COGNH‘IVE ADJUSTMENT OF PERCEWED SELF CREDIBIUTY.
PERCEIVED SOURCE CREDIBILITY, AND PERCEIVED TASK
DIFFICULTY AS A RESULT OF FEEDBACK EN TASK
GREENTED DY'ADS

Thesis for the Degree of Ph. D.
MECHIGAN STATE iU‘NWERSlTY

Roberf Owen Engbrefson
1964

THESE;

MllilHllllHlWl“WNW\IHU19l|flU|l9i|l9||LU|Hl ' LIBRAR y

3 1293 10456 Michigan State
University

 

This is to certify that the

thesis entitled

Cognitive Adjustment of Perceived Self Credibility,
Perceived Source Credibility, and Perceived Task
Difficulty as a Result of Feedback in Task
Oriented Dyads

presented by
Robert Owen Engbretson

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Communication

O>45 /Z/ (1%“

Major professor
Date 92L?“ Fl; 17‘?

0-169

 

 

 

ROCi

I '
.

“CC
U41.

MY.

     

 

ABSTRACT
COGNITIVE ADJUSTMENT OF PERCEIVED SELF CREDIBILITY,
PERCEIVED SOURCE CREDIBILITY, AND PERCEIVED TASK

DIFFICULTY AS A RESULT OF FEEDBACK IN TASK
ORIENTED DYADS

by Robert Owen Engbretson

The rationale of this study developed from the premise that in an
interaction situation involving task accomplishment, an individual has
a number of cognitive clusters which are relevant to the situation. These
cognitive clusters were conceptually defined as: Self Credibility, or
the perceptions a person has concerning his own ability to engage in the
interaction; Source Credibility, or the perceptions a person has con-
cerning the abilities of other participants to engage in the interaction;
and, Task Difficulty, or the perceptions a person has concerning the
complexity and difficulty of the task in which the person is engaged.
The rationale further stated that when an individual receives feedback
which is not consistent with these cognitive elements, he is placed in
a state of cognitive imbalance. In order to restore cognitive balance
it is necessary for the individual to make changes in the cognitive elements.
A number of alternative ways to change the cognitive elements are
available to the individual. This research identified Specific changes
occurring in these cognitive elements under varying combinations of Self
Credibility, Source Credibility and Feedback.

Eighty female undergraduates engaged singly in an interaction
situation with a male confederate. To experimently assign subjects

to one of eight treatment conditions involving high or low Self and

Robert Owen Engbretson
Source Credibility, each subject engaged in a practice period, after
which she received information concerning her competency and the com-
petency of the confederate.

Without being visible to each other, the subject and the con-
federate then engaged in a task oriented interaction. The task was to
describe the positions of a geometric puzzle in such a way that the
confederate could reproduce the puzzle. During the interaction, the
confederate responded to the communication efforts of the subject by a
prearranged positive or negative verbal feedback schedule.

Pre and post measures of Source Credibility were obtained with
semantic differential type scales taken from a recent factor analytic
study which identified empirical factors for conceptualizing and operation-
alizing Source Credibility. These scales were generalized for use in
measuring Self Credibility. Task Difficulty was measured with similar
scales.

Analysis of the data supported the experimental manipulations.
To test the hypotheses, £_tests for related samples on the differences
between the means were used. The results of the analysis for the eight
treatment conditions indicate that:

l. Across all treatment conditions, the feedback a subject

receives from a source affects her perception of Self
Credibility, the credibility of the source of the feedback,

and her perception of the difficulty of the task.

Robert Owen Engbretson

2. Self Credibility appears to be the most stable of the
variables; and, hence, least susceptible to change. Three
of the four significant changes in Self Credibility occurred
in positive feedback treatments.

3. Consistent across all treatment conditions, Source
Credibility decreased under negative feedback and increased
under positive feedback conditions.

U. The subjects' perceptions of the difficulty of the task
change under differing conditions of feedback. Under
negative feedback the task is perceived as being more
difficult, while under positive feedback the task is
perceived as being less difficult.

Within a "balance model" framework, twelve of twenty four
predictions were correct. Of twelve incorrect predictions, eleven in-
volved significant changes under conditions in which the "balance model"
would predict no change. A possible explanation was developed in which
it was proposed that the feedback served as a secondary reinforcing or
nonreinforcing stimulus to the statements used in the experimental
manipulation of the independent variables. It was suggested that had
the experimental manipulation of the independent variables been more
successful many of the significant changes which were not predicted would dis-

appear.

COGNITIVE ADJUSTMENT OF PERCEIVED SELF CREDIBILITY,
PERCEIVED SOURCE CREDIBILITY, AND PERCEIVED TASK
DIFFICULTY AS A RESULT OF FEEDBACK IN TASK

ORIENTED DYADS

BY

Robert Owen Engbretson

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Communication

196“

ACKNOWLEDGEMENTS

I would like to express my deepest appreciation to Dr. Gerald
Miller, for the many hours spent on my behalf directing the thinking,
execution, and writing of this dissertation. Mr. Albert Talbott, who
provided a great deal of assistance in the statistical analysis of the
study also deserves my deepest appreciation. Their assistance, as friends
and colleagues, forced a clarity which was often lacking in the random
thoughts of the writer.

For assistance throughout my advanced degree program, I would like
to thank: Dr. David K. Berlo, Chairman, Department of Communication;
Professor Leo Martin, the Chairman of my doctoral committee; and Dr.'s
Charles Blackman, Carl Gross, Hideya Kumata, Colby Lewis, and
Malcolm MacLean, Jr.

For typing of this manuscript and assistance and guidance through-
out my advanced degree program, I would like to express my appreciation
to Mrs. Shirley Sherman. Her cheerfulness and willingness to help has
made many a bleak day easier to accept.

The greatest debt of gratitude is owed to my wife Betty, for her
willingness to accept the role of graduate student wife and provide stability
during a period when it is needed; and to my children Kevin, Kurt, Kerry,
and Kit, who were never sure what degree and dissertation meant to them

but were willing to go without their father for long periods of time.

ii

CHAPTER

I

II

III

IV

TAB LE OF CONTENTS

RATIONALE AND HYPOTHESES . . . . . . . .

Rational? o e e e e e o o e e e e e e
HypotheSeS o o o o e e e e o e e o 0

METHOD 0 O O O O O O O O O O O O O I O

SUbjeCtS e o o o o o e e e o o o e 0
Procedure 0 o e e e e e e c o e o e e
Dependent Variables e e e o o o o o 0
Self and Source Credibility . . . .
TaSk DiffiCUltY e e e e o o o o e 0
Additional Dependent Measures . . .

RESULTS 0 O O O O O O O O O O O O O O O

Manipulation of Independent Variables
SGlf credibility e e e e e e e e 0
Source Credibility . . . . . . . .
FeedbaCk e e e e e e e e e o e e e

Hypotheses of the Study . . . . . . .

Additional Analysis . . . . . . . . .

CONCLUSIONS, DISCUSSION, AND IMPLICATIONS FOR

FURTHER RESEARCH . . . . . . . . . . .

COUClUSiOnS e o o o o e o e o o e o e
DiSCUSSion o e o e e e o e o e o e 0
Implications for Further Research . .

REFERENCED BIBLIOGRAPHY . . . . . . . . . . . .

APPEXWDIX A 0 O I O O O O O O I O O O O O O O O O

Questionnaire . . . . . . . . . . . . . . . .
Positive Feedback Schedule . . . . . . . . . .
Negative Feedback Schedule .... . . . . . . . .
Diagrams of Puzzles . . . . . . . . . . . . .

iii

Page

In

la
la
19
2O
21
23

2Q

2Q

_ 2M

27
30
31
NO

#6
#6
52
56
59
62
63
72

73
71+

‘ll‘ll-lll'l'l

LIST OF TABLES

TABLE Page

)

1. Pre Test Means for Subjects Assigned to High or 24
Low Self Credibility

2. Summary Table of the Three Factor Analysis of
Variance of Pre Test Scores of Self Credibility 25
3. Pre Test Self Credibility Means for Subjects in Each
of the Eight Treatments 25
u. Critical Difference Test Between Self Credibility
Treatment Means 26
5. Pre Test Means for Subjects Assigned to High or
Low Source Credibility 27
6. Summary Table of the Three Factor Analysis of
Variance of Pre Test Scores of Source Credibility 27
7. Pre Test Source Credibility Means for Subjects in
Each of the Eight Treatments 28
8. Critical Difference Test Between Pre Test Source
Credibility Treatment Means 28
9. Sign Tests for Large Samples to Test the Experimental
Manipulation of Self Credibility and Source
Credibility within Individual Treatments 29
10. Post Test Means for Subjects Assigned to Positive
or Negative Feedback 30
11. Summary Table of the Three Factor Analysis of
Variance of Post Test Positive --- Negative Scores 31
12. Post Test Positive --- Negative Means for Subjects in
Each of the Eight Treatments 31
13. Critical Difference Test Between Post Test Positive ---
Negative Treatments Means 32
1a. Pre Test Minus Post Test Mean Differences for Subjects
in the High Self Credibility, High Source Credibility,
and Positive Feedback Condition 33
15. Pre Test Minus Post Test Mean Differences for Subjects
in the High Self Credibility, High Source Credibility, and
Negative Feedback Condition 3”

iv

TABLE

16.

17.

18.

19.

20.

21.

22.

23.

2H.

25.

26.
27.
28.

29.

30.

31.

LIST OF TABLES (con't.)

Pre Test Minus Post Test Differences for Subjects in
the High Self Credibility, Low Source Credibility, and
Positive Feedback Condition

Pre Test Minus Post Test Mean Differences for Subjects
in the High Self Credibility, Low Source Credibility,
and Negative Feedback Condition

Pre Test Minus Post Test Mean Differences for Subjects
in the Low Self Credibility, High Source Credibility,
and Positive Feedback Condition

Pre Test Minus Post Test Mean Differences for Subjects
in the Low Self Credibility, High Source Credibility,
and Negative Feedback Condition

Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, Low Source Credibility,

and Positive Feedback Condition

Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, Low Source Credibility,

and Negative Feedback Condition

Grand Summary of Pre --- Post Mean Differences for the
Dependent Variables, Self Credibility, Source Credibility,
and Task Difficulty

Mean Number of Seconds Taken to Describe the Puzzle

Summary Table of the Three Factor Analysis of Variance
of Time Taken to Describe the Puzzle

Critical Difference Test for Treatment Means on Time
Taken to Describe the Puzzle

Choice of Stopping Time and Self Credibility
Choice of Stopping Time and Source Credibility
Choice of Stopping Time and Feedback

Mean Number of Information Questions Asked by Subjects
in the Eight Treatments

Summary Table of the Three Factor Analysis of Variance
of Number of Information Questions Asked

Critical Difference Test for Treatment Means on Number of
Information Questions Asked

V

Page

35

36

36

37

38

39

40

Ml

42

#2
U3
#3

Q3

”4

#5

us

FIGURE

1.

LIST OF FIGURES

 

Possible conditions of the independent
Summary of the hypotheses . . . . . .
Geometric puzzle . . . . . . . . . . .
Top view of separation device . . . .
Puzzle shown at end of experiment . .

Scales for self and source credibility

vi

variables

Page

l3
17
18
19

2O

LIST OF APPENDICES

 

Page

APPENDIX A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Positive Feedback Schedule . . . . . . . . . . . . . . . . . . . 72
Negative Feedback Schedule . . . . . . . . . . . . . . . . . . . 73

Diagrams Of PPaCtiCG PUZZleS o e o o o o e o e o o e e e o o o 0 e 7“

vii

CHAPTER I

RATIONALE AND HYPOTHESES

Rationale

An individual (A) enters any task oriented interaction with a set
of expectations resulting from prior experience in similar situations.
These expectations involve at least three major factors: first, A's
expectations concerning the perceived difficulty or complexity of any task
associated with the interaction; second, A's expectations concerning his
ability to perform any task associated with the interaction; and third,
A's expectations concerning the task abilities of other individuals in
the verbal interaction situation. These three factors will hereafter be
labeled Task Difficulty, Self Credibility, and Source Credibility.

