was
made with the intention of controlling for sex biases which
have often appeared in such studies (Silberman, 1971).
While the use of more than one of each could conceivably
alter results through a factor of additive influences, there
is little in the literature to suggest that this would occur
i
in a setting where the responses of the children are regu
lated by the script used to simulate the lesson. In the
same sense, different children with different personality
characteristics could result in nonreplicatability of the
findings of this experiment, but personality of course en
compasses cognitive style elements as well.
Procedure
The classification instrument, the GEFT, and the
videotapes were administered to two late afternoon teacher
education classes at the University of Southern California
during the Spring 1975 semester. The classes in each case
were told by their respective professors that they were
being asked to participate in an important educational ex
periment and that their cooperation was appreciated. Test
booklets for the GEFT were distributed with instructions
being read verbatim from the test manual.
The initial instructions for the first nine simple
figures were:
"Now start reading the Directions, which include 2
practice problems for you to do. When you get to the
end of the Directions on Page 3, please stop. Do not '
go beyond Page 3 .1 1 !
"Are there any questions about the directions?" " j
(E should pause to allow questions.) "Raise your hand |
: if you need a new pencil during the test."
E then says: "When I give the signal, turn the page
and start the First Section. You will have 2 minutes
for the 7 problems in the First Section. Stop when you
reach the end of this section. Go ahead!"
Two minutes were allowed for completion, at which
time the stop all work instructions were issued. During
'completion of this initial set, little cross-conversation
was apparent. The second and third complex sets of figures
■were administered with the following instructions:
"When I give the signal, turn the page and start the
Second Section. You will have 5 minutes for the 9 prob
lems in the Second Section. You may not finish all of
them, but work as quickly and accurately as you can.
Raise your hand if you need a new pencil during the
test. Ready, go ahead."
Five minutes were allowed for solving each of these
'sets of figures and as a result of the increased complexity,
many subjects failed to complete all solutions. Cross
conversation was discouraged during these administrations,
however, as contrasted with the earlier (first) administra
tion, during which there was considerably more vocalized
concern shown over the inability to "see" the simple figure
embedded in a more complex one.
59
The test booklets were identified and collected and
the subjects were told that they would next be required to
view two videotapes and make several judgments about them.
A list of characteristics was written on the blackboard and
the subjects were told to watch for these during the ensuing;
videotapes. The list comprised the same personal charac
teristics as used in the prevalidation and is identical to 1
i
that shown in Appendix I. These characteristics were:
Factual orientation
Social orientation
Degree of independent thinking
Sex role dependency
Peer group orientation
Social pressure susceptibility
Sociability
Shyness
Rigidity in thinking
Distractability j
Expression of positive feelings
Seeking of social rewards 1
Task orientation
i
While this list was long for the short stimulus
presentation planned, it was indicated that many of these
characteristics were on opposite ends of a continuum with
others on the list, such as shyness and sociability. The i
object of using the list was to sensitize the subjects to
the differences in the personality characteristics of the
children they would be seeing and it was further indicated
that they would not have to rate or rank these children on
j ' 60 1
the blackboard list of items. ,
The videotape ranking record (Appendix G) was then
distributed. It included a printed set of instructions,
identification information required, and the six questions '
related to the dependent variable measures— two questions on
ranking of on-task performance, two on expected school per
formance in related areas, one on IQ estimate (rank only),
and a final question on the overall preference for those
children shown on the tape. The instructions were read ver
batim and the ranking scheme was explained. There were no
'questions asked on the meaning of ranking or the need to
i i
rank order the children who were to be seen.
i
It was indicated that the videotapes represented
two microteaching sequences using a real teacher and four
sixth-grade students she had not seen before (prior to that
day of videotaping). The subjects were told that the tape
'would be stopped at the end of the first sequence, Aero
dynamics, at which time they were to respond to the first
question which requested a ranking of the four students on
their understanding of that subject matter. A diagram of
the seating arrangement along with the names of the children
was placed on the blackboard and it was further indicated
that nameplates would be in front of each participant. As
61
l
there were no questions about the procedure, the room was
i
dimmed and the first videotape was shown, using a SONY
:(Model 3600) Videotape Recorder and Player along with a
Packard-Bell 24" (class A) Monitor-Receiver mounted at a
48" height in front of the classroom. No interruptions
occurred during this or subsequent portions of the videotape!
showing. 1
At the end of the tape, the room was illuminated and
the first response was completed. It was indicated that
i
1 several minutes had been provided for a break for the chil
dren during the initial videotaping but that the time was
not apparent from watching the tape. The room was again
dimmed and the remaining tape, Crime and Punishment, was
presented. This tape lasted approximately 20 minutes and
was viewed uninterrupted by any external event. The room
was again illuminated and the balance of the questionnaire/
I •
ranking form was completed.
The experimenter paced the completion as far as pos
sible by saying, "If everybody has completed the response,
please look up so we can go on to the next one." A few
students went ahead of this pace but by and large the re
sponses were made simultaneously by all. All material was
collected and the videotape equipment was disassembled. The
class was subsequently told that they would be debriefed on
the intent of the experiment and the outcome, as well as the1
implications. !
I
|
Tabulation of GEFT Data
The GEFT were scored using the test manual key and
scoring recommendations (Witkin et al.-, 1971). Only fig
ures in the right perspectives constituted right responses:
extraneous lines invalidated a response, lines drawn in
1 rather than traced similarly negated a response and partial
solutions were wrong. Only one of the students in the two
classes apparently did not understand the instructions and
attempted only one of the problems. Judging by his name, he
appeared to be a foreign student. ;
i
Additionally, forms were devised to collect the
absolute numerical rankings in a form which permitted tabu
lations of mean comparisons: one for each subject (46) for
each of the six questions, and for each of the four stimulus
children. A total of 1,104 data points was taken.
Tabulation of Ranking Response Data
Two methods of response tabulation were used. The
first summed and averaged the rankings assigned to each
stimulus child over the 23 similarly classified subjects and
63 :
for each dependent variable question. Since each of the 4 6 ,
subject ranked each of the 4 stimulus children on each of 6 :
questions, the total of these data points was 1,104 points
; (46 x 4 x 6). Note that the ranks were restricted to 1, 2,
3, and 4, or one for each child.
The second method tabulated the rank frequencies
for each category of interest. Tabulations were made by j
I
stimulus child and by dependent variable questions for each
Respondent group. High rankings were 1 or 2 and low rank
ings 3 or 4. The total of these points was again 1,104
points (46 x 4 x 6). The data were analyzed using the chi
square statistic.
Anaty sis of the Ranking Respons e Data
Analysis of data was performed by using the Statis
tical Package for the Social Sciences routines for chi
square analysis. The International Business Machine (IBM)
360/60 computer at the University of Southern California
Computer Center (UCC) was used. See Figure 1 for an illus
tration of the experimental matrix.
The videotapes depicted the same four children, two
field dependent (boy and girl) and two field independent
(boy and girl) participating in two microteaching sequences.
One of these sequences was designed to maximize the response
opportunities for the field-dependent students and the
other to maximize the response opportunities for the field-
independent students (no presentation order effects were
anticipated, based on the pilot study conducted).
R E S P O N S E S T O
Q U E S T IO N S
O V E R A L L P R E F .
2X2X2
E X P E R IM E N T A L
D E S IG N M A T R IX
P R E D I C T E D • P E R F O R M A N C E (2)
• O N - T A S K P E R F O R M A N C E (2)
V I D E O T A P E D S T U D E N T
COGNITIVE S T Y L E S
F I E L D I N D E P
MAUI
S U B J E C T S
F I E L D
I N D E P .
S U B J E C T S
F I E L D
D E P E N D .
S U B J E C T S
R A N K I N G ON
D E P E N D E N T
V A R I A B L E S
Fig. 1.
Experimental Matrix
CHAPTER IV
RESULTS
As indicated in the previous chapter, chi square
analysis was performed for each of the 6 tabulations of v
ranking responses. These are listed in Table 4 and relate
to the 4 research hypotheses as shown.
i
Table 5 presents the results of computation of 15
chi square 2 x 2 contingency tables. While in theory 24 chi
'square computations were possible, 9 such tables revealed,
upon calculation of expected values, cell frequencies of
well under 5. Chi square statistics were computed but not
presented for these contingency tables (see Appendix H for
complete data). (Computation of Fisher*s exact probabili
ties was not deemed of importance because of the extreme
values observed.)
For the on-task performance ranking hypothesis,
specifically that respondent groups would tend to show a
bias toward the stimulus children closest to their own cog
nitive style regardless o*f their actual on-task performance
65
Table 4
Ranking Responses as Related to Research Hypotheses
Research Hypothesis Ranking Category Ranking Response Number
1. Cognitive Style bias On-Task Performance:
will appear in ranking Crime & Punishment 1
for on-task performance Aerodynamics 2
2. Cognitive Style bias Intelligence:
will appear in ranking Ranking for IQ 3
for overall intelligence
3. Cognitive Style bias Predicted Performance:
will appear in ranking Math & Sciences 4
for predicted performance
in similar (to on-task)
areas
Social Studies 5
4. Cognitive Style bias
will appear in ranking
for general preference
Overall Preference 6
cn
C V
Table 5
Chi-Square* Comparisons Between Field-Dependent and Field-Independent Subjects
Stimulus Children
Field-Dependent Field-Independent
Response Number Male Female Male Female
1. On-Task Performance
Crime & Punishment
Not**
Computed
Not
Computed
Not
Computed
Not
Computed
2. On-Task Performance
Aerodynamics
.034***
Not
Computed
Not
Computed
0
3. Intelligence Estimate 1.36 0 .08 .92
4. Predicted Performance
Math/Science
.37 .35
Not
Computed
.39
5. Predicted Performance
Social Studies
.11
Not
Computed
Not
Computed
0
6. Overall Preference 1.71 .10 1.38 2.21
*Chi squares corrected for continuity, all df = 1.
**Not computed because expected cell frequencies were under
***p > .10
5 (see Appendix H).
i
ov
68 |
in the videotape, there were no significant chi squares
observed. Initially, analysis was attempted for the two
i
sets of on-task performance rankings: Response 2 for the
Aerodynamics subject matter and Response 1 for the Crime and
Punishment material. No computations were advisable, how- ;
i
ever (Seigel, 1956), for any of the rankings for Response l,i
i
Crime and Punishment, because of the numerically inadequate :
expected cell frequencies computed in every case, and only
two of the four were possible for Response 2, Aerodynamics•
For the intelligence ranking hypothesis (Response
3) that respondent groups would show a bias toward those
i
children closest to their cognitive style given the video
tape exposure, no significant chi squares were observed for
the four computed.
The predicted performance hypothesis was measured by
means of two ranking responses: Response 4 (predicted per
formance in the mathematics/science areas) and Response 5
(predicted performance in the social studies areas). The
research hypothesis, as in on-task performance, was that
respondents would show a bias toward the stimulus children
closest to their cognitive styles regardless of the previous
on-task performance evidence, which suggested generalizable
abilities in a particular subject area. That is, given
69
objective evidence of superior performance by those children'
of differing cognitive styles, respondents would still show
i
a bias toward the children closer to their own cognitive j
style. This research hypothesis was not supported. Chi :
squares for Response 4 were computed for three of the four
stimulus children. None were significant at the .10 level.
i
Chi squares for two of the four contingency tables for 1
Response 5 were computed, the others having expected fre
quencies of under five.
The final hypothesis was a general or overall pref-
ierence hypothesis. The prediction was that respondents
I
would rank the stimulus children similar to their own cog-
I
nitive style higher than others based on a general attrac
tion. Four chi squares were computed for this response and .
none were significant at the .10 level.
An additional set of data computations provided
i
means of ranks (cell means) for each stimulus child on each
of the six ranking responses and for each respondent group.
This was done in order to enable a numerical comparison be- j
tween the response of each subject group for the same stim
ulus child. Table 6 presents these comparisons grouped by
the six ranking categories. Column marginals as well as
percentages for these comparisons were readily obtainable
Table 6
C e l l M e a n s 3 f o r A l l R a n k i n g R e s p o n s e C a t e g o r i e s
Response No. 1:
On-Task Performance for Crime and Punishment
Response No. 2:
On-Task Performance for Aerodynamics
Stimulus Students Stimulus Students
Respondent
Subjects
Field-
Dependent
Male Female
Field-
Independent
Male Female
Row
Means
Respondent
Subjects
Field-
Dependent
Male Female
Field
Independent
Male Female
Row
Means
Field-
Dependent 1.91 1.30 3.04 3.73 2.50
Field-
Dependent 2.30 3.60 1.56 2.52 2.50
Field-
Independent 1.91 1.21 3.08 3.78 2.50
Field-
Independent 2. 30 3.47 1.52 2.69 2.50
Column Means*3 1.91 1.25 3.06 3.75 Column Means 2.30 3.53 1.54 2.60
Classification
Meansc . 1.58 3.41
Classification
Means 2.92 2.07
Response No. 3:
Intelligence Estimate
Response No. 4:
Predicted Performance in Math and Science
Stimulus Students Stimulus Students
Respondent
Subjects
Field-
Dependent
Male Female
Field-
Independent
Male Female
Row
Means
Respondent
Subjects
Field-
Dependent
Male Female
Field-
Independent
Male Female
Row
Means
Field-
Dependent 2.56 1.81 2.43 3.21 2.50
Field-
Dependent 2.56 2.86 1.43 3.13 2.50
Field-
Independent 2.39 1.60 2.30 3.69 2.50
Field-
Independent 2.91 2.82 1.17 3.08 2.50
Column Means 2.47 1.30 2.36 3.45 Column Means 2.73 2.84 1.30 3.10
Classification
Means 2.11 2.91
Classification
Means 2.79 2.20
Predicted
Response No.
Performance in
5:
Social Studies
Response No. 6:
Overall Preference
Stimulus Students Stimulus Students
Respondent
Subjects
Field-
Dependent
Male Female
Field-
Independent
Male Female
Row
Means
Respondent
Subjects
Field-
Dependent
Male Female
Field-
Independent
Male Female
Row
Means
Field-
Dependent 2.08 1.56 3.04 3.30 2.50
Field-
Dependent 1.52 2.78 2.73 2.95 2.50
Field-
Independent 2.08 1.30 3.13 3 ,.47 2.50
Field-
Independent 1.78 3.13 2.69 2.73 2.50
Column Means 2.08 1.43 3.04 3.38 Column Means 1.65 2.95 2.71 2.84
Classification
Means 1.76 3.23
Classification
Means 2.30 2.78
a
N = 23 responses per cell
b
Column means are for each vertical column
c . .
Classification means are for four cells
and were therefore computed and presented. These marginals
i
represent the average of the four cell means for the field-
dependent stimulus children and the four cell means for the
i
field-independent stimulus children. !
Of some interest is the numerical similarity between
the respondent populations for a given cell pair. It should
be noted that all of these means should be regarded as devi
ations from the mathematically fixed grand mean of 2.50.
Upon examination of the marginal means for each of
the stimulus children over both respondent groups, that is,