During a particular interaction, individual A either maintains
previous behavior that he has found rewarding or adjusts his behavior to
meet the exigencies of the situation. If such an adjustment becomes
necessary, it is facilitated by the response that the receiver(s) makes
to the communication efforts of individual A. While these responses may
be verbal or non-verbal, this experiment is concerned only with the adjust-
ment of expectations resulting from verbal responses.

Receiver responses that lead to adjustment of source expectations
and behavior are commonly labeled "feedback" (10). The concept "feedback"
was first used in the field of cybernetics (28). Later it was extended as
a central concept in theoretical formulations concerning more complex
social behavior (13)(29). While these theories also provide for intra-

personal feedback loops, this study is concerned with the effects of an

interpersonal feedback loop. Specifically, it focuses upon the adjustment
of expectations and behavior on the part of individual A which result
from feedback from another individual.

Murphy defines "feedback" as, "a verbal pattern [that] may be
positive (maintaining) or negative (correctiveY'(16). In an extensive
review of the experimental manipulation of verbal behavior, Salzinger
relates the concepts of feedback and reinforcement. He defines the values
of the variable "reinforcement" as either:

Positive - any event produced by one individual after the
behaVLor of another individual which strengthens that behavior.

Ne ative - any event, produced by an individual, the

elimination or avoidance of which strengthens the

behavior which eliminates or avoids it (23).

By integrating these two definitions, it is possible to define
conceptually, for individual A, two values of the variable "feedback" as:

Positive Feedback - any verbal pattern produced by individual

B after a communication by individual A which strengthens

and maintains the behavior of A.

Negative Feedback - any verbal pattern produced by an

individual B, after a communication by individual A, the

elimination or avoidance of which strengthens the corrective

communication behavior by individual A which eliminates or

avoids it.

Research concerning the effects of feedback has dealt primarily
with observable behavior. Leavitt and Mueller (11) found that feedback
increases the accuracy with which information is transmitted. Salzinger,
summarizing the extensive research on the effect of positive reinforcement
(feedback) in verbal conditioning experiments, stated that "there have

been very few studies on verbal behavior which report on the use of

negative reinforcements" (23).

Aside from verbal conditioning research, there is a body of
research dealing with the effects of negative feedback on encoding
variables. Greenberg and Tannenbaum (8), using written responses, used
two time measures and five grammar measures. In analyzing verbal
responses, such dependent measures as total utterance and non-fluencies
(1n) disfluencies (25), voice volume, time Spent encoding, and pauses
and frequencies of occurrence of various types of Speech hesitation
phenomena (26) have been employed. Generally, this research demonstrates
that negative feedback produces a significant effect on encoding variables.

It is probable, however, that such changes or adjustments in ob-
servable encoding behavior are accompanied by changes in cognitive expecta-
tions. Support for such an assumption is provided by Miller and Mischel.
Miller varied all possible combinations of approval-neutral-disapproval
conditions for a confederate first speaker and a naive second speaker.

He suggests that - aside from the experimental effect on observable
encoding variables - those subjects who received treatment inferior to
that extended to the confederate also manifested cognitive change.
Specifically, the questionnaire data for subjects in the inferior group
revealed that they:

...were significantly more dissatisfied with their Speeches

following presentation...felt they had done a significantly

poorer job of Speaking in comparison with the first speaker...

and had a significantly"more unfavorable reaction toward the

experimenter than subjects in the Same and Superior groups. (lu)
Mischel investigated the relation between the public-private nature of
the Situation in which expectancy (goal) statements are elicited and the

amount of change in related expectancy statements following negative feed-

back for performance on a first task. After giving his subjects eleven

points lower than what they expected on the first task, he found that the
subjects lowered their expectancy statements for a second task. Subjects
in the private situation lowered their expectancy statements significantly
more than subjects in the public Situation (15).

This research supports the assumption that changes in encoding be-
havior as the result of feedback are accompanied by cognitive change. It
further supports the assumption that cognitive change may manifest itself
in any or all of the three conceptually defined areas of Task Difficulty,
Self Credibility, or Source Credibility.

Having examined the factor of Task Difficulty in the previously cited
research by Miller and Mischel, a brief review of two bodies of research
literature is necessary to develop conceptually the cognitive concepts
Self Credibility and Source Credibility.

Research concerning the properties which a receiver perceives in a
source has been categorized under the label "Source Credibility." This
research has tended to conceptualize Source Credibility in various ways.

A recent study attempted to provide empirical factors for conceptualizing
and operationalizing Source Credibility. The Study utilized:

1. Nine public sources, with no consideration of their
relation to any subject matter.

2. Three personally-known sources: 'a person you know' whose
'opinion you respect'; neither 'respect nor disrespect';
and 'do not respect'.

3. Six public sources, placed in a 'relevant' or
'irrelevant' relation to a topic (12).

and submitted these sources to subjects for rating on 83 bipolar scales.

Factor Analysis of the obtained ratings identified three major factors:

 

safety, qualification, and dynamism.l The study indicated that any
measurement of source credibility involving qualification would be best
represented by solution three, the solution with the Six public sources
related to a tOpic.

In a verbal interaction situation, the roles of source and receiver
are interchangeable. From the position of individual A, it follows that if
there are properties which he, as a receiver, perceives in a source, then
he will perceive similar properties in himself when he plays the role of
a source. For clarity I will refer to those properties which individual

A, as a receiver, perceives in a source as Source Credibility_and those

 

similar properties which individual A, as a source, perceives in himself

as Self Credibility.

 

The concept Self Credibility, as developed here, is not meant to be
equated with what has been developed in the psychoanalytic literature (5)
(30) as the individual's "self concept." An individual's "self concept"
is viewed in this literature as a generalized perception of the self, a
view of one's own self which holds across all interaction situations.

Self Credibility, being developed conceptually from the literature concerning
Source Credibility, is viewed as a portion of the individual's "self
concept;" specifically, that portion which is related to the task the
individual is engaged in during an interaction.

It is further proposed that when applied to a specific social

 

1Examples of scales identifying each factor are (a) safety: stable-

unstable, reasonable-unreasonable (b) qualification: educated-uneducated,
trained-untrained (c) dynamism: bold-timid, energetic-tired.

situation, the concept "self esteem" can be viewed as "self credibility."
The research literature suggests that the "self esteem" of an individual
is a significant variable in situations involving social interaction.
Stotland and Cottrell, using a Q Sort as a measure of self esteem, found
that, "persons low in self esteem are more subject to social influence
than persons high in self esteem" (27). Rosenbaum and deCharms, measuring
residual hostility toward an attacker, found that high self esteem subjects
were not very strongly instigated toward aggression and showed some
insensitivity to cues presented by other persons (2l). Cohen indicated
that persons with high self esteem are characterized by tendencies to
protect themselves from negative self evaluation (A).

Generalizing the cited research on self-esteem to a verbal in-
teraction situation, it is proposed that any adjustment of cognitive
expectations concerning individual A's perceived Self Credibility and
the credibility of the source of feedback will be a function of the
relationship between the two and the nature of the feedback individual
A receives from the source.

In terms of individual A, four variables which effect a verbal in-
teraction situation have been conceptually defined. These four variables
are Task Difficulty, Self Credibility, Source Credibility, and Feedback.
While each may be viewed as a continuous variable, Task Difficulty takes
the extreme values of difficult and easy, Self Credibility and Source
Credibility the extreme values of high and low, and Feedback the extreme
values of positive and negative.

Given these extreme values, the independent variables of Self

Credibility, Source Credibility, and Feedback yield eight possible

conditions which individual A can perceive as existing. These posSible
conditions are summarized in Figure 1.

Thus far it has been suggested that individual A uses the feedback
he receives in a verbal interaction situation as a mechanism for adjust-
ment, and that this adjustment subsequently is manifested in both
observable encoding behavior and cognitive change. Two Questions remain
to be answered: first, under what conditions of Self Credibility, Source
Credibility, and Feedback can cognitive change be expected; and second,
what predictions can be made relating to the specific forms this

cognitive chance may take?

Perceived Nature of Perceived Credibility
Self Credibility Feedback of the Source of Feedback
High Pos. High
High Neg. High
High Pos. Low
High Neg. Low
Low Pos. High
Low Neg. High
Low Pos. Low
Low Neg. Low

"Fig. 1. Possible conditions of the independent variables."

For example; can individual A, who perceives himself as highly credible,
be expected to experience a cognitive readjustment if he receives negative

feedback from a low credible source? And, if so, will this readjustment

take the form of changes in his perceived Self Credibility, changes in
perceived Source Credibility, changes in perceived Task Difficulty, or
changes in some combination of these three variables?

In the literature of psychology, some light is thrown on these
questions by a group of theories commonly listed under the general rubric
of "balance theories" (7)(9)(l7)(19). These theories have been reviewed
in a number of publications (3)(18). While they differ in certain of
their elements, they share a set of common assumptions. Bettinghaus
summarizes these assumptions as follows:

1. All cognitive balance theories assume that the individual
places values on the objects and signs which he perceives.

2. All balance theories postulate that when two or more objects
are placed within the same cognitive or perceptual field,
the field may be described in terms of the values which
the individual originally placed on the objects within the
fields

3. All balance theories hypothesize that when the values held
for objects within the same perceptual or cognitive field
are internally consistent, a state of balance exists, and
the individual will do nothing to change the values he holds
toward the objects.

u. All balance theories assume that when the values placed on
objects within the perceptual or cognitive field are not
consistent for the individual, a state of imbalance exists,
and an individual will take steps to restore balance to
the perceptual or cognitive field.

5. All balance theories hypothesize that the strength of the
pressures to reduce imbalance are a direct function of the
magnitude of the imbalance (1).

While all of these assumptions are germane to the present Study,

assumption four is of Special Significance. This assumption indicates

that cognitive change can be expected when the relationships between

cognitive elements are not consistent, or balanced. The relationship

between the cognitive elements, Self Credibility, Source Credibility,

and Task Difficulty has been discussed above. When an individual receives
feedback which produces inconsistency among these cognitive elements,
these elements can be expected to change. As Festinger states:

The simplest definition of dissonance (Festinger's term for

cognitive inconsistency), can, perhaps, be given in terms of

a person's expectations. In the course of our lives we have

all accumulated a large number of expectations about what

things go together and what things do not. When such an ex-

pectation is not fulfilled, dissonance occurs (6).

Among the various balance theories presently found in the literature,
the theoretical formulation which most closely approximates the experi-
mental situation developed in this study is that of Newcomb (l7).

Newcomb is the only balance theorist to incorporate an interpersonal
dimension in his formulation. In Newcomb's system the communicative act
occurs when:

A communicates to B about some event, X. The important

variables are A's orientation toward X, A's orientation toward

B, B's orientation toward X, and B's orientation toward A,

each such orientation conceived as being potentially positive

or negative in sign (18).

The cognitive elements of an interaction situation, involving a
task, described in this study, closely parallel the elements described
by Newcomb, A's orientation toward X can be equated with the variable
Task Difficulty. A's orientation toward B can be equated with the variable
Source Credibility. Both B's orientation toward X and A can be equated
with A's perception of the feedback he receives from B. The only place
where a close parallel cannot be established is the area of Self

Credibility. Neither Newcomb nor the other balance theorists explicitly

incorporate such a variable into their theoretic system.

lO

Newcomb States that when the relationship between the elements is
not consistent for the individual, an imbalanced state exists and the in-
dividual will make an effort to restore balance. Restoration of balance,
however, may be achieved by Several different means. In the interaction
situation of interest in this study, several specific mechanisms may be
suggested by which individual A can restore balance. For instance, if A
has high perceived Self Credibility regarding a particular task, and if
he receives negative feedback regarding his performance on the task from
a high credible source, cognitive imbalance should result. Here A has
at least four avenues available for restoring balance: he may subsequently
lower his perception of his own Self Credibility, he may subsequently
lower his perception of the credibility of the source, he may subsequently
perceive the task as more difficult, or he may change the value of more
than one of these three variables. Since the general assumptions of
balance theories provide no means for Specifying which of these alternatives
will be chosen, it is felt that one of the major contributions of this
study will be to provide information on this problem for interaction
situations involving various combinations of the values of the independent
variables Self Credibility, Source Credibility, and Feedback.

Given the above limitation, the hypotheses that follow are offered
with reservation and are to be regarded as tentative predictions derived
from the general assumptions of balance theories. The predictions are
based on the stipulation of those interaction situations in which the
cognitive elements described in this study are consistent and the stipulation
of those interaction situations in which the cognitive elements are in-

consistent. The direction of change in the cognitive elements, in order

11

to restore balance, is also predicted.

Hypotheses

 

The following hypotheses were investigated in this study:

1. Under conditions of High Self Credibility and Positive
Feedback from a High Credible Source, no cognitive imbalance
will result, and; therefore there will be no cognitive change
in perceived Self Credibility, perceived Source Credibility,
or perceived Task Difficulty.

2. Under conditions of High Self Credibility and Negative
Feedback from a High Credible Source, cognitive imbalance will
result. Reduction of this imbalance will take the form of a
decrease in perceived Source Credibility and an increase in
perceived Task Difficulty.

3. Under conditions of High Self Credibility and Positive
Feedback from a Low Credible Source, no cognitive imbalance
will result, and; therefore, there will be no cognitive change
in perceived Self Credibility, perceived Source Credibility,
or perceived Task Difficulty.

u. Under conditions of High Self Credibility and Negative Feedback
from a Low Credible Source, cognitive imbalance will result.
Reduction of this imbalance will take the form of a decrease
in perceived Source Credibility and an increase in perceived

Task Difficulty.

l2

5. Under conditions of Low Self Credibility and Positive Feedback
from a High Credible Source, cognitive imbalance will result.
Reduction of this imbalance will take the form of an increase in
perceived Self Credibility and a decrease in perceived Task

Difficulty.

6. Under conditions of Low Self Credibility and Negative Feedback
from a High Credible Source, no cognitive imbalance will result,
and; therefore, there will be no cognitive change in perceived
Self Credibility, perceived Source Credibility, or perceived

Task Difficulty.

7. Under conditions of Low Self Credibility and Positive Feedback
from a Low Credible Source, cognitive imbalance will result.
Reduction of this imbalance will take the form of a decrease

in perceived Task Difficulty.

8. Under conditions of Low Self Credibility and Negative Feedback
from a Low Credible Source, no cognitive imbalance will result,
and; therefore, there will be no change in perceived Self
Credibility, perceived Source Credibility, and perceived Task

Difficulty.

The stated hypotheses are summarized in Figure 2.

l3

 

 

Assigned condition Predicted Charge in Dependent Variables
SLF SOC FB SLF SOC TD
High High Positive 0 O 0
High High Negative 0 - +
High Low Positive 0 O 0
High Low Negative 0 - +
Low High Positive + O -
Low High Negative 0 O 0
Low Low Positive 0 O -
Low Low Negative 0 0 0

"Fig. 2. Summary of the hypotheses."

CHAPTER II

METHOD

Subjects

Pretesting data indicated that the sex of the subject may
Significantly affect the interaction patterns. Because of this possibility,
only female subjects were utilized in this experiment. Eighty female
subjects, all undergraduate students enrolled in the Communication 100
class at Michigan State University, volunteered to participate in the
experiment. While the experimenter was an instructor in the course, the
subjects had no knowledge of the nature of the experiment. Subjects

were randomly assigned to treatment cells in groups of eight.

Procedure

The confederates employed in the experiment were four undergraduate
males. Each confederate received training during the pretesting period
and again prior to the actual experiment. To minimize variability,
assignment of the four confederates was counter-balanced across all
treatments. Each confederate was paired with a minimum of two subjects
in each cell, and a maximum of three subjects in four of the eight cells.
Bach confederate interacted with twenty subjects, ten positively and ten
negatively.

Three people were involved in the experimental procedure: the subject
(A), a confederate (B) and the experimenter (B).

After A and B were seated, E read this statement to them:

in

15

I would like to ask each of you not to discuss this study

with any of your classmates or friends, as many of them will

be participating in it. Later in the quarter I will be

coming into each of your classes to explain the results of

the Study. O.K.?

A and B were then given questionnaires, and after reading the first
page were asked to answer the questions on page two dealing with age, sex,
year in school, and other biographical data. The complete questionnaire
may be found in Appendix A.

After indicating that they were ready, A and B were read this
statement.

The task you are about to engage in involves cooperation

between both of you in putting together a geometric puzzle.

In this particular part of the experiment subject A will

be describing the puzzle while subject 8 will actually be

putting it together. Our experience has shown us that people

do better at this task if they are given a chance to practice,

so we have provided for a Short acclimation and training

period for each of you before we start the actual task.

Subject B (confederate), if you will take this card to

Mr. Weston in room 319, the first door to your right, he

will give you some training in putting the pieces together.

Subject A (subject), if you will stay here I will give you

some training in describing a similar puzzle.

After B left the room, A was given one or two similar puzzles to
describe. (Diagrams of these puzzles are contained in Appendix A.)

This particular task was chosen for two reasons. First, it could be
assumed that the subjects would have had very little prior experience with
a task of this nature. This would reduce any contaminating effects caused
by prior experience, and would allow for more effective manipulation of
the independent variables of Self and Source Credibility. Second, the

task was of such a nature that it could be assumed that each subject

would approach the description of the puzzle on a piece by piece basis.

This factor

16

allowed for construction of a feedback schedule where

feedback was provided after the subject had described the position of

each piece,

thus allowing for more effective manipulation of the in-

dependent variable of Feedback. Pretesting data indicated that both of

these assumptions were valid.

A then participated in the practice period with E. Irrespective

of A's performance during the practice period, A was stopped after five

minutes and

thirty seconds. Depending upon the random assignment of A,

one of the two following statements was read to him in a conversational

tone of voice.

If A

If A

were assigned to a condition involving high Self Credibility:

You have included many of the details. You do real
well at explaining the puzzles, much better than
most people do. We don't have any more time to
practice.

were assigned to a condition involving low Self Credibility:

You have missed many of the details. You don't do
too well at explaining the puzzles. Most people
do better. However, we don't have any more time
to practice.

After precisely six minutes had elapsed, B returned. B was seated

and asked by B to return the card from Mr. Weston which had previously

been given by E to B. After receiving the card from B, E read one of

the following statements aloud in a conversational tone of voice.

If A

were assigned to a condition involving high Source Credibility:

Oh, I see that Mr. Weston says that you do real
well at putting together the pieces.

17

If A were assigned to a condition involving low Source Credibility:

Oh, I see that Mr. Weston says that you don't do
too well at putting together the pieces.

After hearing the statement, the confederate (B) replied with a con-
versational "Yeah".

Both A and B were then asked to turn the page of the questionnaire.
After reading a brief description of the task and an explanation of how
to mark the scales, A completed the pre measures of Self Credibility,
Source Credibility, and Task Difficulty.

A was then given the geometric puzzle in Figure 3 to describe to

"Fig. 3. Geometric puzzle."

The interaction involving the description of the puzzle by A took
place in a large, quiet room. To control the possible communication
effects of having A and B visible to each other, a device was constructed
which separated the subject, confederate, and the experimenter. This
device was 32 inches high and was placed on a large table. A diagram of

the device may be found in Figure A.

18

 
 
    

Experimenter

Confederate

"Fig. u. Top view of separation device."

During the interaction, B responded to the efforts of A to
describe the puzzle by using a pre-arranged positive or negative feedback
schedule. Appendix A contains copies of the schedules. The positive
and negative feedback schedules were revised slightly as a result of
pretesting. Every attempt was made to equate the strength of the two
schedules. Pretesting data indicated that subjects perceived the two
schedules as positive or negative. Of twenty pretest subjects, eighteen
responded on the positive-negative scale consistent with the treatment
to which they were assigned, (p. = (.05). Confederates were instructed
to adhere to the schedule as rigidly as possible. Where it was necessary
to respond to questions by the subject, confederates were instructed
to reply with a positive answer if the subject was assigned to a positive
treatment group, and a noncommital answer if the subject was assigned to
a negative treatment group. During the interaction, E kept a record of
the amount of time it took each subject to explain the puzzle. Additional
measures of the subject's choice of stopping time and the number of in-
formation questions directed to B by A were also obtained. E remained

silent during the entire interaction.

19

Upon completion of the interaction, post measures of Self Credibility,
Source Credibility, and Task Difficulty, were obtained from each subject.
Before being excused, A and B were shown the puzzle and the card
containing the description. The puzzle in Figure 5 was presented to all
subjects. The puzzle was presented with two minor errors for two reasons.
First, it was done to relieve any feelings of inadequacy which may have
developed in subjects under negative feedback conditions. Subsequent
written remarks by many of the subjects under negative feedback conditions
indicated that the puzzle shown at the conclusion of the interaction
served this purpose. Secondly, it was decided that the completed puzzle
with a few minor mistakes was a reasonable expectation for all subjects.
A and B were then asked again not to reveal details of the study to friends

or classmates and were excused.

"Fig. 5. Puzzle shown at end of experiment."

Dependent Variables

 

Measures required were pretest and post-test scores of: (1) Self
Credibility, (2) Source Credibility, and (3) Task Difficulty. In

addition, measures of the amount of time to complete the task, the subject's

20

choice of stopping time, and information questions asked by the subject
were utilized.

Self and Source Credibility - The measuring instrument for these dependent
measures was constructed on the basis of the factor analytic study by
Lemert (12) mentioned above. Ten scales from the qualification factor
were chosen for pretesting. Pretest data indicated that three of the
scales did not discriminate adequately; consequently they were eliminated
from the final analysis. The remaining seven scales were utilized as a
measure of Self and Source Credibility. Figure 6 provides a listing of

these seven scales, along with the factor loading for each scale.

Qualification Scales Factor Loadings
expert---ignorant .90
competent--—incompetent .88
capable---incapable .87
knowledgeable---unknowledgeable .87
efficient---inefficient .78
successful---unsuccessful .76
confident---unsure .59

"Fig. 6. Scales for self and source credibility."

The seven scales were randomly distributed on the pre and post questionnaires
and were systematically reversed to eliminate order effects. Each sub-
ject's score was obtained by asking her to mark each scale in relation to
the following statements:
For pre-test Self Credibility:
"Myself, in relation to the task about to be performed."
For pre-test Source Credibility:
"Subject B, the person c00perating with you, in relation

to the task about to be performed."

21

For post-test Self Credibility:

"Myself, in relation to the task just completed."

For post-test Source Credibility:

"Subject B, the person c00perating with you, in relation
to the task just completed."

Each subject's score was arrived at by summing across the seven
scales. The mean of the subject's raw pre-test and post-test scores was
used in the analyses to test the hypotheses.