the column mean of 46 ranking observations, for each of the
six ranking response categories, some index of general pref
erence toward stimulus students may be obtained. This index
is reflected in Table 7.
i
I
One stimulus child, the field-dependent female, was
ranked highest on three of the six response categories and
one child, the field-independent female, was not ranked
highest on any of these response categories. This same
field-independent child received four of the six lowest
rankings. Collapsing Table 7 over the two high and two low
categories reveals that the field-dependent children ac
counted for nine of .the twelve highest rankings.
72
Table 7
Frequency Comparison of Response Magnitudes
for All Six Response Categories Combined
Second Next
Stimulus Children Highest Highest Lowest Lowest
Field-Dependent
Male
1 4 2 2
Field-Dependent
3 I
0 0
Female
J.
FieId-Independent
Male
2 1 2 0
FieId-Independent
0 0
o
4
Female
CHAPTER V
DISCUSSION !
The major research question stated at the conclusion!
of Chapter III was to determine whether the cognitive styles|
i
,of teacher and student interact in such a way as to produce |
favorable teacher expectation effects when these styles are
i
similar and/less/or unfavorable expectation effects when they
are dissimilar. Subordinate research hypotheses developed
from this question were whether these bias effects would ap
pear in ranking for (1) on-task performance assessment, (2)
overall intelligence estimation, (3) predicted performance, :
and (4) general preference for children with similar styles.
These research questions were operationalized using a series
,of six ranking operations.
The results failed to support research hypothesis 1
for on-task performance ranking, specifically, that subjects
classified as field dependent or field independent would
tend to attribute superior performance to those stimulus stu
dents who were similar to their own cognitive styles. The
prediction was that regardless of superior performance by
73
stimulus children of the dissimilar cognitive style, sub
jects would "cross over" and rank performance of the similar
cognitive style children higher. The experimental design
having constructed a teaching (stimulus) situation in which
children of the field dependent orientation excelled with
the socially derived subject matter {Crime and Punishment)
and those of the field independent orientation excelled with
the factual subject matter (Aerodynamics) apparently pro
vided a strong impetus for subjects to stay on the same side
of the line and rank what they saw objectively. Performance
was clearly ranked in most cases as it actually was by both
respondent groups. Both of the field-dependent children
received a greater number of higher rankings for their
socially derived lesson performance, Response 1, and both of
the field-independent children received a high number of
higher rankings for the factual material, Response 2.
Examination of the cell means for the two on-task
performance rankings, that is, the mean of the 23 rankings
attributed to each of the stimulus children by each of the
respondent groups, also revealed close similarity between
the respondents. If the style bias was operating, it would
have been apparent in the magnitudes of these means such
that diametrically opposite cell means for stimulus children
of similar cognitive styles would be either high or low;
this representative of the two inherent sides of any bias,
for and against.
It seems fair to assume that either the predicted
cognitive style bias does not extend to on-task performance
ranking or that the implicit "testing" nature of the experi
mental situation influenced the respondents to rank objec
tively and not on the basis of other factors,
j The hypothesis for the intelligence ranking, re
search hypothesis 2, were also not significant although it
was predicted that this category would have provided an op
portunity for respondents to show a cognitive style pref
erence without the concomitant objectivity problems associ
ated with the on-task performance ranking category. That is
in this ranking, subjects were free to single out any of the
stimulus children's observed characteristics as indicative
of superior intelligence and rank the children accordingly.
Failure to support research hypothesis 2, that of
cognitive style bias in the intelligence ranking, was of
considerable importance for several reasons. First, it was
not performance bound as were Responses 1, 2, 4, and 5. The
videotape exposure, while brief by classroom standards,
offered a variety of stimuli, any of which could have been
76
subjectively regarded as clear-cut evidence of intelligence j
superiority. The observation of nonsignificant results
points toward a rejection of an important cognitive style
bias premise.
The results also failed to provide any evidence of a
bias in expected performance ranking, research hypothesis 3.■
• |
Here, as in the case of on-task performance, both respondent
groups remained close in their judgments: that the field- !
dependent stimulus children would excel in the social science
areas and that the field-independent stimulus children would
excell in math/science.
i
For Response 4, math/science predicted performance,
respondents tended to consistently rank the field-indepen
dent children higher than the field-dependent children.
Examination of the cell means and marginals also
reflected a consistency between the respondent groups.
Field independents individually and taken as a group were
ranked higher in math/science predicted performance and
field-dependent children were ranked higher for the social
science predicted performance.
In general, for predicted performance, research
hypothesis 3, the same patterns prevailed as did for on-task
performance. Accuracy and objectivity again appeared to be
77
the predominant motivation underlying the rankings. The
40-minute-long stimulus situation apparently provided little
justification for deviating from what was present in terms
of overt behaviors in the videotape. Predicted rankings
aligned with those of the on-task ratings. Point for point,j
the rankings for Aerodynamics and math/science prediction
. i
and those for Crime and Punishment and social science pre
dictions revealed no difference between the respondent pop- ,
ulations. I
Final research hypothesis 4 was not supported. As
in the case of intelligence estimation, this hypothesis pre
dicted a cognitive style bias in the overall preference
ranking of the stimulus children. As in research hypothesis
2, the rankings were not bound in any direct way to overall
performance. Respondents were free and encouraged' to show
overall preference toward the stimulus children without
I
tying this preference to any objective element of the video
taped performance. Since the subjects were teacher education
students, aware of classroom behavior dynamics and problems,
any specific behavior or combination of behaviors would have
been singled out and ranked highly. If, for example, or
derly classroom activity was high in one's preference sys
tem, a prospective teacher would have chosen (i.e., ranked
78 I
!
higher) those children of a reserved and quiet demeanor who !
were otherwise quite intelligent. The research hypothesis j
predicted such a bias among the field-independent respon
dents and the opposite bias among the field-dependent re
spondents . The results did not support this contention. '
Examination of the cell means and related marginal j
(column and classification) means revealed little or no dif-1
ference in the ranking tendencies of the respondent groups.
Table 7 in Chapter IV provided a tabulation of
rankings for all six responses combined as they were at
tributed to each of the stimulus children. While no other
statistical manipulations are justified, it is apparent that
for all questions taken together, irrespective of respondent1
i
group, more highest rankings are accumulated by the field- |
dependent children. Considering that for six responses
there are only twelve possible higher rankings and twelve
i
lower ones, the field-dependent children accumulated nine
of these higher twelve. Looking at the reverse, eight of
the twelve lowest rankings went to the field-independent i
stimulus children, with one child alone— the field-
independent female— accounting for six of the eight. This
may be a point for further experimentation on those stimulus
qualities which are projected in the videotape medium.
79 |
I
The general hypothesis that teacher (student teach- j