Task Difficulty--Three scales utilizing the polar adjectives, easy---
difficulty, simple---complex, and clear---unclear, were employed to
measure Task Difficulty. The scales were randomly distributed on the

pre and post questionnaires and were systematically reversed to eliminate
order effects. Each subject's score was obtained by asking her to mark
each scale in relation to the following statements:

For pre-test Task Difficulty:

"Respond to the following scales in relation to how you
perceive the cooperative task in which you are about
to engage."

For post-test Task Difficulty:

"Respond to the following scales in relation to how you
perceive the cooperative task in which you have just
engaged."

Each subject's score was arrived at by summing across the three
scales. The mean of the subject's raw pre-test and post-test scores

was used in the analyses to test the hypotheses.

22

In addition, a series of six scales, appearing under the heading
"Information," were included in the questionnaire. The primary purpose
of these scales was not to provide data bearing on the hypotheses, but
rather to provide a means for assessing the effectiveness of the manipu-
lation of the independent variable of feedback. The six scales utilizing the
polar adjectives, intimate---remote, good---bad, positive---negative,
sensible---not sensible, clear——-not clear, and helpful---not helpful,
were randomly distributed on the pre and post questionnaires and were
systematically reversed to eliminate order effects. Each subject was
asked to mark each scale in relation to the following statements:

For pre-test Information:

"ReSpond to the following scales in relation to the
information you expect to receive from your partner."

For post-test Information:

"Respond to the following scales in relation to the
information you received from your partner."

The positive---negative scale was chosen to test the experimental
manipulation of feedback, because pretesting data indicated that it
represented most accurately the subjects perception of the feedback they
had received.

The measures for Self Credibility and Source Credibility were on
separate pages of the questionnaire. The measures for Task Difficulty and
Information were combined on one page. Pages were systematically rotated
to counter-balance any effects which might have been produced by having

all the subjects scoring the dependent measures in the same order.

23

The additional measures kept by the experimenter during the
interaction were:
Time to complete the task - Two measures of time taken to complete the task
were obtained. Timing began when the subject was given the puzzle. The
elapsed time was recorded when the subject had completed describing the
puzzle for the first time, and again when the subject decided to stop
work on the puzzle.
Subject choice of stopping time - After the subject had completely
described the puzzle for the first time, She was afforded the opportunity
to stop or to go over the puzzle again. Measures of the choice of stopping
time by the subject were kept by the experimenter. A "l" was recorded
if the subject Stopped at the first opportunity, and a "2" was recorded
if the subject stopped other than at the first Opportunity. Only one
chose to go over the puzzle more than two complete times.
Information questions asked by the subject - The experimenter kept a record
of the number of interrogative attempts by the subject to obtain in-

formation from the confederate.

CHAPTER III

RESULTS

This chapter contains an analysis of the data to test the efficacy
of the eXperimental manipulation of the independent variables and an

analysis of the data to test the hypotheses.

Manipulation of the Independent Variables

 

Several analyses were conducted to test the efficacy of the ex-
perimental manipulation of Self Credibility, Source Credibility and

Feedback. Following is a summary of the results of these analyses.

Self Credibility
The pre test means for subjects in the High and Low Self Credibility
treatments, when collapsed across all conditions of Source Credibility

and Feedback, are found in Table 1.2

Table 1. Pre Test Means for Subjects Assigned
to High and Low Self Credibility

Assigned to: Mean
High Self Credibility 3.28
Low Self Credibility 3.93

Table 2 contains a summary of the three factor analysis of variance,

utilized to test the significance of these data.

 

2When observing the means in all of the analyses in this chapter,
as the mean approaches the value of 0.00, this is indicative of a high
value of the variable under consideration. AS the mean approaches the

value of 7.00, this is indicative of a low value of the variable under con-
sideration. 29

25

Table 2. Summary of the Three Factor Analysis of

Variance of Pre Test Scores of Self Credibility

 

 

Source of variance .9:; _Jfii_

Between
Self credibility (A) l 8.45
Source credibility (B) l .80
Feedback (C) l .01
A/B 1 .HI
A/C 1 .07
8/0 1 .SM
A/B/C 1 .5”

Within error 72 1.16

* = p <'.05 two tailed

F

7.26
41.00
(1.00
(1.00
(1.00
(1.00

(1.00

5’:

The Significant main effect on Self Credibility supports the ex—

perimental procedures used to establish conditions of High and Low Self

Credibility.

The Self Credibility means for subjects in the eight treatment

conditions may be found in Table 3. It can be seen that the means for

subjects in the four treatments utilizing High Self Credibility are con-

sistently higher than those for subjects in the four Low Self Credibility

Table 3. Pre Test Self Credibility Means for

Subjects in Each of the Eight Treatments

 

HSOC- HSOC- LSOC-
PFB NFB PFB

 

 

High Self Credibility 3.15 3.07 3.50

Low Self Credibility ”.05 3.76 3.79

LSOC-
NFB

 

3.41

n.1u

treatments.

of the 28 possible comparisons among means.

of these comparisons.

26

A critical difference technique was utilized to make each

Table A contains a summary

It can be seen that the significant main effect in Table 2 is

caused primarily by the Low Self Credibility, Low Source Credibility,

Negative Feedback treatment and the High Self Credibility, High Source

Credibility, Negative Feedback treatment.

Even though not all of the

High and Low Self Credibility treatments differ Significantly, the

Table 14.

Critical Difference Test Between Pre Test Self Credibility
Treatment Means

 

Assigned condition

SLF SOC FB

H

H

consistently higher means for subjects in the former conditions indicate

relatively successful manipulation of the variable of Self Credibility.

H

H

P

N

P

N

P

N

* = p«( .05

H-H-P

 

 

-.08

 

+.35

+.H2

critical difference

 

= .96

 

 

L-L-P

 

 

+.53

+.O9

+.38

+.35

27

Source Credibility

 

The pre test means for subjects in the High and Low Source
Credibility treatments, when collapsed across all conditions of Self
Credibility and Feedback, are found in Table 5.

Table 5. Pre Test Means for Subjects Assigned
to High or Low Source Credibility

Assigned to: Mean
High Source Credibility 2.28
Low Source Credibility 3.95

Table 6 contains a summary of the three factor analysis of variance,

utilized to test the significance of these data.

Table 6. Summary Table of the Three Factor Analysis of Variance
of Pre Test Scores of Source Credibility

 

Source of variation if; _M_§__ L
Between
Self Credibility (A) 1 1.97 3.95
Source Credibility (B) l 7.72 18.09*
Feedback (C) l .05 l 1.00
A/B 1 .91 < 1.00
MC 1 .141 < 1.00
B/C l .98 2.30
A/B/C 1 .10 4, 1.00
Within error 72 .ua

* = p (.05 two tailed

28

The significant main effect on Source Credibility supports the
experimental procedures used to establish conditions of High and Low
Source Credibility.

The Source Credibility means for subjects in the eight treatment
conditions may be found in Table 7. It can be seen that the means for

Table 7. Pre Test Source Credibility Means for
Subjects in Each of the Eight Treatments

HSLF- HSLF- LSLF- LSLF-

 

 

 

 

PFB NFB PFB NFB
High Source Credibility 3.05 3.01 2.92 2.81
Low Source Credibility 3.68 3.3a 3.99 3.29

subjects in the treatments utilizing High Source Credibility are con-
sistently higher than the means for subjects in the four Low Source
Credibility treatments. A critical difference technique was utilized to
make each of the 28 possible comparisons among means. Table 8 contains a
Table 8. Critical Difference Test Between Pre Test Source Credibility

Treatment Means
Assigned condition

 

 

 

 

 

 

 

 

 

 

SLF. SOC £§_ H-H-P H-H-N L-H-P_ L-H-N H-LPP H-L-N L-L-P L-L-N
H H P ---- -.04 -.63* -.2A +.63* +.29 +.uu +.2H
H H N ---- -.59* -.20 +.67* +.33 +.u8 +.28
L H P ---- +.39 +1.26* +.92* +1.07* +.87*
L H N ---- +.87* +.53 +.68* +.u8
H L P ---- -.3u -.19 -.39
H L N ---- +.15 -.05
L L P ---- -.20
L L N ----

J.

-~ = p (.05 two tailed critical difference = .59

 

oil I ll:
. v.11

 

lIlI Illllll

29

summary of these comparisons.

Investigation of Table 8 indicates that eight of the ten significant
differences are between High and Low Source Credibility conditions, pro-
viding support for the effectiveness of the experimental manipulation
of Source Credibility.

Having established that the experimental manipulation of Self
Credibility and Source Credibility was generally successful, two Sign
Tests were utilized to establish the efficacy of the manipulation of Self
Credibility and Source Credibility within each individual treatment. The
results of these Sign Tests are found in Table 9.

Table 9. Sign Tests for Large Samples to Test the Experimental
Manipulation of Self Credibility and Source Credibility
Within Individual Treatments

Assigned conditions of Self Credibility

 

and Source Credibility_ N z p
SAME 33 1.01; {.28
DIFFERENT 36 2.83 <}01 2T

It can be seen from Table 9 that when subjects were assigned to
different conditions ofSelf Credibility and Source Credibility, they
reflected these differences in their pre test Self Credibility and Source
Credibility scores, at a greater than chance frequency. On the other
hand, when subjects were assigned to the same Self Credibility and Source
Credibility conditions no significant differences in their pre test Self

Credibility and Source Credibility scores were obtained.

30

Feedback

Finally, to test the experimental manipulation of Feedback, the
post test positive-negative scores for subjects in the eight treatments
were analyzed. These means, when collapsed over all conditions of Self
Credibility and Source Credibility, are found in Table 10.

Table 10. Post Test Means for Subjects Assigned
to Positive or Negative Feedback

 

 

 

Assigned to: Mean
Positive Feedback 1.60
Negative Feedback 4.63

Table 11 contains a summary of the three factor analysis of
variance, utilized to test the Significance of these data.

The Significant main effect for Feedback supports the experimental
procedures used to manipulate the variable Feedback.

The Positive -- Negative means for subjects in the eight treatment
conditions can be found in Table 12. It can be seen that the means for
subjects in the four treatments utilizing Positive Feedback are con-
sistently higher than those for subjects in the Negative Feedback treat-
ments. A critical difference technique was utilized to make each of the

28 possible comparisons among means.

31

Table 11. Summary Table of the Three Factor Analysis of Variance
of Post Test Positive -- Negative Scores

 

 

Source of variation if; 115— L

Between
Self Credibility (A) 1 1.01 ( 1.00
Source Credibility (B) 1 .11 (1400
Feedback (C) 1 183.01 118.38*
A/B l .61 ( 1.00
A/C 1 .11 (1.00
B/C 1 1.01 (1.00
A/B/C 1 2.81 1.81

Within error 72 1.55

* = p < .05 two tailed
Table 12. Post Test Positive -- Negative Means for
Subjects in Each of the Eight Treatments

HSLF- HSLF- LSLF- LSLF-
HSOC LSOC HSOC LSOC

 

 

 

 

Positive Feedback 1.70 1.20 1.80 1.70

Negative Feedback 4.20 4.29 4.90 4.50

Table 13 contains a summary of these comparisons.

Table 13 indicates that all possible comparisons between Positive
and Negative Feedback conditions differ significantly in the predicted
direction. Thus, the experimental manipulation of Feedback was completely

successful.

Hypotheses of the Study

The analyses reported below were conducted to test the hypotheses

32

of the study; i.e., those hypotheses specifying the cognitive changes
occurring in each of the eight treatment groups. Since the hypotheses

concerned changes within cells, the t test for related samples was

employed to test differences between pre test and post test means.

Table 13. Critical Difference Test Between Post Test Positive --
Negative Treatment Means

Assigned condition

 

 

 

 

 

 

 

 

 

§p£_ §p9_ §§_ H-H-P H-L-P L-H-P L-L-P H-H-N H-L-N L-H-N L-L—N
H H P ---- -.50 +.lO .00 +2.50* +3.20* +3.20* +2.80*
H L P ---- +.60 +.50 +3.00* +3.70! +3.70* +3.30*
L H P ---- -.10 +2.40* +2.90% +2.90* +2.70%
L L P ---- +2.50% +3.20* +3.20* +2.80*
H H N ---- + .09 +.70 +.30
H L N ---- +.79 +.21
L H N ---- -.40
L L N ----

f: = p < .05 two tailed Critical Difference = 1.18

In all cases where predictions were made concerning direction of
change, one tailed tests of Significance were employed. In those cases

where no change was predicted, two tailed tests of significance were utilized.

Hypothesis 1

 

Under conditions of High Self Credibility and Positive Feedback from
a High Credible Source, no cognitive imbalance will result, and; therefore,
there will be no change in perceived Self Credibility, Source Credibility,

or Task Difficulty.

I'llll Illllllll ‘II!|||'I¢|III|II|'I

 

33

Table 14 contains a summary of the mean differences on each of the
three dependent measures for subjects in the High Self Credibility, High
Source Credibility, Positive Feedback condition. It can be seen that the
data generally fail to support the hypothesis of no change. On only one

Table 14. Pre Test Minus Post Test Mean Differences for Subjects

in the High Self Credibility, High Source Credibility,
and Positive Feedback Condition

 

 

Pre- mean Post- mean D E p_
SLF 3.15* 2.67** +.u8 2.61 (.05-2T
SOC 3.05 1.75 +1.30 10.59 1(.01-2T
TD 3.86 3.80 -.06*** .22 N.S.

*Means are based on a seven point scale, i.e., good = 1, bad = 7.

**A post test mean which is lower than a pre test mean indicates an
increase in the dependent measure.

***Task difficulty signs have been reflected to make them consistent with
the other dependent measures. A plus (+) indicates an increase in task
difficulty while a minus (-) indicates a decrease in task difficulty.

of the dependent variables, Task Difficulty, did no significant change

occur. Perceived Self Credibility and perceived Source Credibility in-
creased significantly on the post test measurement. Thus, contrary to

prediction, it appears that subjects in this condition made a number of

cognitive adjustments as a result of the interaction in which they engaged.

Hypothesis 2
Under conditions of High Self Credibility and Negative Feedback from

a High Credible Source, cognitive imbalance will result. Reduction of this

31+

imbalance will take the form of a decrease in perceived Source Credibility
and an increase in perceived Task Difficulty.
Table 15 contains a summary of the mean differences on each of
the dependent measures for subjects in the High Self Credibility, High
Source Credibility, and Positive Feedback condition. It can be seen that
Table 15. Pre Test Minus Post Test Mean Differences for Subjects

in the High Self Credibility, High Source Credibility,
and Negative Feedback Condition

 

Pre- mean Post- mean D E p_
SLF 3.07 3.u8 -.u1 1.87 N.S.
SOC 3.01 3.30 -029 1028 N.S.
TD 3.76 u.so +.7u n.12 (.01-1T

the data support the predicted change in Task Difficulty. Contrary to

prediction, Source Credibility did not change significantly.

Hypothesis 3

 

Under conditions of High Self Credibility and Positive Feedback
from a Low Credible Source, no cognitive imbalance will result, and;
therefore, there will be no cognitive change in perceived Self Credibility,
perceived Source Credibility, or perceived Task Difficulty.

Table 16 contains a summary of the mean differences on each of the
dependent measures for subjects in the High Self Credibility, Low Source
Credibility, and Positive Feedback condition. It can be seen that the
data fail to support the hypothesis of no change. Perceived Self

Credibility and perceived Source Credibility increased significantly on

35

Table 16. Pre Test Minus Post Test Differences for Subjects in
the High Self Credibility, Low Source Credibility,
and Negative Feedback Condition

 

Pre- mean Post mean D E p
SLC 3.50 2.37 +1.13 u.93 (.01-2T
soc 3.68 1.78 +1.90 9.90 (.01-2T
TD 3.70 2.93 -.77 3.31 (.05-2T

the post test measure. Perceived Task Difficulty decreased significantly
on the post test measure. Thus, contrary to prediction, it appears that
subjects in this condition made a number of cognitive adjustments as a

result of the interaction in which they engaged.

Hypothesis 0

 

Under conditions of High Self Credibility and Negative Feedback from
a Low Credible Source, cognitive imbalance will result. Reduction of this
imbalance will take the form of a decrease in perceived Source Credibility
and an increase in perceived Task Difficulty.

Table 17 contains a summary of the mean differences on each of the
dependent measures for subjects in the High Self Credibility, Low Source
Credibility, and Negative Feedback condition. It can be seen that the
data support the hypothesis. Perceived Source Credibility significantly
decreased on the post test measurement while perceived Task Difficulty
showed a significant increase. Thus, it appears that subjects in this
condition made the predicted cognitive adjustments as a result of the

interaction in which they engaged.

36

Table 17. Pre Test Minus Post Test Mean Differences for Subjects
in the High Self Credibility, Low Source Credibility,
and Positive Feedback Condition

 

Pre- mean Post-mean D ;t_ p_
SLC 3.91 3.95 -.71+ 1.30 N.S.
soc 3.34 14.08 -.714 5.15 (.01-1T
TD 4.00 14.86 +.86 2.1+9 (.05-1T

Hypothesis 5

Under conditions of Low Self Credibility and Positive Feedback from
a High Credible Source, cognitive imbalance will result. Reduction of
this imbalance will take the form of an increase in perceived Self
Credibility and a decrease in perceived Task Difficulty.

Table 18 contains a summary of the mean differences on each of the
dependent measures for subjects in the Low Self Credibility, High Source
Credibility, Positive Feedback condition. It can be seen that the data
generally support the predictions made. Perceived Self Credibility

Table 18. Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, High Source Credibility,
and Positive Feedback Condition

 

Pre-mean Post- mean D E p
SLF 14.05 3.38 +.67 3.69 ( .01-1T
800 2.142 1.90 +.s2 2.57 (.05-2T

TD 4.146 3.56 -.90 1.86 (.05-1T

37

significantly increased on the post test measurement, while perceived

Task Difficulty showed a significant decrease. Contrary to prediction,
perceived Source Credibility increased significantly on the post test
measurement. It appears that, generally as predicted, subjects made a number
of cognitive adjustments as a result of the interaction in which they

engaged.

Hypothesis 6

Under conditions of Low Self Credibility and Negative Feedback from
a High Credible Source, no cognitive imbalance will result, and; therefore,
there will be no cognitive change in perceived Self Credibility, perceived
Source Credibility, or perceived Task Difficulty.

Table 19 contains a summary of the mean differences on each of the
three dependent measures for subjects in the Low Self Credibility, High
Source Credibility, Negative Feedback condition. It can be seen that the

Table 19. Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, High Source Credibility,
and Negative Feedback Condition

 

Pre- mean Post- mean D t p
SLF 3.76 5.31 -1.55 3.80 (.01-2T
soc 2.81 3.66 -.85 2.85 (.05-21‘
TD n.17 14.73 +.56 2.uu (.05-2T

data fail to support the hypothesis of no change. Both perceived Self
Credibility and perceived Source Credibility decreased significantly while
perceived Task Difficulty showed a significant increase. It appears that

this condition created more cognitive imbalance than expected, and thus

llilllll'illl I I‘ll. IIIIIIIIII‘I'!

38

more cognitive change by the subjects who engaged in the interaction.

Hypothesis 7

 

Under conditions of Low Self Credibility and Positive Feedback
from a Low Credible Source, cognitive imbalance will result. Reduction
of this imbalance will take the form of a decrease in perceived Task
Difficulty.

Table 20 contains a summary of the mean differences on each of the
dependent measures for subjects in the Low Self Credibility, Low Source
Credibility, Positive Feedback condition. It can be seen that the data
generally support the prediction of cognitive change. Perceived Task
Difficulty decreased significantly, as predicted. Contrary to predictions,
perceived Source Credibility increased significantly. The predicted
non-significant change in perceived Self Credibility was supported.

Table 20. Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, Low Source Credibility,
and Positive Feedback Condition

 

Pre- mean Post- mean D E p_
SLF 3.79 2.93 +.86 2.19 N.S.
TD 9.33 3.143 -.90 2.614 (.05-1T

Hypothesis 8
Under conditions of Low Self Credibility and Negative Feedback from
a Low Credible Source, no cognitive imbalance will result, and; therefore,

there will be no change in perceived Self Credibility, perceived Source

39

Credibility, or perceived Task Difficulty.
Table 21 contains a summary of the mean differences on each of
the three dependent measures for subjects in the Low Self Credibility,
Low Source Credibility, Negative Feedback condition. It can be seen that
the data generally fail to support the hypothesis of no change. The change
Table 21. Pre Test Minus Post Test Mean Differences for Subjects

in the Low Self Credibility, Low Source Credibility,
and Negative Feedback Condition

 

Pre- mean Post- mean D _t_ p
SLF ”.1” l".59 -.'45 10'4” N.S.
soc 3.29 9.19 -.90 2.26 ( .05-2T
TD n.53 n.80 +.27 .80 N.S.

that did occur took the form of a reduction in perceived Source Credibility.
As predicted there were no changes in perceived Self Credibility or Per-
ceived Task Difficulty.

A grand summary of the differences between the pre test and post
test means for each dependent variable in each treatment is contained
in Table 22. It can be seen that the predictions based on the balance
model were correct for twelve of the twenty-four hypotheses; i.e.,
predictions were correct 50 per cent of the time. Eleven of the twelve
incorrect predictions involve significant changes in the dependent
variables not implied by balance theories. Possible reason for this large

number of unpredicted significant changes will be explored in Chapter IV.

MO

 

 

Table 22. Grand Summary of Pre -- Post Mean Differences
for the Dependent Variables, Self Credibility,
Source Credibility, and Task Difficulty
Assigned Dependent Pre Post
Condition measures Mean Mean D E_ _p_ Predicted
SLF SOC PB
H H P SLF 3.15 2.67 +.98 2.16 (.05-2T NO
SOC 3.05 1.75 +1.30 10.59 (.01-1T NO
TD 3.86 3.80 ‘.06* .22 N.S. YES
H H N SLF 3.07 3.48 -.91 1.87 N.S. YES
SOC 3.01 3.30 -.29 1.28 N.S. NO
TD 3.76 9.50 +.74 9.12 (.01-1T YES
H L P SLF 3.50 2.37 +1.13 9.93 (.01-2T NO
SOC 3.68 1.78 +1.90 9.90 (.01-2T NO
TD 3.70 2.93 -.77 3.31 (.01-1T NO
[‘1 L “I SLF 30.41 3.95 -05” 1030 N080 YES
SOC 3.34 9.08 -.7H 5.15 ‘(.Ol-1T YES
TD 9.00 9.86 +.86 2.99 (.05-1T YES
L H P SLF 9.05 3.38 +.67 3.69 (.01-1T YES
SOC 2.92 1.90 +.52 2.57 (.05-2T NO
TD “0146 3.56 -090 1086 (.05-11‘ YES
L H N SLF 3.76 5.31 -1.55 3.80 (.01-2T NO
SOC 2.81 3.66 -.85 2.95 (.05-2T NO
TD 9.17 9.73 +.56 2.29 (.05-2T NO
L L P SLF 3.79 2.93 +.86 2.19 N.S. YES
SOC 3.99 1.93 +1.56 6.17 (.01-2T NO
TD “.33 3.0.3 -.90 2.64 (.05-1T YES
L L N SLF "’01” 4059 “0'45 lot-“4 N.S. YES
TD “.53 4.80 +.27 '80 N.S. YES
*all signs for task difficulty have been reflected. = increase in TD

and - =

decrease in TD

Additional Analysis

Additional analyses were performed on those measures which were kept

by the experimenter during the interaction.

These measures were obtained

91

to gain further useful information about potential ways in which the

subjects' behavior might be affected by the experimental treatments.

Time taken to complete the description of the puzzle.
For each subject a measure was obtained of the amount of time in
seconds taken to complete the puzzle. Table 23 contains a summary of the

mean number of seconds for subjects in each of the eight treatments.

Table 23. Mean Number of Seconds Taken to Describe the Puzzle

HSLF- HSLF- LSLF- LSLF-

 

 

 

 

HSOC LSOC HSOC LSOC
Positive Feedback 332 279 371 330
Negative Feedback 612 856 692 792

These data were analyzed by means of a three factor analysis of
variance. Table 29 contains a summary of this analysis. The analysis
yielded a significant £_of 96.50 for the Feedback effect.

Table 25 contains a summary of the 28 possible comparisons of
treatment means. It can be seen that all of the comparisons between
Positive and Negative Feedback conditions result in significant differences.
Subjects in the Negative Feedback conditions took longer to describe the
puzzle than did subjects in the Positive Feedback conditions. In addition,
one of the comparisons between Negative Feedback conditions resulted in

a significant difference.

92

Table 29. Summary Table of the Three Factor Analysis of Variance

of Time* Taken to Describe the Puzzle

 

Source of variation .2:; _Jfii_ _£;_

Between
Self credibility (A) l 52 ( 1.00
Source credibility (8) 1 77937 1.22
Feedback (C) 1 2966925 96.50**
A/B 1 21615 < 1.00
A/C 1 37959 < 1.00
B/C 1 238165 3.73
A/B/C 1 30225 < 1.00

Within error 72 63789

* = Time in seconds

9“" = p < .05 two tailed

 

Table 25. Critical Difference Test for Treatment Means on
Time Taken to Describe the Puzzle
Treatment
H H P 332 ---- -053 +039 +002 +280* +529* +310* +910*
H L P 279 ---- +092 +051 +3339 +5776 +363* +963k
H P 371 ---- -091 +291* +985* +2716 +37l*
L P 330 ---- +282* +526* +312* +9l2*
H N 612 ---- +299* +030 +130
L N 856 ---- -219 -119
H N 692 ---- +100
L N 792 ----
"‘ = p < .05 two tailed Critical Difference = 239

93

Choice of stopping time

 

After each subject had explained the puzzle once, she was given
the opportunity to stop or to go over the puzzle again. Since the
resultant data for each subject meet only the assumptions of nominal

2

measurement, a x was computed for each of the independent variables.

The results are found in Tables 26, 27, and 28.

Table 26. Choice of Stppping Time and Self Credibility
Stopped first time StOpped second time

High SLF 20 20

Low SLF 10 30

x2 = 5.33; (p = < .05)

Table 27. Choice of Stopping Time and Source Credibility
StOpped first time Stopped second time

High soc 15 25

Low SOC 13 27

X2 =(l N.S.

Table 28. Choice of Stopping Time and Feedback

Stgpped first time Stopped second time
Positive FB 25 15
Negative FB 5 35

x2 = 21.33; (p =<.01)

99

It is apparent that those subjects under Negative Feedback con-
ditions stopped less often at the first opportunity than did those
subjects under Positive Feedback conditions. In addition, Table 26
indicates that those subjects under Low Self Credibility treatments stOpped
less often at the first opportunity than did subjects under High Self

Credibility treatments.

Number of information questions asked by the subject

For each subject, a measure was taken of the number of information
questions asked of the confederate. Table 29 contains a summary of the
mean number of information questions asked by subjects in each of the

eight treatments.

Table 29. Mean Number of Information Questions Asked by
Subjgots in the Eight Treatments

HSLF- HSLF- LSLF- LSLF-

 

 

 

HSOC LSOC HSOC LSOC
Positive Feedback 1.20 .60 1.10 .80
Negative Feedback 3.50 5.60 5.00 7.60

These data were analyzed by means of a three factor analysis of
variance. Table 30 contains a summary of the analysis. The Feedback
effect yielded a significant F_of 39.63. No other effects were significant.

Table 31 contains a summary of the 28 possible comparisons between
treatment means. All of the 13 significant comparisons indicate that
subjects in the Negative Feedback conditions asked a significantly greater

number of questions than did subjects in the Positive Feedback conditions.

95

Table 30. Summary Table of the Three Factor Analysis of Variance
of Number of Information Questions Asked

 

Source of variance if: __M_S__ _F__

Between
Self credibility (A) l 16.20 1.59
Source credibility (B) l 18.05 1.77
Feedback (0) 1 905.00 39.63='=
A/B 1 .80 (1.00
A/C 1 19.95 1.91
B/C 1 39.20 3.83
A/B/C l .05 ( 1.00

Within error 72 10.22

* = p (.05 two tailed

Table 31. Critical Difference Test for Treatment Means
on Number of Information Questions Asked
Treatment
_s_1_._§ _s_c_>_q pg Mean H-H-P H-L-P L-H-P L-L-P H-H-N H-L-N L-H-N L-L-N

 

 

 

 

 

 

 

H H P 1.20 ---- -.60 -.10 -.90 +2.30 +9.90* +3.80* +6.90
H L P .60 ---- +.50 +.20 +2.90*+5.00* +9.90* +7.00*
L H P 1.10 ---- -.2O +2.90 +9.50* +3.90* +6.50*
L L P .80 ---- +2.70 +9.90* +9.20* +6.80*
H H N 3.50 ---- +2.10 +1.50 +2.10
H L N 5.60 ---- -.60 +2.00
L H N 5.00 ---- +2.60
L L N 7.60 ----

* = p <’.05 two tailed Critical Difference = 2.86

CHAPTER IV

CONCLUSIONS, DISCUSSION, AND IMPLICATIONS FOR FURTHER RESEARCH

All conclusions and generalizations made in this chapter are limited
to those interaction situations which closely parallel the experimental
subjects and conditions. As stated earlier, pretesting data indicated
that different results might be expected under different sex combinations.
The nature of the task, the control of visual feedback, and numberous
other potentially relevant variables also place restrictions on the

generalization of the findings.

Conclusions

 

Across all treatment conditions, it is clear that the feedback a
subject receives from a source does affect her perception of Self Credibility,
the credibility of the source of the feedback and her perception of the
difficulty of the task.

Self Credibility appears to be the most stable of all of the
variables; and, hence, least susceptible to change. Three of the four
significant changes in Self Credibility occurred in Positive Feedback
treatments. This suggests that the subjects in this experiment were more
willing to increase their perceived Self Credibility than to decrease it.

Consistent across all treatment conditions, Source Credibility de-
creased under Negative Feedback and increased under Positive Feedback.
The ambiguity of the situation and the fact that the subjects did not

know the confederates seems to create a situation in which Source

96

97

Credibility is very susceptible to change.

Finally, the subjects' perceptions of the difficulty of the task
change consistently with feedback conditions. Subjects who receive
Negative Feedback perceive the task as being more difficult, while
subjects who receive Positive Feedback perceive the task as being less
difficult.

In this study, predictions were made concerning changes in the
dependent variables as a result of combining various values of the in-
dependent variables. It is necessary, then, to establish some conclusions
involving each of the eight treatment conditions.

(1) It was predicted that in the High Self Credibility, High Source
Credibility, and Positive Feedback condition, no cognitive imbalance
would result. One possible explanation of why changes did in fact occur
when not predicted, is provided below in the discussion section. Subjects
in this treatment increased significantly their ratings of their own Self
Credibility and their ratings of the source providing the feedback. One
might also expect a significant decrease in Task Difficulty; however,
given the experimental conditions there is no reason for that variable
to change. If an individual feels that both participants in the interaction
function effectively at the task, and if she receives no information that
throws doubt on this assumption, then the perceptions of the difficulty
of the task should remain constant.

(2) It was correctly predicted that the High Self Credibility, High
Source Credibility, Negative Feedback condition would produce cognitive
imbalance. This imbalance is reduced by increasing the perception of the

difficulty of the task. It was also predicted that imbalance would be

'98

reduced by decreasing Source Credibility. It appears that the combination
of High Self Credibility and High Source Credibility makes these two
variables resistant to change. It seems reasonable to assume, that,

if an individual perceives that both participants in an interaction
function effectively, the easiest way to restore balance caused by the
negative feedback is to perceive the task as being more difficult rather
than decreasing the high credibility of the source.

(3) It was predicted that in the High Self Credibility, Low Source
Credibility, and Positive Feedback condition no cognitive imbalance would
result. One possible explanation of why changes did in fact occur when
not predicted, is provided below in the discussion section. Subjects in
this treatment increased significantly their ratings of their own Self
Credibility and the credibility of the source providing the feedback. To
balance these cognitive changes, subjects in this treatment decreased
their perception of the difficulty of the task.

(9) It was predicted that in the High Self Credibility, Low Source
Credibility, Negative Feedback condition, cognitive imbalance would result.
The variables changed as predicted. When people receive what they perceive
as Negative Feedback from a Low Credible Source, they would not be ex-
pected to reduce the resulting imbalance by decreasing their Self
Credibility. Subjects in this condition reduced the imbalance by decreasing
significantly the credibility of the source. To balance this change they
increased their perception of the difficulty of the task.

(5) It was predicted that in the Low Self Credibility, High Source

Credibility, Positive Feedback condition, cognitive imbalance would result.

99

After being told that the person was not very good at this task the
resultant imbalance is most easily reduced by increasing significantly
Self Credibility. One possible explanation of the reason for the con-
comitant significant increase~in perceived Source Credibility which was
not predicted, is in a later section of this chapter. To balance the
change in perceived Self Credibility and perceived Source Credibility,
there was a significant decrease in perceived Task Difficulty.

(6) It was predicted that in the Low Self Credibility, High Source
Credibility, Negative Feedback condition, no cognitive imbalance would
result. The prediction of no cognitive change was based on the assumption
that, when an individual who does not perceive herself as being very good
at a task, receives information that does not challenge this assumption
then there is no reason to change her perceptions of herself or the
source of the information. A discussion of why Self Credibility and Source
Credibility did in fact significantly decrease is provided in the discussion
section. To balance these changes, subjects in this treatment significantly
increased their perceptions of the difficulty of the task.

(7) It was predicted that in the Low Self Credibility, Low Source
Credibility, Positive Feedback condition, cognitive imbalance would result.
As predicted, Self Credibility did not change significantly. Not predicted
was a significant increase in Source Credibility. To balance the sig-
nificant increase in Source Credibility, subjects in this treatment
significantly decreased their perceptions of the difficulty of the task,
as predicted.

(8) It was predicted that in the Low Self Credibility, Low Source

50

Credibility, Negative Feedback condition, no cognitive imbalance would
result. Not predicted was a significant decrease in the ratings of the
source of the feedback. One possible explanation of why this change
occurred is provided below in the discussion section. With the sig-
nificant decrease in perceived Source Credibility one might also expect
a significant increase in the perception of the difficulty of the task.
However, given the experimental conditions, there is no reason for that
variable to change. If an individual feels that both participants in
the interaction do not function very effectively at the task, and if she
receives no information that throws doubt on this assumption, then per-
ceptions of the difficulty of the task should remain constant.

In addition to the conclusions relating to the eight hypotheses of
interest in the study, additional data were gathered by the experimenter
during the observation of the interaction between the subject and the con-
federate. Following is a summary of the conclusions relevant to these ob-
servations.

(1) Subjects exposed to Negative Feedback generally took longer to
describe the puzzle than did subjects exposed to Positive Feedback. Such
instances of Negative Feedback served to indicate to the subject that her
messages were ambiguous. As a result, subjects in the Negative Feedback
conditions engaged in a greater number of communication attempts aimed at
revising their instructions so that they would be clear to the confederate.
It is these increased number of attempts at clarification that largely
account for the observed time differences. This finding provides further

support for the previously mentioned research (8)(19)(25)(26) concerning

51

the effects of negative feedback on observable encoding behavior. As
Newcomb (17) suggests, changes caused by an imbalanced state can be
either behavioral, cognitive, or both. The emphasis on cognitive change
in this study made it necessary to control attempts by the subject to
clarify the situation by having the confederate reply to interrogative
questions with a noncommital answer. In other words, the attempts of the
subject to reduce the imbalance by increasing the volume of communication
directed toward the confederate were rejected by the confederate.

(2) Subjects exposed to Negative Feedback asked a significantly
greater number of questions of the confederate than did subjects exposed
to Positive Feedback. One explanation for this finding is contained in
the immediately preceding discussion of differences in time Spent
describing the puzzle. Most of the communication attempts to increase
clarity were couched in the form of interrogative statements. This fact
accounts for the significant differences observed between subjects in
Positive and Negative Feedback conditions.

(3) Subjects exposed to Negative Feedback chose to describe the
puzzle a second time significantly more frequently than did subjects
exposed to Positive Feedback. This finding is somewhat more surprising
than those summarized above. It might be expected that subjects receiving
Negative Feedback would seize upon the first available opportunity to
withdraw from a potentially threatening and anxiety-arousing situation,
while those subjects receiving Positive Feedback would choose to maintain
the existing pleasant state of psychological affairs for as long as

possible. Such, however, was not the case. One possible explanation for

52

the willingness of subjects in the Negative Feedback condition to con-
tinue work on the puzzle may have been the student-instructor relationship
existing between the experimenter and themselves. These subjects may

have been driven by a desire to do as well as possible, in order to
maintain positive relationships with the experimenter.

(9) Subjects assigned to conditions of Low Self Credibility chose
to describe the puzzle a second time significantly more frequently than
did subjects assigned to conditions of High Self Credibility. On the
surface, this result may seem unexpected; however, it should be remembered
that half of these subjects were in conditions in which they received
Positive Feedback from the confederate. It may well be that a number of
these subjects chose to continue because they desired additional rein-
forcing information to offset their original perceptions of Low Self

Credibility.

Discussion
The rationale of this study developed the premise that in an

interaction situation involving task accomplishment, an individual has a
number of cognitive clusters which are relevant to the situation. These
cognitive clusters were conceptually defined as: Self Credibility, or the
perceptions a person has concerning his own ability to engage in the
interaction; Source Credibility, or the perceptions a person has con-
cerning the abilities of other participants to engage in the interaction;
and, Task Difficulty, or the perceptions a person has concerning the

complexity and difficulty of the task in which the person is engaged.

53

The rationale further stated that when an individual receives feedback
which is not consistent with the cognitive elements he is placed in a
state of cognitive imbalance. In order to restore cognitive balance it
is necessary for the individual to make changes in the cognitive elements.
A number of alternative ways to change the cognitive elements are
available to the individual. The value of this research study has been
to identify the specific changes occurring in these cognitive elements
under varying combinations of Self Credibility, Source Credibility and
Feedback.

It is clear from the results of this study that there are some
limitations on the predictive ability of the balance model in this
situation. Of the 29 predictions made in this study, 12 were correct.

Of the 12 incorrect predictions, eleven involved significant changes
when none had been predicted. An inSpection of the data and the ex-
perimental procedures suggest some possible explanations for the failure
to predict these changes.

First, comments made by the subjects after the experiment indicate
that the situation was very ambiguous. Prior experience in a situation
like this was, as far as could be determined, nonexistent. The only
indication that the subjects had concerning their own and the confederates'
competencies was provided by the experimenter. The perceptions of Self
Credibility and Source Credibility, then, were based on a single learning
trial. It appears that this single learning trial was not sufficient
to provide for the unequivocal manipulation of the independent variables.

The data in Table 22 indicate that while the experimental manipulations

59

were sufficient to produce significant differences between high and low
conditions, all of these differences lie on the positive side of the
seven point scale used to measure the variables. The range of pre test
means for Self Credibility and Source Credibility suggest that the ex-
perimental differences were created, not between high and low, but
between neutral and high. In particular, the experimental manipulation
of Self Credibility was not as successful as would be ideally desired.
Although the means for High Self Credibility conditions are consistently
larger, only three of the sixteen comparisons between High and Low Self
Credibility conditions are statistically significant.

It is suggested, then, that the experimental conditions served as
additional positive or negative reinforcers to the single learning trial
provided by the experimenter. For example, subjects assigned to High Self
Credibility, High Source Credibility, and Positive Feedback, had pre test
Self Credibility and Source Credibility means of 3.15 and 3.07 respectively.
The resultant Positive Feedback produced post test means of 2.67 and 1.75
respectively. The differences in pre test and post means are significant.
However, had the experimental procedures produced pre test means further
from the midpoint of the scale the post test means may not have been
significant. This at least suggests that the elements of a situation can
be described as "balanced" only after the values of these elements have been
firmly established through a large number of experiences of learning
trials. Thus, one potential limitation of balance models may be that
they are powerful predictors only in situations which have had a long period

of time to stabilize.

55

It appears, then, that the changes in the cognitive elements within
each individual treatment condition can be explained in part within a
balance theory framework. There is a further implication that the impetus
to produce the imbalanced condition, which in turn creates the need for
cognitive readjustment, rests in part on principles of learning theory.
The significant changes, particularly in those conditions in which no
cognitive change was predicted, resulted from the reinforcing or non-
reinforcing nature of the feedback in relation to the prior perceptions
the subjects had developed from the information given them by the ex-
perimenter. The predictions the experimenter made concerning cognitive
change were based on the assumption of rigorous experimental procedures
to produce appropriate experimental conditions. The apparent limitations
of the experimental procedures provide some suggestions for further re-
search, discussed in a later section of this chapter.

The predictions proposed in the rationale of this study were
tentative, since the theoretic model used to generate the predictions did
not incorporate a concept which could be equated with the variable Self
Credibility, and since the model provides no means for predicting Specific
changes that will occur on each of the variables. A value of this study
has been to provide information on various specific cognitive changes
resulting from particular kinds of interaction situations. This study
further demonstrates the inherent complexity of even the most simple appearing
interaction situations, and fits well with a process philosophy view of
communication. Verbal interaction produces complex and varied cognitive

changes in relevant Self, Source, and Task variables.

56

Undoubtly, other variables also affect and are affected by the
interaction. For example, this study attempted to provide some low
level generalizations concerning the effect of a completely positive
or negative feedback condition. The experimenter recognizes that is is
highly unlikely that any interaction situation will contain information
that is perceived as completely positive or completely negative. It was
assumed that if differences were not significant when dealing with the
ends of a hypothetical, postive---negative, continuum, it would be un-
likely that differences would exist under combinations of positive or
negative statements.

The concern for the further analysis of the variables in this study
and other variables not used in this study, suggests some directions for
continued research effort, specifically:

l. A study designed to allow the subject to gain more ex-
perience in the task. Additional experience before under-
going the experimental treatment may establish more stable
conditions of Self Credibility.

2. A study in which the subject is allowed to establish her
own level of Self Credibility. Methodologically, this
could be accomplished by asking the subject, after a
training period, how good she thinks whe will be at this
task. Her remarks would then be used to place her in a
particular experimental condition. This would control to
some degree the problem of the experimenter randomly

assigning subjects to experimental treatments which were

5.

57

not commensurate with their own perceptions of their ability.

A series of studies in which the major independent variable
would consist of manipulations within the feedback schedule.
This series of studies would attempt to provide generalizations
in regard to the relative weights of combinations of positive

and negative statements.

Studies dealing with the ways in which grades and grading
procedures function as feedback mechanisms for the receiver

of the grades may provide further knowledge of the communi-
cation processes involved in the student-teacher interaction.
A recent review (29) suggests that this problem area has

not been extensively researched. One study (2) has established
a relationship between the grade received on an essay and
attitude change concerning the topic the essay was written
about. Another recent study (22) has classified various

forms of social power used by teachers to influence students.
These forms of social power were compared with the students'
grade aspirations and the students' perceived level of
capacity. The study suggests that the conceptualization of
feedback utilized in this study is rather broad and that feed-
back may take on more values than positive or negative,

depending on the level of analysis.

This study established that the qualification factor of the
previously mentioned factor analytic study (12), is useful

as a measuring instrument to measure the concept Self

58

Credibility, at least within the content of an interaction
situation involving a task. An investigation of the effect
of feedback using the other two factors, safety and
dynamism, along with the qualification factor as measuring
instruments may produce significant differences in the

measurement potential of each factor in this task situation.

This study involved only two individuals and a simple
task. A series of studies involving more than two people
and different task situations will serve to further clarify

the effect of feedback in interaction situations.

l.

10.

ll.

12.

13.

REFERENCES

Bettinghaus, E. Communication theory and the use of the new media.
Paper presented at the Conference on Theory for the New Media,
Michigan State University, 1962.

Bostrom, R. N., Vlandis, J. w., 6 Rosenbaum, M. E. Grades as rein—
forcing contingencies and attitude change. Journal of Educational
Psychology, 1961, 52) 112-115.

 

 

Brown, R. Models of attitude change. In New directions in psycholggz.
New York: Holt Rinehart and Winston, 1962. pp. 3-85.

Cohen, A. R. The effects of individual self esteem and situational
structure on threat ordered reactions to power. Unpublished doctoral
dissertation, University of Michigan, 1953.

Combs, A. W. (Chairman) Perceiving, behaving, becoming. Yearbook
Association for Supervision and Curriculum Develgpment, 1962, National
Education Association, Washington, D.C.

Festinger, L. Cognitive dissonance. Scientific American, 207,
93-100.

 

Festinger, L. A theory of cognitive dissonance. Stanford, California:
Stanford University Press, 1957.

 

Greenberg, B., 6 Tannenbaum, P. H. Communicator performance under
cognitive stress. Journalism Quarterly, 1962, 39(2), 168-178.

 

Heider, F. The psychology of interpersonal relations. New York:
John Wiley and Sons, 1958.

Johnson, F. G., 8 Klara, G. R. Feedback: principles and analogies.
Journal of Communication, 1962, 12, 150-159.

Leavitt, H. J., 8 Mueller, R. A. Some effects of feedback on
communication. Human Relations, 1951, 9, 901-910.

 

Lemert, J. Dimensions of source credibility. Paper read at
Association for Education in Journalism, August, 1963.

Miller, G. A., Galenter, E., 8 Pribram, K. H. Plans and the structure
of behavior. New York: Henry Holt, 1960.

 

 

59

19.

15.

16.

17.

18.

19.

20.

21.

22.

23.

29.

26.

27.

60

Miller, G. R. Variations in the verbal behavior of a second
speaker as a function of varying audience response. Speech
Monographs, 1969, 31, 109-116.

 

Mischel, W. The effect of the commitment situation on the
generalization of expectancies. Journal of Personality, 1958, 22,
508-5160

 

Murphy, G. Toward a field theory of communications. Journal of
Communication, 1961, 11, 196-201.

 

 

Newcomb, T. An approach to the study of communication acts.
Psychological Review, 1953, 60, 393-909.

 

Osgood, C. E. Cognitive dynamics in human affairs. Public Opinion
Quarterly, 1960, 23, 391-365.

 

Osgood, C. E. 8 Tannenbaum, P. H. The principle of congruity in
the prediction of attitude change. Psychological Review, 1955,
223 ”2.55.

 

Powers, W. T., Clark, R. K., 8 McFarland, R. L. A general feedback
theory of human behavior: part II. Perceptual and Motor Skills,
1960, _1_1_, 309-323.

 

Rosenbaum, M. E., 8 deCharms, R. Direct and vicarious reduction
of hostility. Journal of Abnormal and Social Psychology, 1960,
.11, 309-323

Rosenfeld, H., 8 Zander, A. The influence of teachers on aspirations
of students. Journal of Educational Psychology, 1961, p29 1-11.

Salzinger, K. Experimental manipulation of verbal behavior; a review.
Journal of General Psychology, 1959, Ei) 65-99.

 

Smith, Ann Z., 8 Dobbin, J. E., Marks and marking systems. In C. H.
Harris (Ed.), Encyplopedia of educational research. New York:
Macmillan, 1960. pp. 783-789.

 

Stassi, E. J. Disfluency of normal Speakers and reinforcement.
Journal of Speech and Hearinngesearch, 1961, 9, 358-361.

 

Stolz, W. Some effects of cognitive feedback on encoding behavior.
Unpublished masters thesis, University of Wisconsin, 1962.

Stotland, E., 8 Cottrell, N. B. Self esteem, group interaction,
and group influence on performance. Journal of Personality, 29,
273-289.

 

61

28. Wiener, N. Cybernetics. New York: John Wiley, 1998.

 

29. Wiener, N. The human use of human beings. New York: Doubleday, 1956.

30. Wylie, Ruth. The self concept. Lincoln: University of Nebraska Press,
1961.

APPENDIX A

Information Flow Study

Department of Communication

Michigan State University

This research study is one of a series concerned with the flow of in-

formation in communication situations.

All instructions on how to proceed will be given to you either in this
booklet or by the researcher. Please do not turn ahead to any part of
the booklet, unless you are told to do so either in the booklet or by the
researcher. Also, once you have completed any part of the booklet please

do not turn back.

You will be called by the letter which is visible to you at eye level

on the partition.

At times you will be asked to pass information through to the researcher.

Please wait until this information is requested before passing it.

Do not be concerned with the numbers on the right side of some of the
pages. These numbers are there for the purpose of coding the information

to IBM cards.

Do you have any questions at this point? If you do, please ask them now.

If you have no questions, please say NO QUESTIONS

 

DO NOT TURN THIS PAGE UNTIL TOLD TO DO SO

63

I Would you please complete the following information:
1. Class designation: (Check one)
Freshman __ Sophomore __ Junior __ Senior __ Grad.___ (l9) __
2. Sex: Male __ Female __ (15) __

3. College enrolled in: (Check one)

 

 

 

 

 

 

 

 

Agriculture Engineering
Arts 8 Letters Home Economics
Business Natural Science
Communication Arts _____ Social Science
Education University
Other (Specify) (16-17)____

 

When you are ready to begin say READY

DO NOT TURN THIS PAGE UNTIL TOLD TO DO SO

69

This part of the study will involve the completion of a task similar to
the one you have been practicing on. For this particular part:
Subject A will--
1. receive a card with the position of the seven
figures on it
2. have the task of describing the figure so that
Subject B can reproduce it
Subject B will--
1. have the task of actually putting together the
figure
2. provide the information to subject A by responding

verbally

You will have as much time to work at this task as you wish. The task
will not be stopped until either Subjects A or B wish to stop working.
Before proceeding with the task we would like some further information
from you. At the top of each of the following three pages you will find
a statement listed above a set of seven-point scales. Please indicate
your reaction to the statement by placing a check mark on each scale.

For example, here is a statement and a single scale:

Classes at M.S.U.

Good : : : : Bad

 

If you feel that classes at M.S.U. are in general extremely good, you

would place a check mark in the Space closest to the word good.

TURN THE PAGE

 

65

In general, consider the positions on the scale to represent the following
judgments:
Good:

extremely good

quite good

slightly good

neither good or bad,

"I don't know"

"This scale doesn't apply to the question"

slightly bad

quite bad

extremely bad

Bad:

Be sure to put one check mark, and only one1_a1opg_each scale.

Do not omit any scales.

 

Go ahead and complete the next three pages. When you are finished with

the three pages, say READY

TURN THE PAGE AND COMPLETE THE NEXT THREE PAGES

66

Statement:

Subject B, the person cooperating with you, in relation to the

task about to be performed

 

 

 

Reasonable : : : : : ___Unreasonable
Unsuccessful;__}___y : : : : Successful
Expert__:____:__:__:___:__:__I gnorant
Crude___:__;:___:____:__:__:___Gracious
Unfriendly;__3____i___: : : : Friendly
Inefficient__:__:___:___:__:__:___Efficient
Admirable_:____:__:__:___:__:___Contemptable
Patient___:__:___:____:_:__:__lmpatient
Incompetent____:___:___:__:__:__:__Competent
Know ledgeable___:___:__:_:__:__:__Unknowledgeable
Closedminded___:___:__:__:__:____:____0penminded
Capable_:__:__:__:__:__:__ln capable
J us t__:_:__:__:___:___:__Un 3' us t
Confident : : : : : : Unsure

TURN THE PAGE

(32)__
(33)___
(39)___
(35)___
(36)__
(37)____
(38)___
(39)__
(90)—
(91)___
(92)__
(93)__
(99)—

(95)

Statement:

Myself, in relation

Just

to the task

about to be performed

Unjust

 

Incapable

Capable

 

Competent

Incompetent

 

Admirable

Successful
Reasonable
Ignorant

Unfriendly :

Efficient

Unsuccessful

 

 

Unreasonable

Expert

Friendly

 

Inefficient

Patient

 

Unknowledgeable
Closedminded

Confident

TURN THE PAGE

 

Impatient
Knowledgeable
: Openminded

Unsure

(18{____
(19)____
(20)__
(21)__
(22)___
(23)___
(2a)__
(25)___
(26)__
(27)____
(28)_____
(29)____
(30)___

(31)

 

lit-"I

68

Respond to the following scales in relation to how you perceive the

cooperative task you are about to engage in:

 

 

meaningful : : : : : : not meaningful (96)_____
easy___:___: : : :__:___difficult (97 )_
fast__:__:_:___:__:__:__s low (98)

co lorful___:___:_:__:___:___:___dull (99 )_
complex____:__:__:__:___:_:_s imple (50)
clear__:__:__:___:___:__:_unclear (51)

Respond to the following scales in relation to the information you expect

 

to receive from your partner

 

helpful__:__:_:__:__:__:__harmful (52 )_
unc1ear__:___:__:__:__:__:_c1ear ( 53)
sensible___:__:__:__:__:_:___not sensible (59)___
negative__:___:__:___:___:___:__positive ( 55 )_
good : : : : : : bad (56)____
remote___:___:__:____:___:__:___intimate (57)___

DO NOT TURN THE PAGE

WHEN YOU FINISH THIS PAGE1 SAY READY

Statement:

Myself, in relation

Unsure

to the task

just completed

Confident

 

Openminded

Closedminded

 

Knowledgeable

O.

 

Impatient

Inefficient

Friendly

Unknowledgeable

Patient

 

 

Efficient

Unfriendly

 

Expert

Ignorant

Unreasonable

Unsuccessful

O
.

Reasonable

 

Successful

 

Gracious

0.

Crude

 

Contemptable

O.

 

Incompetent

Capable

Admirable

 

Competent

 

 

Unjust

Incapable

Just

 

TURN THE PAGE

 

(18{____
(19)___
(20)__
(21)___
(22)___
(23{____
(29)___
(25)____
(26)____
(27)___
(28)___
(29)____
(30)___

(31)

 

70

Statement:
Subject B, the person cooperating with you, in relation to

the task just completed

 

 

 

 

Unsure : : : : : : Confident (32)_____
Unjust : : : : : : Just (33l____
Incapable___: : : : : : Capable (39)____
Openminded___g___: : : : : Closedminded (35{____
Unknowledgeab1e___:____:___:____:____:___:__Knowledgeable (36 )_
Competent__:__:__:___:__:__:__Incompetent ( 37)
Impatient___:_:__:____:__:___:__Patient (38)___
Contemptable__:___:___:__:___:___:___Admirable (39 )_
Efficient____:____:___:___:__:___:___Inefficient (90)
Friendly___:____:___:____:____:___:__Unfriendly (9l)_____
Gracious___:____:__:___:___:___:___Crude (1+2)
Ignorant___:___:__:__:___:___:__EXpert (93 )_
Successful___:____:__:__:____:__:__Unsuccessful (99)____
Unre as onable___:__:___: : : : Reas on ab 1e ( 95)

 

TURN THE PAGE

 

71

Respond to the following scales in relation to the information you

 

received from your partner

 

 

 

 

 

Intimate : : : : : : Remote (52 )—
Bad___:__: : : :___:____Good (53)__

Positive : : : : : :__Negative (59)____

Not Sens ible__:___:__:__:_:__:__Sensib 1e (55 )_
Clear : : : : : : Unclear (56)____
Harmful__:___:__: : : : Helpful (57)_____

Respond to the following scales in relation to how you perceive the

cooperative task you have just engaged in

Unclear__:___:__:__:__:___:___Clear (96 )_
Simple__:___:___:__:__:____:__Comp1ex ( 97 )_
Dull__:__:__:__:___:__:__Colorful ( 98)____
Slow____:_:___:__:____:___:__Fast (‘49 )_
Difficult__:__:____:____:__:_____:____Easy (50 )_

Not Meaningful_:___:___:__:__:___:__Meaningful (Sl)___

TURN THE PAGE

 

WHEN YOU ARE FINISHED WITH THIS PAGE, SAY FINISHED

72

Reinforcement schedule-—POSITIVE
Respond to all questions with a positive answer--O.K.,

 

 

 

To any Opening statement Fine, Let's go on
After the lst piece Good, Let's go on
After 2nd piece It was a good idea to start

with the pieces

 

After the 3rd piece I'm sure you have explained these
correctly, but could we go over

the first three briefly so we

 

 

 

can check
After reviewing the 3 pieces O.K., Let's go on
After the 9th piece Good
After the 5th piece I've got it. I think I see

where the last two pieces fit

 

After the 6th piece You're doing fine, I think we're

going to have this right

 

After the 7th piece I would have gone about this the
same way you did, anyhow you can
go over it again briefly or we

can stop, you decide

 

After the ending statement Fine - I think we did all right

73

Reinforcement schedule--Negative

Try to keep on schedule

01-5me
0

Try to avoid direct answers except when necessary

Use a tone of voice which is firm and strong
Try to make the statements sound natural
Move a piece or two together once in awhile

 

To any opening statement

That's not clear. Will you

go over it again

 

After lst piece

I'm not sure this is right, but

let's go on

 

After 2nd piece

After 3rd piece

It seems to me that it would have
been better to start with the
pieces (choose pieces other than

those the subject started with)
I'm not sure this is clear, go

back over the first 3 so I can check

 

After reviewing the first 3

I'm not sure, but let's go ahead'

 

After the 9th piece

Uh-huh

 

After 5th piece

It doesn't look to me like the

last two pieces left will fit

 

After the 6th piece

Boy,"if we get this right it'll

be a miracle

 

After the 7th piece

I would have done this a different
way anyhow, you can go over it
again briefly or we can stop now,

you decide

 

After the last statement

I don't think we did very well

79

Practice Puzzles

 

 

"TILITIIIINIITII