ers in the present study) would be influenced more by the
i
similarity of the stimulus child's cognitive style than by
other indices of performance was not sustained. No differ
ences between respondent groups were found to be signifi- ;
cant. As astute observers, the subjects were able to over-
I
come any affective preferences while observing either on- I
i
task performance or predicting future performance. In these
.rankings, subjects were clearly motivated by the need for
I
accuracy and objectivity, much as if they were themselves
,being tested for alertness. I
' ■ i
i
1 Consideration must also be given to the fact that
the respondents were part of a classroom situation where
performance evaluation is always implicit (Campbell & Stan- ,
ley, 1966) . This would in part explain the consistent re
sults between the respondents for on-task and predicted per-
i
formance rankings. It would be stretching this point, how
ever, if the same explanation were applied to the intelli
gence and preference rankings.
The remaining point of discussion must be whether or
not this experiment supported Witkin's (197 3) premise on the
influence of cognitive style similarity on the educational
process. The experiment attempted to determine whether such
80 J
style similarity between teacher and student could influence.
\
a teacher's judgments on several important measurements. j
i
While the evidence is not supportive, two factors may have !
\
contributed to this failure: the strength of the treatments1
and the passive characteristic of the measurement task. To
overcome the treatment strength in each of the ranking re
sponses in which children of the dissimilar cognitive style (
excelled would have required a denial of the superiority of
|their performance. In general, subjects ranked performance
objectively. The experiment was passive; the videotape of
fered no chance for interaction with the stimulus children.
Classroom interaction may be a necessary prerequisite for
the cognitive style bias to occur. These will be elaborated
I
.in Chapter VI.
i
CHAPTER VI
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ;
Summary
Review of the literature in educational psychology 1
revealed a large number of unresolved issues surrounding thej
expectation phenomena, one of the most intractable of which
I
has been the selectivity of its effects. Recent research by
Witkin and his associates suggested that cognitive style
similarity between teacher and student could be an important
variable. The present experiment set out to determine
whether cognitive style similarity between teacher and stu- ~
dent measured along the field-dependent/field-independent
continuum could bias a teacher's objectivity in a variety of
itypical teaching situations. A stimulus in the form of a
videotape was developed to elicit responses from a group of
teacher education students classified along the field-
dependent/field-independent dimension. The prediction that
these students would show a bias toward stimulus students of
like cognitive styles was tested for four dependent vari
ables: on-task performance ranking, intelligence ranking,
81
82 |
I
I
performance prediction, and general preference. j
Conclusions i
Results of the experiment, analyzed through non- j
i
parametric techniques, revealed no significant differences
between subjects classified and field dependent or field
independent for any of the research hypotheses.
I
Results of the experiment may be interpreted in
several ways. The situation in which the videotapes were
presented was a classroom situation in which evaluation of
performance is a predominant concern. Given the strength of
the treatment, that is, evidence of superior performance by
children of differing cognitive styles, showing a preference
for children of the same cognitive style would be tantamount1
to ignoring objective evidence of superiority. Hence, the
strong consistency of observations between the respondents
of both cognitive style classifications may have been more
due to the implicit testing nature of the stimulus situation.
Even so, when the ranking task was separated from performance
per se, as in the overall intelligence and overall preference
ranking, no significant differences were observed.
The question of attenuation of characteristics in
the videotape medium must also be addressed. In a perfor
mance based situation, where a relatively rigid teaching
83 |
i
format is employed, the stimulus value of a number of impor
tant cognitive style characteristics may be reduced over all,
leaving only a relatively small number of stimuli with whichj
subjects may identify. If sufficiently small, the ambiguity^
which then accompanies the stimulus may result in the types
of rankings observed— with few exceptions, inconclusive. j
i
i
Building characteristics into a videotape, in short, is no
guarantee that these characteristics will be projected with
equal salience. Performance may dominate as it did in this 1
case.
In the same sense, the task of ranking was passive
i
with no possibility for interaction. No words or concepts
were taught by the subjects and no classroom dialogue was
present between subject and student. This is another way in
which the results may be interpreted. Students taught by
i
teachers of similar cognitive styles performed better, for
i
example. Interaction with students by teachers may be a
prerequisite for observation of a cognitive style bias.
Finally, the classification of field independent-
dependent based upon :the GEFT alone may result in the gen
eration of two subject pools who do not differ on anything
more than the ability to solve a perceptual problem.
I
I
Recommendations
Recommendations for further definitive study are
advanced herein. To test the influence of experience and
test sensitivity, the same videotapes could be administered
to student teachers in an evaluation-free environment and
aslo to in-service teachers under the same nonjudgmental
I
circumstances.
i
Another recommendation is to provide more stimuli 1
for a given unit of time through the use of several addi
tional students without structuring a lesson. Related to
this would be the construction of a videotape with only
I
field-dependent of field^independent stimulus children and
administering these to similar respondent groups. Rather
i
than ranking data, rating (interval scale) data could be i
taken with orthogonal comparisons made between the video
tape presenting field-dependent and field-independent
children.
Finally, similar experimental situations could be
conceived, capitalizing on the infinite reusability of video
tape with many matched or varying subject populations. In
these experiments, it would be important to minimize the test
sensitivity aspect in that teacher education students, like
others, are aware of the fact that judgments are being made
85
I
on their measurements; consequently, in striving for objec
tivity and precision, they may tend to minimize things such
as likes and dislikes and stress observables. These stud-
lies could also utilize * larger stimulus student groups, per
haps in a naturalized setting, which would provide increased
measurement opportunities.
I
I
J
APPENDIXES
86
APPENDIX A
GEFT SCORE DISTRIBUTION
87
APPENDIX A
GEFT SCORE DISTRIBUTION
Scores eliminated
/
SUBJECT GEFT SCORE
APPENDIX B
SAMPLE PAGE OF GEF TEST
89
APPENDIX B
SAMPLE PAGE OF GEF TEST
FIGURES TEST
EMBEDDED
GROUP
By Philip K. Oltman, Evelyn Raskin, & Herman A. Witkin
Name
Sex
Today’s date Birth date
INSTRUCTIONS: This is a test of your ability to find a simple form when
it is hidden within a complex pattern.
Here is a simple form which we have labeled “X ” :
This simple form, named “X” , is hidden within the more complex figure
below:
Try to find the simple form in the complex figure and trace it in pencil
directly over the lines of the complex figure. It is the SAME SIZE, in the
SAME PROPORTIONS, and FACES IN THE SAME DIRECTION within the
complex figure as when it appeared alone.
When you finish, turn the page to check your solution.
Copyright, 1971, Consulting Psycholgists Press, Inc.
X
90
APPENDIX C
CRITERIA FOR SELECTION OF CHILDREN
91
APPENDIX C
CRITERIA FOR SELECTION OF CHILDREN
Field Independent
Karen McWilliams
High IQ
Physically inactive
Average stature, -12 yrs.
Very quiet
Works independently
Task oriented, hard to
distract
Seeks nonsocial rewards
Factual subject matter
orientation
GEFT Score 19
Field Dependent
Shari Spiegler
High IQ
Physically active
Small stature, -12 yrs.
Very social/gregarious
Seeks social rewards
Openly expresses positive
feelings
Assists others
GEFT Score 9
Greg Welker
High IQ
Physically active
Average stature, -12+ yrs.
Quiet but friendly
Works independently
Task oriented
Partly responsive to
social rewards
Prefers discovery method
GEFT Score 18
Mitch Spindell
High IQ
Physically very active
Large stature, -12 yrs.
Very social/gregarious
Responsive to social
rewards
Openly expresses positive
feelings
Competitive
Assists others
GEFT Score 9
(12-minute administration of GEFT to all subjects)
92
APPENDIX D
VIDEOTAPE OBJECTIVES AND INSTRUCTIONS
TO THE TEACHER
93
APPENDIX D
VIDEOTAPE OBJECTIVES AND INSTRUCTIONS
TO THE TEACHER 1
The thesis is that teachers of a particular cognitive style,'
specifically field dependent or field independent, will tend'
to show a measurable bias or preference toward students of
similar cognitive styles. By measurable is meant some
quantifiable indices measured in terms of the teachers'
,rating of on-task performance, predicted performance, appa
rent IQ, and simply preference for students of similar cog
nitive styles.
i
Cognitive style as defined as ones characteristic or global
method of perceiving the world around him; field dependent
characterized as a definite sensitivity to context or imbed-
dedness of a perceptual element (that is, the relation of
the item perceived to all background related and situation
related entities) and field independent characterized by an
extremely analytical way of perceiving the world with judg- ;
ments based upon internal strategies and situation-indepen
dent elements. When a field-dependent person is confronted
with a problem or when he perceives a situation, his think
ing is mediated by almost everything in his perceptual field
taken as a whole; not as separate elements. When a field-
independent person is confronted with the same situation,
many of the elements may be "removed" from the context and
evaluated or analyzed separately. In the same way, solu
tions to various problems are typically internally formu
lated rather than being driven or defined by the situation. 1
By presenting a videotape which depicts students of these
two cognitive styles to teachers of similar or dissimilar
cognitive styles it is expected that various biases will
appear which may be measured and analyzed; these biases
toward like styles or matched styles.
94
• ! ' : 95 [
?
In order to measure this bias, a stimulus situation was ;
designed in which students of known cognitive styles would !
be depicted in 2 microteaching sequences. These sequences
were designed to enable students of the two cognitive style
categories to interact optimally with two different subject
matter areas; one factual and one social. In this way the !
students all show up well with some subject matter but, if 1
the hypothesis of bias-effect holds true, it is expected
that the teachers of a particular cognitive style will tend :
to favor overall students of that same style. I
\
The two microteaching sequences will be recorded on video
tape and ultimately shown to a large number of teachers who
will themselves be classifed (using a classification test
instrument) as field dependent or field independent.
The objectives of the first videotape, Aerodynamics, are to 1
teach the students elementary flight physics; the nature of
forces acting on an aircraft, how it propels itself and how
it stays aloft. The second videotape, Crime and Punishment,
is also an interactive lesson and discusses the reasons for
punishment, the preventative elements of imprisonment or
other punishment and the concept of rehabilitation. In the
Aerodynamics videotape it is expected that the factually
oriented field-independent students will be more salient or
will come across as brighter, whereas the field-dependent 1
students will come across more effectively in the Crime and
Punishment videotape. Notwithstanding this, it is assumed '
that teacher-subjects will cross these lines of subject mas
tery and rank the students closer to their own cognitive i
styles higher than others.
APPENDIX E
CONCEPT AND VIDEOTAPE WORD LIST
96
APPENDIX E
KEY WORDS
Lift
Thrust
Drag
Gravity
CONCEPT AND VIDEOTAPE WORD LIST
Aerodynamics and Flight
OTHER IMPORTANT WORDS
Friction
Overcome
Power
Air
Gas
Action-Reaction
BACK
TOP
AIR
FRONT BOTTOM
WING SHAPE
97
98l
Crime and Punishment
KEY WORDS
Prevent
Protect
Punish
Rehabilitate
Treatment
Freedom
OTHER IMPORTANT WORDS
Prison
Fines
Law
Rules
Stealing
Welfare
LIMIT
M.P.H
PRISON
IIIIII
APPENDIX F
EFFECTIVENESS OF VIDEOTAPE
STIMULUS CHARACTERISTICS
99
APPENDIX F
EFFECTIVENESS OF VIDEOTAPE
STIMULUS CHARACTERISTICS
Average Ratings of Eight Raters
F A C T U A L O R I E N T A T I O N - T h e d e g r e e M i t c h
t o w h i c h t h e s t u d e n t a p p e a r s t o S h a r i
b e s t i m u l a t e d b y f a c t u a l s u b j e c t G r e g g
m a t t e r a n d t h e m o r e c o n c r e t e K a r e n
i s s u e s
L o w 1
1
1
1
1
5 High
5
5
5
5
I N D E P E N D E N T T H I N K I N G - T h e d e g r e e G r e g g 1
t o w h i c h t h e s t u d e n t s h o w s t h i n k - K a r e n 1
i n g s t y l e s w h i c h a r e i n d e p e n d e n t S h a r i 1
o f g r o u p s n o r m s a n d c o n s e n s u s M i t c h 1
© 5
A ^ 4 5
(2) 3 4 5
2 ( 3 ) 4 5
d )
S O C I A L O R I E N T A T I O N - T h e d e g r e e S h a r i
t o w h i c h t h e s t u d e n t a p p e a r s t o ' ■ G r e g g
b e s e n s i t i v e t o t h e s o c i a l t h e m e M i t c h
i n t h e d i s c u s s i o n s K a r e n
S E X R O L E D E P E N D E N C Y - T h e d e g r e e S h a r i
t o w h i c h t h e s t u d e n t s a p p e a r t o M i t c h
b e i n f l u e n c e d b y t h e i r t r a d i t i o n a l G r e g g
s e x r o l e s i n t h e i r t h i n k i n g a n d K a r e n
r e s p o n d i n g
E X P R E S S I O N O F P O S I T I V E F E E L I N G -
E x p r e s s e s p o s i t i v e f e e l i n g s i n
d e a l i n g w i t h o t h e r s
P E E R G R O U P O R I E N T A T I O N - T h e d e g r e e
t o w h i c h t h e s t u d e n t a p p e a r s t o
b e i n f l u e n c e d b y t h e a c t i o n s o f
t h e p e e r g r o u p
G R E G A R I O U S N E S S - T e n d e n c y t o
g r a v i t a t e s o c i a l l y t o w a r d t h e
o t h e r s t u d e n t s
M i t c h
S h a r i
K a r e n
G r e g g
G r e g g
S h a r i
K a r e n
M i t c h
K a r e n
M i t c h
G r e g g
S h a r i
1 2 3
1 2 3
© 2 3
V © 3
©
i 9
2 3 4 5
2 3 (4) 5
3 4 5
3 © 5
3 4 5
I 9 I
3 © 5
R E T I C E N C E - T h e d e g r e e o f
s o c i a l h e s i t a n c y s h o w n b y
t h e s t u d e n t
M i t c h
G r e g g
K a r e n
S h a r i
1 ( 2) 3 4 5
1 2 ( 3) 4 5
1 2 V © 5
! 0 3 4 5
100
R I G I D I T Y - T h e l a c k o f
f l e x i b i l i t y i n t h i n k i n g
a n d o t h e r a p p a r e n t b e
h a v i o r
S E E K S S O C I A L R E W A R D S - S e e k s a n d
r e s p o n d s t o t e a c h e r s s o c i a l r e
w a r d s
T A S K O R I E N T A T I O N - S t a y s w i t h t h e
t a s k a t a l l t i m e s . N o t e a s i l y
d i s t r a c t e d
S h a r i
K a r e n
M i t c h
G r e g g
G r e g g
K a r e n
S h a r i
M i t c h .
S h a r i
M i t c h
K a r e n
G r e g g
APPENDIX G
VIDEOTAPE RESPONSE RECORD
102
APPENDIX G
VIDEOTAPE RESPONSE RECORD
NAME _________________________________________! _______ AGE /________
TEACHING SPECIALTY___________________________________ ACADEMIC YEAR
Many teachers, are required to make important judgments based on very limited
exposure and information about children. You are about to see- two videotapes or
microteaching sequences after each of which you will be asked to make some impor
tant judgments. We would like you to assess the strengths and weaknesses of the
children you will see in the videotapes. Of special interest is your evaluation
of performance, your prediction of future performance, your estimate of IQ, and
finally your preference or your liking for the children you saw in the video
tapes. Thank you again for your help.
Rank order the children on their understanding of the crime and punishment
subject._________________________________________________1______________
2 _________
3 _________
4 ______________
Rank order the children on their understanding of Aerodynamics.
*1_._________
2_____________
3 __________________ .
4 ______________
Rank order the children on overall intelligence.
1__________
2_____________
3 ______________
4 ______________
Rank order the children on their future performance in Math/Science.
1__________
2 _________
3 _________ . .
. 4______________
Rank order the children on their future performance in Social Science.
1__________
2_____________
3______________
4 ______________
Rank very honestly your overall liking for the children you saw.
1 _______
2 _________
3______________
4______________
By Ranking we mean list in order. Even in the event that two or more children
are "tied" for Number 1 or 3, please rank them anyway.
103
APPENDIX H
CONTINGENCY TABLES FOR RANKING RESPONSES
104
105
APPENDIX H
CONTINGENCY TABLES3 FOR RANKING RESPONSES*5
RESPONSE NUMBER 1— GN-TASK PERFORMANCE: CRIME AND PUNISHMENT
Field-D ependent Student (M ale) Field-Dependent Student (Fem ale)
Respondents
Ranked as
1st o r 2nd
Ranked as-
3rd o r 4th
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
F i eld -In d ep endent
Student-Teachers 22 19
Field-Dependent
Student-T eachers 23
45
20
33
Chi square— not computed Chi square— not computed0
Field-Independent Student (M ale) Field-Independent Student (F em ale)
R espondents
Ranked as
1st. o r 2nd
Ranked as
3rd o r 4th
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
F ie l c l -In d ep end ent
Student-Teachers 20 21
Field-D ependent
Student-T eachers 2 1
41
22
43
Chi square--not computed0 Chi square— not computed0
a
Degrees of freedom =1
^Column totals all sum to N - 23
°Cell (expected)frequencies inadequate for chi-square computation
CONTINGENCY TABLES FOR RANKING RESPONSES
RESPONSE NUMBER 2— ON-TASK PERFORMANCE: AERODYNAMICS
Pield-Deoendent Student (M ale)
Respondents
F i eld-Ind ep end ent
Student ~T eachers
Field-Dependent
Studeni-T eachers
Ranked as
1st o r 2nd
11
Ranked as
3rd o r 4th
12
12_
23
1L
23
Chi square = .034 p > .10°
Field-Independent Student (M ale)
Respondents
Fi eld-Ind ependent
Student-T eachers
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
Field-Dependent Student (Fem ale)
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
O l
k . 1
X
4
II
42
Chi square— not computed^
F I eld-Ind eo end ent Student (F em ale)
Rankea as
1st o r 2nd
13
Ranked as
3rd o r 4th
10
Field-Dependent
Student-T eachers
M
40
Chi square— not compared^
Degrees of freedom = i
Column totals all sum to N = 23
Includes Yates' correction
12.
C ' R .
11
21
Chi square =0 p > .10
Cell (expected) frequencies inadequate for chi-square computation
CONTINGENCY TABLES3 FOR RANKING KESP0NS2Sb
RESFONSE NUMBER 3— INTELLIGENCE ESTIMATE
Field-Dependent Student (M ale) Field-Denendent Student (Fem ale)
Respondents
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
Field-Independent
Student-T eachers 16
Field-Dependent
Student-T eachers X4
23
9
23
II
33
_6
13
Chi square =1.36 p > .10
Chi square =0 p > .10
Field-Independent Student (M ale)
ispondents
Ranked as
1st. o r 2nd
Ranked as
3 rd o r 4th
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
Field-Independent
Student-T eachers 12 11 14
Field-Dependent
Student-T eachers 10.
22
13
24
_5
14
18
32
Chi square = .08 p > Chi square = .82 p > .10
Degrees of freedom = 1
Column totals all sum to N = 23
Includes Yates' correction
107
a b
C O N T IN G E N C Y T A B L E S F O R R A N K IN G R E S P O N S E S
R E S P O N S E N U M B E R 4 — P R E D IC T E D P E R F O R M A N C E : M A TH A N D S C IE N C E
F ield-Dependent Student (M ale)
Field-Independent
Student-T eachers
Field-Dependent
Student-T eachers
Ranked as
1st o r 2nd
7 •
17
Chi scruare
Ranked as
3rd o r 4th
16
13
29
,37 p > .10C
Field-D ependent Student (Fem ale)
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
15
11
19
11
27
Chi square = .35 p >..10
Field-Independent Student (M ale)
Respondents
Field-Independent
Student-T eachers
Field-Dependent
Student-Teachers
Ranked as
1st o r 2nd
22
19
41
Ranked as
3rd o r 4th
Jt
5
Chi square— not computed
Field-Independent Student (Fem ale)
Ranked as
1st o r 2nd
_1
15
Ranked as
3rd o r 4th
14
17
31
Chi square - .39 p > .10
’ ’ ’ Degrees of freedom * = 1
$
Column totals all sum to N = 23
^Includes Yates' correction
i
Cell (expected) frequencies inadequate for chi square computation
CONTINGENCY TABLES FOR RANKING RESPONSES
RESPONSE NUMBER 5--PREDICTED PERFORMANCE: SOCIAL STUDIES
Fieid-Dependent Student (M ale)
Respondents
F ield-Independer.t
Student-T eachers
Field-Dependent
S tadent-T eachers
Ranked as
1st o r 2nd
16
Ranked as
3rd o r 4th
18.
34
5
12
Chi square = .11 p > .10'
Fie Id-Dependent Student (Fem ale)
Ranked as
1st o r 2nd
2 0
Ranked as
3 x'd o r 4th
20
40
Chi square—-not computed'
*d
Field-Independent Student (M ale)
Respondents
Field-Independent
Student-Teache rs
. Field-Dependent
Student-T eachers
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
19
2 S L
39
Chi square— not computed
Degrees of freedom - 1
Column totals all sura to N = 23
Includes Yates1 correction
Field-Independent Student (Fem ale)
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
11
18
35
Chi square =0 p > .10
Cell (expected) frequencies inadequate for chi square computation
CONTINGENCY TABLES FOR RANKING RESPONSES
RESPONSE NUMBER 6— OVERALL PREFERENCE
Field-Dependent Student (M ale)
Respondents
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
F i eld -Ind ep end ent
Student-Teachers 14
Field-D ependent
Student-T eachers
II
33
4
15
Chi square = 1.73. p > .10
Field-Denendent Student (Fem ale)
Ranked as
1st o r 2nd
Ranked as
3rd o r 4th
17
14
15_
32
Chi square = .10 p > .10
Field-Independent Student (M ale)
Respondents
Field-Independent
Student-T eachers
Ranked as
1st o r 2nd
l ' c
Ranked as
3rd o r 4th
1 0
Field-D ependent
Student-T eachers _8_
2 1
15.
25
Chi square =1.38 p > .10
'Degrees of freedom = . 1
Field-Independent Student (Fem ale)
Ranked as
1st o r 2nd
13
JL
20
Ranked as
3 rd o r 4th
10
16
26
Chi square =2.21 p > .10C
Column totals all sum to N = 23
Includes Yates1 correction
110
I
i
REFERENCES
111
REFERENCES
Amidon, E., & Flanders, N. The effects of direct and indi
rect teacher influence on dependent-prone students
learning geometry. Journal of Educational Psy
chology, 1961, 52, 286-291.
Baird, L. Teaching styles: An exploratory study of dimen
sions and effects. Journal of Educational Psy
chology, 1973, 64, 15-21.
Beez, W. Influence of biased psychological reports on
teacher behavior and pupil performance. Proceedings
of the 76th Annual Convention of the American Psy
chological Association, 1968, 3, 605-606.
Brophy, J., & Good, T. Communication of different expecta
tion for children's classroom performance: Some
behavioral data. Journal of Education Psychology ,
1970, 61, 365-374.
Brophy, J., & Good, T. Teacher-student relationships:
Causes and consequences. New York: Holt, Rinehart
& Winston, 1974.
Brown, W. The influence of student information on the for
mulation of teacher expectancy (Doctoral disserta
tion, Indiana University, 19^9). Dissertation
Abstracts International, 1970, 30, 4822-A.
Burstall, C. French from eight: A national experiment.
Slough, Great Britain: NFER, 1968.
Campbell, D., & Stanley, J. Experimental and quasi-
experimental designs. Chicago: Rand McNally, 1963.
Castaneda, A., & Gray, T. Bicognitive processes in multi
cultural education. Educational Leadership, 1974,
32(3), 203-207.
112
113 |
i
Claiborn, W. Expectancy effects in the classroom: A fail- ;
ure to replicate. Journal of Educational Psychology,*
1969, 60, 377-383.
Dalton, W. The relations between classroom interaction and j
teacher reatings of pupils: An explanation of one j
means by which a teacher may communicate her ;
expectancies. Peabody Papers In Human Development,
1969, 7,6.
DiStephano, J. Interpersonal perceptions of field dependent
and field Independentlteachers and students. Unpub-j
lished doctoral dissertation, Cornell University, 1
1969.
Doob, L. W. Behavior and grammatical style. Journal of
Abnormal and Social Psychology, 1958, 56, 398-400.
Feitler, F., Weiner, W., & Blumberg, A. The relationship
between Interpersonal relations and preferred class-,
room settings . Paper presented at the annual meet
ing of the American Educational Research Association,
1970.
Feshbach, N. Student teacher preferences for elementary
school pupils varying in personality characteristics!
Journal of Educational Psy chology, 1969, 60, 126- j
132.
Fielder, W., Cohen, R., & Feeney, S. An attempt to repli
cate the teacher expectancy effect. Psychologlcal
Reports, 1971, 29, 1223-1228.
Flanders, N. Analysing teacher behavior. Reading, Massa
chusetts: Addison-Wesley, 1970.
Fleming, E., & Anttonen, R. Teacher expectancy or My Fair
Lady. American Education Research Journal, 1971, 8,
241-252.
Goldenberg, I. Social class differences in teacher atti
tudes toward children. Child Development, 1971, 42,
1637-1640.
Good, T
Good, T
Greene,
Heil, L
Hunt, D
Hunt, D
Jackson
James,
i
James,
., & Brophy, J. Teacher-child dyadic interactions: j
A new method of classroom interactions. Journal of
School Psychology, 1970, 8, 131-138.
., & Brophy, J. Behavioral expression of teacher
attitudes.. Journal of Educational Psychology, 1972,:
63, 617-624. j
M.' Client perception of the relationship as a func
tion of worker-client cognitive styles. . Unpublished
doctoral dissertation, Columbia University, 197 2. i
i
i
., Powell, M., & Feifer, I. Characteristics of
teacher behavior related to the achievement of chil
dren in several elementary grades. Washington,
D.C.: Department of Health, Education and Welfare,
Office of Education, Cooperative Research Branch,
1960.
Matching models in education: The coordination of
teaching methods with student characteristics .
Toronto: Ontario Institute for Studies in Educa
tion, Monograph Series No. 10, 1971.
., & Sullivan, E. Between psychology and education.
Hinsdale, Illinois: Dryden Press, 1974.
I
, P., & Lahaderne, H. Inequalities of teacher-pupil 1
contacts. Psychology in the Schools, 1967, 4, 204-
211.
2. A cognitive style approach to teacher-pupil inter
action and the academic performance of black chil
dren. Unpublished master's thesis, Rutgers Uni
versity, 197 3.
t f . Personal preferences for method as a factor in 1
learning. Journal of Educational Psychology, 1962,
53, 43-47.
Jenkins, B. Teachers' views of particular students and
their behavior in the classroom. Unpublished doc
toral dissertation, University of Chicago, 1972.
Joyce, C. , & Hudson, L. Student style and teacher style:
An experimental study. British Journal of Medical
Education, 1968, 2, 28-32.
Justice, M. Field dependency ^ intimacy of topic and inter
person distance. Unpublished doctoral dissertation,
University of Florida, 1969.
Kelley, H. The warm-cold variable in first impressions of
people. Journal of Personality, 1950, 18, 431—439.
Kester, S., & Letchworth, G. Communication of teacher ex
pectations and their effects on achievement and
attitudes of secondary school children. The Journal
of Educational Research, 1972, 66, 51-55.
Kleinfeld, J. Instructional style and the intellectual per
formance of Indian and Eskimo students . (Final
report, Department of Health, Education and Welfare,
Office of Education, Project No. l-J-027). Washing
ton, D.C.: U.S. Government Printing Office, 1972).
I
Kranz, P. & Tyo, A. Do teachers’ perception of pupils
affect their behavior toward those pupils? Unpub
lished manuscript, Millersville State College, 1973.
Lange, C. A study of the effects on learning of matching
cognitive styles of students and instructors in
nursing education. Dissertation Abstracts Inter
national, 1972, 33(9), Section A.
Lindzey, G., & Byrne, D. Measurement of social choice and
interpersonal attractiveness. In G. Lindzey & E.
Aronson (Eds.), Handbook of social psychology (2nd
ed.) (Vol. 2). Reading, Massachusetts: Addison-
Wesley, 1968.
Luborsky, L. Individual differences in cognitive style as a
. determinant of vasoconstrictive orienting responses.
Paper presented at the meeting of Transactions of an
International Colloquium, Bratislava and Smolenice,
September 1965.
Mackler, B. Grouping in the ghetto. Education and Urban
Society, 1969, 2, 80-96.
, nei
Marlow, D., & Gergen, K. Personality and social interac
tion. In G. Lindzey & E. Aronson (Eds.), Handbook
of social psychology (2nd ed.) (Vol. 3). Reading,
Massachusetts: Addison-Wesley, 1969.
McNemar, Q. Psychological statistics (2nd ed.). New York:
Wiley & Co., 1955. .
Merton, R. The self-fulfilling prophecy. Antioch Review,
1948, 8, 193-210.
I
Middleman, R. An experimental field study of the impact of j
nonverbal communication of affect on children from
two socio-economic backgrounds. Paper presented at
the annual meeting of the American Educational
Research Association, 1972. '
'Ohnmacht, F. Teacher characteristics and their relation
ships to some cognitive styles. Journal of Educa
tional Research, 1967, 60, 201-204.
Ostfeld, B., & Katz, P. The effect of threat severity on
children of varying SES levels. Developmental
Psychology, 1969, 1, 205-210.
Panda, K., & Guskin, S. Effect ofsocial reinforcement*
locus of control* and cognitive style on concept of i
learning among retarded children. Annual report, I
Center for Educational Research and Development for
Handicapped Children, Indiana University, 1970.
(Monograph)
'Phillips, M., & Sinclair, R. Conceptual systems and educa
tional environments: Relationships between teacher
conceptual sy stems * student conceptual Systems * and
classroom environment as perceived by fifth- and
sixth-grade students. Paper presented at the annual
meeting of the American Educational Research Asso
ciation, Chicago, 1973.
Rist, R. Student social class and teacher expectations.
Harvard Educational Review, 1970, 40, 411-451.
Rosenthal, R., & Jacobson, L. Pygmalion in the classroom:
Teacher expectation and pupils ' intellectual devel
opment. New York: Holt, Rinehart & Winston, 1968.
117
Rosenthal, R. , & Rubin, D. Appendix C: Pygmalion reaf
firmed. In J. Elashoff & R. Snow (Eds.), Pygmalion
reconsidered. Belmont, California: Wadsworth Pub
lishing Co., 1971.
Rubin, L. A study on teaching style. Paper presented at
the annual meeting of the American Educational
Research Association, Chicago, 1971.
Rubovits, P., & Maehr, M. Pygmalion analyzed: Toward an
explanation of the Rosenthal-Jacobson findings.
Journal of Personality and Social Psychology, 1971, j
19, 197-203.
Sears, P., & Hilgard, E. The teacher's role in the motiva
tion of the learner. In Theories of learning and
( instruction. Sixty-third Yearbook of the National
Society for the Study of Education, Part 1, 1964.
Seaver, W. Effects of naturally induced expectancies on the
performances of pupils in primary grades. Unpub
lished doctoral dissertation, University of Illinois,
1971.
Seigel, S. Nonparametric methods for the behavioral
sciences. New York: McGraw-Hill, 1956.
Silberman, M. Teachers' attitudes and actions toward their
students. In M. Silberman (Ed.), The experience of
schooling. New York: Holt, Rinehart & Winston, ,
1971.
Sperry, L. Learning performance and individual differences.
Glenview, Illinois: Scott, Foresman & Co., 1972.
Thelen, H. Matching teachers and pupils. National Educa
tion Association Journal, 1967, 55(4), 18-20. j
Thorndike, R. Review of Pygmalion in the classroom. Ameri-<
can Educational Research Journal, 1968, 5, 708-711.
Tobias, S. Preference for instructional method and achieve
ment. Paper .presented at the annual meeting of the
Eastern Psychological Association, 197 2.
Tuckman, B. Study of the interactive effects of teaching
style and student personality. Proceedings of the
77th annual convention,. American Psychological
Association, 1969.
Williams, F., Whitehead, J., & Miller, L. Relations between
language attitudes and teacher expectancy. American
Educational Research Journal, 1972, 9, 263-277.
Willis, S., & Brophy, J. The origins of teachers' attitudes
towards young children. Journal of Educational
Psychology, 1974, 66(4), 520-529.
Winer, B. J. Statistical principles in experimental design.
New York: McGraw-Hill, 1962.
Witkin, H. The role of cognitive style in academic per
formance and in teacher-student relations. Educa
tional Testing Service Research Bulletin, 1973, 73
(11, Whole issue).
Witkin, H., Dyk, R., Faterson, H., Goodenough, D., & Karp,
S. Psychological differentiation studies of devel
opment. New York: Wiley, 1962. (Republished:
Potomac, Maryland: Laurence Erlbaum Associates,
1974).
Witkin, H., & Moore, C. Cognitive style and the teaching-
learning process. Paper presented at the annual
meeting of the American Educational Research Associ
ation, Chicago, April 1974.
I
Witkin, H., Oltman, P., Raskin, E., & Karp, S. Manual for
the embedded figures test. Palo Alto, California:
Consulting Psychologists Press, 1971.
Wu, J. Cognitive style and task performance. Unpublished
doctoral dissertation, University of Minnesota,
1967.
Yando, R., & Kagan, J. The effects of teacher tempo on the
child. Child Development, 1968, 39, 27-34.
119
Zussman, D. , & Pascal, C. The 'interaction of divergence and\
convergence of students and teachers with personal- !
ity and instructional variables affecting educa
tional outcomes. Paper presented at the annual ;
meeting of the American Educational Research Asso
ciation, Chicago, 1973.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The interactive effects of cognitive style and learner preparation on cognitive learning

PDF

The effects of guided daydream experiences on level of self-esteem and depression in alcoholics

PDF

The relationship of mother's teaching style on the pre-schooler's cognitive performance

PDF

The effects of verbal rehearsal and visual imagery on the performances of learning disabled and normal students

PDF

The interactive effects of global cognitive style and general mental ability with three instructional strategies on a mathematical concept task

PDF

The effects of cognitive and experiential instructional methods on attitude and behavior change

PDF

Differential effects of audiotape feedback on client self-perception

PDF

The lonely profession: A study of the psychological rewards and negative aspects of the practice of psychotherapy

PDF

The effect of seating position on performance and personality in a college classroom

PDF

The role of stress, cognitive complexity and career maturity in teacher socialization

PDF

The interactive effects of computerized social reinforcement and sociability of students during computer assisted instruction

PDF

The interactive effects of arousal level, cognitive style, and color on affective learning

PDF

The relationship of cognitive styles and academic majors with university-level academic achievement

PDF

Sex differences, spatial visualization ability and the effects of induced and imposed imagery on problem solving performance

PDF

The construct of integrative cognitive style and its measure as an indicator of advanced ego development

PDF

An investigation of the relation between student prior knowledge, perception of learning environment and teacher orientation on cognitive achievement for elementary school children

PDF

The effect of self-knowledge and career information on career maturity

PDF

The effects of meditation on general anxiety, test anxiety, and non-verbal intelligence

PDF

The differential effectiveness of three cognitive coping strategies on examination performance and test anxiety

PDF

The interrelationship of personality characteristics, levels of cognition, and academic achievement in junior college students

Asset Metadata

Creator Spindell, William Arden (author)

Core Title The effects of teacher-student cognitive style similarity on performance evaluation, performance prediction, intelligence estimation and general preference

Degree Doctor of Philosophy

Degree Program Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, educational psychology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c26-450431

Unique identifier UC11245248

Identifier usctheses-c26-450431 (legacy record id)

Legacy Identifier DP24172.pdf

Dmrecord 450431

Document Type Dissertation

Rights Spindell, William Arden

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA