University of Southern California Dissertations and Theses

Dissonance and Self-Perception Analysis of "Forced Compliance": When Two Theories Make Competing Predictions

(USC Thesis Other)

Dissonance and Self-Perception Analysis of "Forced Compliance": When Two Theories Make Competing Predictions

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DISSONANCE AND SELF - PERCEPT ION ANALYSES OF
"FORCED COMPLIANCE": WHEN TWO THEORIES
MAKE COMPETING PREDICTIONS
by '
Duane Green
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
August 1972
INFORMATION TO USERS
This material was produced from a microfilm copy of the original document. While
the m ost advanced technological means to photograph and reproduce this document
have been used, the quality is heavily dependent upon the quality of the original
submitted.
The following explanation of techniques is provided to help you understand
markings or patterns which may appear on this reproduction.
1. The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the missing
page(s) or section, they are spliced into the film along with adjacent pages.
This may have necessitated cutting thru an image and duplicating adjacent
pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black mark, it
is an indication that the photographer suspected that the copy may have
moved during exposure and thus cause a blurred image. You will find a
good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being
photographed the photographer followed a definite method in
"sectioning" the material. It is customary to begin photoing at the upper
left hand corner of a large sheet and to continue photoing from left to
right in equal sections with a small overlap. If necessary, sectioning is
continued again — beginning below the first row and continuing on until
complete.
4. The majority of users indicate that the textual content is of greatest value,
however, a somewhat higher quality reproduction could be made from
"photographs" if essential to the understanding of the dissertation. Silver
prints of "photographs" may be ordered at additional charge by writing
the Order Department, giving the catalog number, title, author and
specific pages you wish reproduced.
5. PLEASE NOTE: Some pages may have indistinct print. Filmed as
received.
Xerox University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 48106
73-31,349
GREEN, Duane, 1945-
DISSONANCE AND SELF-PERCEPTION ANALYSES OF
"FORCED COMPLIANCE": WHEN TWO THEORIES
MAKE COMPETING PREDICTIONS.
University of Southern California, Ph.D.,
1972
Psychology, experimental
University Microfilms, A XEROX Company, Ann Arbor, Michigan
UNIVERSITY O F SOUTHERN CALIFORNIA
T H E GRADUATE SC H O O L
UNIVERSITY PA R K
LOS A N G ELES, CA LIFO RN IA 9 0 0 0 7
This dissertation, written by
Duane Green
under the direction of h..La. Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
DOCTOR OF P H IL O SO P H Y
Dean
Dafe September 1972
DISSERTATIO: 'Mmitte;
Chat
ACKNOWLEDGMENT
The advice and assistance of Dr. Edward S. Conolley
the dissertation chairman, and of Dr. Norman Miller,
Dr. A. Steven Frankel, and Dr. Vern L. Bengtson, the
dissertation committee members, is gratefully acknowl
edged.
ABSTRACT OP THE DISSERTATION
Dissonance and Self-Perception Analyses of
"Forced Compliance": When Two Theories
Make Competing Predictions
by
Duane Green
Doctor of Philosophy, Graduate Program in Psychology
University of Southern California, August, 1972
Professor Edward S. Conolley, Chairman
Festinger's (1953* 1957) analysis of "forced com
pliance" and his study supporting that analysis (Festinger
and Carlsmith, 1959) have resulted in one of the most pro
vocative issues surrounding dissonance theory. One contro
versy has centered around the relationship between the
magnitude of an incentive used to induce a person to en
gage in behavior discrepant from some particular private
opinion and the amount of subsequent change in that opin
ion. Dissonance theory postulates an inverse relationship
between the. magnitude of the incentive used to induce com
pliance and post-commitment attitude change. In some
studies post-compliance attitude change has shown a nega
tive relationship to the amount of incentive offered for
ill
compliance (Cohen, 1962; Festinger and Carlsmith, 1959).
In other studies, however, a positive relationship (Rosen
berg, 1965) or lack of relationship (Janis and Gilmore,
1965) was found. Extensive discussions of the dissonance-
incentive controversy have been presented elsewhere (e.g.,
Aronson, 1966, 1968; Elms, 1967; McGuire, 1966; Rosenberg,
1965* 1966; Zajonc, 1968) and since a consistency model
which appears to settle this controversy is presented else
where (Conolley, 1 9 7 0 ; Gerard, Conolley, and Wilhelmy, in
press), it will not be considered here.
A second controversy focuses on an alternative ex
planation of "forced compliance" phenomena suggested by
Bern (1965, 1967b). His position often appears to result
in the same empirical predictions as those made from dis
sonance theory. However, since Bern's interpretation
avoids the use of any motivational concept, the two
theories offer extremely different explanations of the
observed phenomena in the "forced compliance" situation.
Bern postulated that it is only necessary to regard the
subject as an observer who Infers his attitudes on the
basis of external evidence available to anyone who may
have witnessed that behavior. Thus, if the subject per
forms a behavior that is relevant to his attitude and does
not perceive any externally controlling events, he will
use his behavior as the basis for inferring his attitude.
i v
"Interpersonal simulations," the method used to test
Bern’s analysis, and the results obtained using this para- :
digm have aroused considerable controversy (e.g., Bern,
1967a, 1967b; Bern and McConnell, 1970; Harvey and Mills,
1971; Jones, Linder, Kiesler, Zanna, and Brehm, 1 9 6 8).
Although the issue is still debated, it appears that self- ;
perception theory holds that the extremity of initial
attitude is of relatively little phenomenological impor
tance to the subject in a forced compliance situation
(Bern and McConnell, I970). If this is the case, differ
ences in the extremity of initial attitudes would be of
little significance in determining final attitudes. This
is directly opposite the prediction derived fromi.a dis
sonance analysis. Dissonance theory would hypothesize
greater attitude change when the initial attitude was
extreme. Increasing the extremity, and therefore the
discrepancy, heightens support for noncompliance and
would result in more dissonance and greater subsequent
attitude change.
The purpose of this dissertation was to investigate
which of these competing theories best accounts for the
observed phenomena in the "forced compliance" situation.
Therefore, an experiment was designed in which the ex
tremity of the initial attitude and the justification
for commitment to the discrepant act were manipulated.
The study provided strong support for the dissonance
prediction that increasing the extremity of the initial
attitude would enhance subsequent attitude change. Other >
dissonance predictions were supported as well. The self- I
perception predictions received little support.
These results suggest that the dissonance analysis
is most appropriate for explaining the effects of incen
tives and attitude-discrepancy manipulations in the
"forced compliance" paradigm. The finding that increased
discrepancy resulted in increased attitude change is ex
tremely difficult to explain from a self-perception anal
ysis. Also, another study was proposed to: further test
the relative merits of the competing predictions of the
two theories. The conclusion was reached that explana
tions provided by self-perception theory appear to have
little value in the "forced compliance" setting.
v i
TABLE OF CONTENTS
Page
LIST OF TABLES...................................... ix
LIST OF FIGURES.................................... xi
Chapter-
I. INTRODUCTION.......................... 1
II. METHOD .................................... 20
Design Overview ......................... 20
Procedure........................... 21
Subjects............................... 21
Preparation....................... 21
Thirst arousal and precommitment measures 22
Perceptual task.......................... 24
Counterattitudinal commitment..... 24
Control against false compliance .... 26
"Control" groups......................... 27
Postmeasures............................. 27
III. RESULTS .................................. 29
IV. DISCUSSION ........................ ..... 51
REFERENCES ............................... 67
APPENDIX A ............ 74
APPENDIX B . . . ............................. 95
vii.
Chapter Page
APPENDIX C . .............. 97
APPENDIX D ..................................... 103
vlii
LIST OP TABLES
Table
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Page
Correlations Between the First Two Thirst
Scales................................... • 31
Mean Ratings of Thirst on the First Bipolar
Adjective S c a l e ............ 32
Mean Ratings of Thirst on the First Adjective
Checklist Scale ........................... 33
Analysis of Variance of Mean Ratings of
Thirst on the First Bipolar Adjective Scale 34
Analysis of Variance of Mean Ratings of
Thirst on the First Adjective Checklist
Scale .................................... 33
Contrast Analysis of Mean Thirst Ratings on the
First Bipolar Adjective Scale ............. 37
Contrast Analysis of Mean Thirst Ratings on the
First Adjective Checklist Scale ........ 38
Correlations Between the Two Change Scores and
the Behavioral Measure of Thirst............ 39
Mean Change in Thirst Ratings on the Bipolar
Adjective S c a l e................ 4l
Analysis of Variance of Mean Change in Thirst
Ratings on the Bipolar Adjective Scale . . . 42
i x
Table Page j
11. Contrast Analysis of Mean Change in Thirst j
Ratings on the Bipolar Adjective Scale . . . 43 ;
12. Mean Change In Thirst Ratings on the Adjective
Checklist 44 ;
13. Analysis of Variance of Mean Change in Thirst
Ratings on the Adjective Checklist 45 J
14. Contrast Analysis of Mean Change in Thirst
Ratings on the Adjective Checklist ......... 46
1 5. Mean Number of Milliliters of Residual Water
Following Consumption 48 I
16. Analysis of Variance of the Mean Number of
Milliliters of Residual Water Following
Consumption 49 I
17. Contrast Analysis of the Mean Number of Milli
liters of Residual Water Following Consump
tion 50 :
x
I
LIST OP FIGURES
Figure Page ;
1. Theoretical Predictions Derived from Dis-
t
sonance Theory.............. 16
2. Theoretical Predictions Derived from Self-
Perception Theory 18 ,
3. Mean Change in Thirst on the Bipolar Adjec
tive Scale for the Two Levels of Thirst
and the Two Levels of Incentive........... 5^
4. Mean Change in Thirst on the Adjective Check
list Scale for the Two Levels of Thirst and
the Two Levels of Incentive 55 I
5. Mean Milliliters of Residual Water Following
Consumption for the Two Levels of Thirst
and the Two Levels of Incentive 56 ;
x i
CHAPTER I
INTRODUCTION
Festinger’s (1953j 1957) analysis of "forced com
pliance" has resulted in one of the most controversial
Issues surrounding the theory of cognitive dissonance. He !
postulated that performing an act which was discrepant
from one’s private opinion would result in dissonance
since the knowledge of having performed the act would be
dissonant with the knowledge that one believes his opinions
to be true. In this situation the magnitude of dissonance
is assumed to be proportional to the ratio of dissonant
and consonant cognitions weighted by their Importance.
All those cognitive elements which support the compliant
behavior, e.g., promised reward or threatened punishment,
are consonant with the act of compliance. Therefore, the
amount of reward or punishment used to induce compliance
is an important determinant of the magnitude of dissonance.
Dissonance magnitude is assumed to vary Inversely with the
amount of reward or punishment; large rewards or punish
ments result in small magnitudes of dissonance. It
follows, then, that the "maximum possible dissonance would
be created if the reward or punishment was just barely
enough to elicit the desired overt behavior or expression
(Pestinger, 1957* P. 91)."
In the forced compliance situation the overt behav- ;
ior is normally irrevocable. Therefore, unless the person
can distort or deny his act of compliance, dissonance is
reduced by changing one's private opinion to bring it in
line with the compliant behavior. Since attitude change I
is postulated to be directly related to the magnitude of
dissonance aroused, the greatest amount of attitude change I
should occur when the external justification is just barely
sufficient to obtain compliance. Thus, "if one wanted to
obtain private opinion change in addition to mere public i
compliance, the best way to do this would be to offer just :
enough reward or punishment to elicit overt compliance.
If the reward or threat were too strong, only little dis
sonance would be created and one would not expect private
change to follow as often (Pestinger, 1 9 5 7 , p. 9 5 ) . "
An experiment by Pestinger and Carlsmith (1 9 5 9 ) was |
designed to test the predicted effects of monetary justi-
}
fication for compliance on attitudes toward a dull task.
In the low reward condition the subject was first required j
to perform boring and repetitive tasks and then was in- !
duced, by the offer of $1, to lie to a waiting student by ;
telling her that the boring tasks were really enjoyable j
i
and interesting. In the high reward condition the subjects1
i
were paid $20 to do the same thing. The results showed ;
3
that the lower the monetary inducement the more likely it
was that the subject's post-experimental attitudes toward
the tasks would correspond to his compliant act, i.e.,
subjects paid $1 evaluated the tasks and experiment more
favorably than did the $20 subjects.
A subsequent experiment by Cohen (1962) provided
' j
strong support for the above finding. The study was per- |
formed to rule out the alternative explanation that the j
$20 payment in the Festinger and Carlsmith study was so j
large that it engendered suspicion and resistance, leading j
i
subjects to suppress any attitude change which otherwise |
might have occurred. Cohen's subjects were offered $.50, i
j
$1, $5, or $10 to write an essay advocating an opinion
with which they initially disagreed. The post-essay atti- j
s
tudes essentially duplicated the Festinger and Carlsmith I
results: the larger the incentive, the less the attitude |
change in favor of the position advocated in the counter- j
attitudinal essay. j
Several other studies have obtained results that i
support the dissonance predictions in the forced compli- ;
1
ance situation (e.g., Carlsmith, Collins, and Helmreich, j
!
1966, * Conolley, 1970; Conolley, Wilhelmy, and Gerard, 1968; !
Linder, Cooper, and Jones, 1967). Nonetheless, there has
been contradictory empirical evidence as well as competing J
theoretical viewpoints. Extensive discussion of the j
dissonance— incentive controversy is presented elsewhere i
4
(Aronson, 1966, 1968; Chapanis and Chapanis, 1964; Elms,
1967; Janis and Gilmore, 1965; McGuire, 1966; Rosenberg,
1965, 1966; Zajonc, 1968) and a consistency model has been
present which appears to settle this controversy (Conol
ley, 1970; Gerard, Conolley, and Wilhelmy, in press).
Thus, it will not be discussed further.
A second alternative to the dissonance explanation
of "forced compliance" phenomena has been suggested by
Bern (1965, 1967b), who favors a "radical behaviorism"
position. Bern’s position often appears to result in the
same empirical predictions as those made from dissonance
theory. However, since Bem'.s interpretation avoids the
use of any motivational concept, the two theories offer
very different explanations of the observed phenomena in
the "forced compliance" situation.
Self-perception theory centers around two basic
postulates. The first states that "Individuals come to
’know’ their own attitudes, emotions, and other internal
states partially by inferring them from observations of
their own overt behavior and/or the circumstances in which
this behavior occurs (Bern and McConnell, 1970, p. 23)/'and
the second states that "to the extent that internal cues
are weak, ambiguous, or uninterpretable, the Individual is
functionally in the same position as an outside observer,
an observer who must rely upon those same external cues
to infer the individual’s inner states (Bern and McConnell,
................................ 5 ]
1970, P. 23)." On the basis of these two postulates, Bern
I
(1967b) has attempted to provide an alternative to the j
dissonance explanation of studies which found that the
less the justification a person is given for behaving
inconsistently, the more he will change his attitude to j
be consistent with his behavior (e.g., Cohen, 1962;
I
Festinger and Carlsmith, 1959). Bern suggested that the
j
dissonance situation be analyzed from the viewpoint of an
outside observer who is aware of the individual's perform- |
ance of the discrepant task and of the fact that the person
i
was paid to do it. He hypothesized that the observer
would view the behavior of the subject who received a |
i
high incentive as a set of "mands," i.e., that the behavior;
was elicited by the person's desire to gain the extrinsic !
monetary reward rather than by his intrinsic feelings aboutj
the situation. Such a perception would result in the i
observer concluding that the subject's behavior could not j
be used to infer his "true" attitudes since the behavior i
1
1
appears to be under the control of explicit external rein- j
l
forcement contingencies. Thus, the observer's best guess j
would be that the individual's attitude toward the task, j
like anyone else's, would be unfavorable. On the other j
hand, an observer would perceive the behavior of an indi- j
vidual who received a low incentive for the discrepant act j
as a set of "tacts," i.e., that the behavior was elicited j
by the person's intrinsic feelings about the situation. j
! In this case the observer would use the behavior to infer
that the individual's attitude is consistent with his :
behavior and, thus, conclude that the individual's atti
tude toward the task was favorable. |
The self-perception analysis asserts that subjects
in the dissonance experiments are themselves behaving like i
i
these hypothetical observers, i.e., the subject in such an j
:experiment is simply an observer of his own behavior.
Thus, on the basis of the first postulate of self-percep- ;
tion theory, the subject would infer his own attitudes and
;beliefs from his behavior if that behavior appears to be
free from the control of explicit reinforcement contin
gencies. Reinterpretation of dissonance studies proceeds j
by utilizing the second postulate of self-perception theory!
i
:which states that "to the extent the internal cues are j
weak, ambiguous, or uninterpretable, the individual is i
functionally in the same position as an outside observer,
; an observer who must rely on those same external cues to
infer the individual's inner states." If this notion that |
the original dissonance subjects behave the same as outside,
i
I
:observers is correct, then an outside observer who infers j
- ' i
a subject's attitudes should be able to replicate the in- j
verse functional relationship between magnitude of incen- I
; ' • !
: tive and attitude change that exists for the original sub- j
jects. This analysis has been tested by using an experi- !
: i
i i
mental paradigm known as "Interpersonal simulation" ‘ (e.g., J
7
Bern, 1965* 1967a, 1967b, 1968; Bern and McConnell, 1970).
The basic methodology involves asking an observer-subject
to estimate the attitude of a subject who is either de
scribed or overheard performing a behavior under one of the;
typical dissonance experiment conditions. The self-percep-|
tion prediction that observer-subjects can reproduce the ■
S
original dissonance findings has been supported (e.g., Bern,!
1965, 1967b). :
i
Not surprisingly, however, the use of Interpersonal '
simulations to support self-perception theory has been
criticized. Jones, Linder, Kiesler, Zanna, and Brehm
(1968), for example, have suggested that Bern's observer- ;
subjects' reproductions were artifactually produced by the :
descriptions of the experimental procedures they received. |
According to Jones et al. (1968), the descriptions allowed j
I
observer-subjects to infer systematic differences between
conditions in initial attitudes of the involved subjects, j
i.e., to infer differential self-selection in the original |
experiments. They argued that the hypothetical subject in j
i
the description performed the requested behavior even
i
though Bern's descriptions imply that a typical subject
would be quite unwilling to comply. Therefore, observer-
subjects would be expected to infer that their subject was j
atypical and that he was initially more willing to comply
than the average subject. Furthermore, complying for a i
; small incentive would be perceived by the observer as more j
i '
; atypical than complying for a large reward. j
To test this alternative hypothesis that Bern's j
observer-subjects were merely judging differential hypo- j
thetical-subject self-selection, Jones et al. (1968) per- !
formed a series of studies in which they attempted to both j
: i
; replicate Bern's results using his descriptions and to test |
; I
the effect of descriptions which eliminated observer judg- j
i
ments of self-selection. Bern's (1965, 1967b) results were j
repeated when his procedure was used, but not when initial j
attitude information was controlled. Jones, et al. con- j
eluded that Bern's methodology does not provide a test of i
I his theory since it does not provide the observer-subject j
|
with the complete information which is available to the ;
i
original subject, i.e., his initial attitudes. Similar j
criticisms have been expressed by several others (Harvey j
and Mills, 1971; Mills, 1967; Mills and Harvey, 1972; j
• Piliavin, Piliavin, Loewenton, McCauley, and Hammond, j
1969). I
Bern (1968), in responding to his critics, has j
t
I argued that much of the criticism has been based upon a j
misunderstanding of the simulation methodology. He sug- j
gested that Jones et al. (1968) were incorrect both in
assuming that their alternative interpretation of the j
; interpersonal replications differed from the original j
one advanced by self-perception theory and in assuming j
9
that their confirmation of that alternative interpretation
constituted a refutation of the original hypothesis that
the same judgmental process is implicitly employed by
actual subjects in the original dissonance experiments.
He did agree, however, with the Jones et al. warning con
cerning the use of the simulation paradigm. He stated,
’ ’ No 'as if' methodology, including the technique of inter
personal simulation is an adequate substitute for the
intensive study of the actual situation being modeled
(Bern, 1968, p. 273)."
Bern and McConnell (1970) reported a study designed
to evaluate the self-perception prediction that any con
flicting initial attitude that a dissonance experiment
subject may have had prior to the experiment is no longer
salient for him at the time of the posttest. If the self-
perception analysis that a subject arrives at his final
attitude by the "self-selection rule" is correct, then
the pre-manipulation attitude is, phenomenologically, the
same as the post-manipulation attitude, and there is no
attitude "change" as such. The empirical question is,
thus, "After engaging in counterattitudinal behavior, do
subjects in cognitive dissonance experiments 'know' their
initial attitudes or don't they? Do they know their ini
tial attitudes or do they perceive them to be the same as
their postmanipulation attitudes, as the self-perception
analysis requires? (Bern and McConnell, 1970, p. 25)."
10 i
j
Two experiments were performed simultaneously to j
i
answer this question, one which was conceptually identicalj
t j
to the typical dissonance experiment (Attitude-Change) and !
one which examined the salience of initial attitude after !
having performed the discrepant behavior but prior to the j
final attitude assessment (Attitude-Recall). Both experi-j
ments utilized a forced compliance paradigm in which sub- i
: i
jects were induced to write counterattitudinal essays under
varying conditions of freedom of choice to write the essayj
In the Attitude-Change experiment, pretest opinion was
measured and then, in a second session, subjects wrote
the essay and indicated their final opinion on the issue. ;
j
This experiment faithfully recreated the, conditions of the j
typical forced compliance experiment. The results showed '
that subjects given high choice to engage in the counter
attitudinal behavior showed significantly greater attitude j
change in the direction of agreement with that behavior
than did subjects given low choice or control subjects. i
In the Attitude-Recall experiment, after completing their J
essays, subjects were asked to recall the initial attitude |
I
they had expressed in the first session. Then they were
j
asked their final opinion on the issue. According to self-I
i
perception theory, engaging in the behavior provides such j
strong cues for the subject that any control exercised by ;
the initial attitude is "swamped" (Bern, 1968). Therefore, j
(
subjects asked to recall their initial attitudes should |
11
perceive them as being similar to their postmanipulation
attitudes, i.e., they should not perceive any attitude
"change." The results supported this analysis. Subjects i
were unable to recall the premanipulation attitudes cor
rectly and, in fact, perceived their pre- and postmanipu
lation attitudes to be identical. Bern and McConnell (1970)
concluded that the results supported the viability of the
self-perception analysis of forced compliance phenomena.
Since the initial attitude was not salient for dissonance
subjects, the assumption of isomorphism between observer-
subjects and original subjects was considered to be valid. ;
At the conclusion of their article, Bern and McCon
nell (I9 7 0) stated that it seemed unlikely that a "crucial"
experiment for discriminating between dissonance theory and
self-perception theory would ever be executed since the two
theories appear to make identical empirical predictions.
However, it is precisely the controversy over the role of
the initial attitude that does suggest a possible "crucial"
experiment. It appears that the two theories disagree
as to what is the effect of the extremity of the initial
attitude or change that occurs following the commitment
j
to a discrepant act. According to Bern and McConnell
(1970), "If an interpersonal simulation is to comprise a
valid test of the isomorphism between the subject and an
observer, then the theory dictates that a conflicting
’initial' attitude of the original subject must not be
j j
I part of the 'input' description for the observer any more
than it is for the subject himself (p. 24)." As mentioned'
above, Bern and McConnell found that initial attitudes were j
i
not salient for dissonance subjects and, thus, concluded !
that the self-perception assumption of isomorphism between j
observer-subjects and dissonance subjects was valid. Thus,!
it would appear that self-perception theory would postu- !
late that differences in initial attitudes are of little j
significance in determining final attitudes.
Quite obviously, dissonance theory does not make
the same prediction. In the forced compliance situation i
increasing the extremity of the initial attitude provides
i
; cognitive support for noncompliance. Therefore, dissonance
i
; theory would predict that, other things being equal, the I
more extreme the initial attitude, the greater the post- i
■ i
commitment dissonance and subsequent change in attitude |
j !
to make it consistent with the discrepant behavior.
The two theories do agree, however, as to the i
I effect of varying the justification used to induce com- j
; I
i mitment to the discrepant act. Both dissonance theory j
i j
] :
and self-perception theory would predict that a large j
; i
j monetary incentive would result in little or noattitude j
: change. For dissonance theory, increasing the incentive |
! would provide addititional cognitive support for compli-
i ance and would result in little postcommitment dissonance
; and little subsequent attitude change. For self- j
...........................................13...
perception theory, this high incentive would decrease the
importance of the behavior in the final attribution (Bern,
1967b). That is, the subject would view his behavior as
being under the control of the external reinforcement con
tingency of the large incentive, i.e., a "mand." There
fore, the subject would not use his behavior to infer his
attitude 1 As with dissonance theory, the self-perception
prediction would be that the subject would report little
or no attitude change.
The logical tests of these competing predictions is
thus to create a forced compliance situation in which the
extremity of the initial attitude and the justification
for commitment can be manipulated. Previous studies have
shown that aspects of physiologically based states, e.g.,
thirst, may be influenced by a person’s attempt to reduce
dissonance (Brehm, 1962; Brock and Grant, 1963; Mannsson,
1969). This research has shown that a highly thirsty
person who commits himself to further water deprivation
will experience dissonance and show subsequent reduction
in the strength of his thirst motive or drive. Such a
paradigm provides an excellent way of evaluating the
competing predictions made by the two theories. The
extremity of the initial attitude could be varied by
manipulating a person's initial thirst level (low or high)
and the justification for commitment could be varied by
14
manipulating the monetary incentive used to induce the |
x 1 !
commitment (low or high). j
The predictions derived from dissonance theory are j
apparent. The hypothesis would be that the extremity of
I
the initial attitude and the incentive for commitment !
would interact. This is perhaps most easily understood
if one conceptualizes the various initial attitude- j
i
incentive conditions in terms of the amount of justifica- j
tion the person would have for committing himself to per- ;
i
form the discrepant act. If an individual were very j
[
thirsty and yet agreed to further water deprivation for j
i
a small incentive., then he would have very little justi
fication for his highly discrepant behavior. This indi- j
i
I
i
vidual should experience a large magnitude of dissonance ;
i
and display a large amount of subsequent attitude change. I
However, a very thirsty individual who received a large 1
j
incentive for his compliance would be expected to experi- |
ence less dissonance. The large reward should provide himj
i
with at least moderate justification and, consequently, •
n i
Some might question the choice of thirst as the j
attitude which is manipulated. This choice was made for j
several reasons. First, it can be brought under the di- j
rect control of the experimenter and can thus be "born"
in the laboratory. Second, the use of thirst allows one
to obtain both self-report and behavioral data from the
subjects. The advantages of these controls are self-
apparent. Also, it seems of little value to distinguish,
in this case, between attitudes and other internal states
since both are important determinants of behavior. I
i
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
..................................................................... ' ................1 5 ““ "
dissonance arousal and subsequent attitude change would be
expected to be relatively less than in the above condition
Dissonance theory would also predict a main effect for the
extremity of the initial attitude. For a subject who has
a relatively low initial thirst level the commitment to
further deprivation would be low in discrepancy and, thus,
less dissonance arousing than it would be for a subject
who had a relatively high initial thirst level. Conse
quently, the monetary incentive would by itself provide
at least moderate justification for a low thirst subject
since less justification would be necessary in the first
place. Therefore, an individual in either of the two low
initial thirst conditions would be expected to experience
relatively little dissonance and to show little subsequent
attitude change, regardless of the incentive used to in
duce compliance. Either the low or the high incentive
should provide enough justification for commitment to the
future deprivation. These predictions derived from dis
sonance theory are illustrated in Figure 1.
A self-perception analysis of this situation would
result in competing predictions. Dissonance theory would
predict a main effect for extremity of initial attitude;
whereas, it appears that self-perception theory would
predict that the extremity of the initial attitude would
ATTITUDE CHANGE
Highly
Positive
Zero or
Negative
High Thirst
Low Thirst
Low High
INCENTIVE FOR COMPLIANCE
Figure 1. Theoretical predictions derived from
dissonance theory
o\
...............................................17....i
2
have little or no effect on the subject’s final attitudes. |
If this assumption is correct, only the manipulation of
justification would have a significant effect on final ;
attitudes. Self-perception theory would thus predict that j
a subject who received a large incentive would report less j
attitude change than would a subject who received a small !
incentive, regardless of the discrepancy involved. Low
t
incentive subjects would infer their attitudes from the
behavior and report a "more favorable" final attitude. j
High incentive subjects, however, would view their behavior?
i
t
as being under the external control of the incentive and
would not use that behavior to infer their attitudes. j
Therefore, these latter subjects would report little or
i
no attitude change. These predictions derived from self-
perception theory are illustrated in Figure 2. ;
The hypotheses derived from the two theories can be i
summarized quite succinctly. Dissonance theory would pre- ;
i
diet that there would be both a main effect for extremity ;
of initial attitudes (thirst level) and an interaction !
j
effect between the extremity of the initial attitude and j
i
the justification for commitment (incentive). The greatest
amount of attitude change should occur with subjects in
2 i
In fact, self-perception theory is not precise
enough to accurately derive predictions for the effects
of a discrepancy manipulation, but it would seem to imply |
that such a manipulation would be irrelevant to post
commitment attitudes. !
ATTITUDE CHANGE
Highly
Positive
High Thirst
Low Thirst
Zero or
Negative
High Low
INCENTIVE FOR COMPLIANCE
Figure 2. Theoretical predictions derived from
self-perception theory.
19
the high thirst/low incentive condition. The high thirst/
high incentive subjects should also display significant
attitude change but relatively less than the high thirst/
low incentive subjects. Neither the low thirst/low incen
tive nor the low thirst/high incentive subjects would dis
play significant attitude change. Self-perception theory,
however, would predict only a main effect for the justifi
cation manipulation. Subjects in both of the low incen
tive conditions should experience significant and similar
attitude change, while subjects in both high incentive
conditions should display little or no attitude change.
CHAPTER II
I
METHOD |
i
Design Overview
The experimental design required to test the hypoth
eses was a 2 x 2 factorial in which the extremity of the j
initial attitude and the incentive for the induced behav- j
ior were the manipulated variables. Subjects were randomly
assigned to four experimental groups: (1) high thirst/
low incentive, (2) high thirst/high incentive, (3) low j
i
thirst/low incentive, and (4) low thirst/high incentive.
!
i
Two addition "control" groups in which only thirst was j
manipulated were also included for the purpose of estab
lishing baselines of changes in thirst over time. Utiliz- :
j
ing a modified version of the procedure employed by Mann- I
sson (1969), the study was described to all subjects as j
"an experiment which is examining the relationship between j
1
level of thirst and perceptual judgements." They were !
i
told that they would perform a short perceptual judgement |
j
task and that several times throughout the experiment they!
j
would be asked to Indicate on various scales how thirsty j
they felt. During a specified break, all experimental
subjects were asked to volunteer for a future experiment
1
i
t
20 i
............ 21....j
t
which was "examining the psychological consequences of ;
being deprived of water for 24 hours." Following the
experimental subjects* commitment to the "future experi
ment," the "present" experiment was completed. The depend-!
ent measures were changes in subjective thirst ratings on j
i
two self-report scales and the actual amount of water i
I
consumption that occurred. |
i
Procedure !
Subjects. Ninety-three female undergraduates volun-:
teered to participate in the experiment as partial fulfill-'
ment of an introductory psychology course requirement at ;
i
the University of California, Los Angeles. Three of the
93 subjects were eliminated from the experiment and com- j
plete data were not collected from them. They were ex
cluded because one could not see the projection screen j
from a distance of more than three feet, one was allergic j
!
to peanut butter, and one had a kidney problem and could
i
not go without liquids for 24 hours. The remaining 90 j
subjects were randomly assigned to groups in each of the j
j
four experimental and two "control" conditions. j
1
j
Preparation. The subjects, who were tested in j
groups of three to six, were isolated from each other by j
1
partitions between their tables. Each table held a closed !
1
booklet, a pencil, and two covered paper plates and faced j
a large projection screen about 15 feet away. |
"" ' 22 |
After the subjects were seated, they were told to j
open the booklet and follow along as the experimenter read ;
t
i
the instructions (see Appendix A for the complete booklet).:
The instructions stated that the subject would be partici- j
pating in an experimental study of how biological drives
affect psychological variables, specifically the relation- ;
ship between level of thirst and perceptual judgements.
Further, the subject would be performing a short perceptual
i
task and would also be rating her feelings of thirst on i
various scales several times during the experiment.
i
i
Thirst arousal and precommitment measures. Thirst
1
was aroused under controlled laboratory conditions in the
following manner. After the general instructions, the
experimenter read to the subjects instructions that empha- !
sized the importance of having control of how thirsty ;
people are when they perform the perceptual task. He j
i
stated that it was crucial that everyone have approximately
the same level of thirst when performing the perceptual j
task. Therefore, the subjects were asked to remove any
gum, candy, or other substances which they might have in j
j
their mouth and then to uncover the paper plate in front !
of them and eat the entire cracker on the plate. Subjects j
in the high thirst conditions ate a saltine cracker with !
|
one.teaspoon of a specially prepared hot sauce that created
sensations of a hot, dry mouth. The recipe for the sauce j
(Mannsson, 1969) consisted of two parts catsup, two parts ;
i
tobasco, and one part horseradish. Subjects in the low j
thirst conditions ate a saltlne cracker with one teaspoon j
of peanut butter. All subjects in each group were run j
1 i
under only one of the experimental conditions. All sub- !
jects were then told that it was necessary to wait approx- I
I
imately two minutes in order for their thirst level to j
stabilize and that then they would be asked to rate their
1
feelings of thirst.
After the two minutes had passed, the experimenter j
read the instructions that explained the two rating scales I
on which the subjects were to indicate how thirsty they i
felt at that moment. One measure consisted of a bipolar i
scale with the end points labeled: Extremely Thirsty— Not j
At All Thirsty. They were connected by horizontally |
ordered numbers and dots from +30 to -30. Subjects were j
instructed to circle the number or dot which best expressed
their feelings of how thirsty they felt at that moment. j
j
The second measure was an equal appearing interval adjec- j
!
tive taste scale developed by Dillehay, Bruvold, and j
Siegal (1967) which consisted of selective adjectives !
i
describing a glass of water. The adjectives were selected j
i
to differ by about .5 to 1.0 scale units from each other j
i
(see Appendix B for the scale values). Subjects were j
instructed to circle one or more of the adjectives that !
! would best describe how a glass of water would taste to
i !
J
them at that very moment.
Perceptual task. Upon completion of the two thirst •
I
i
measures, the perceptual judgement task was begun. Sub- ;
jects were told that they were going to be shown a series i
of slides of famous paintings, one at a time, and that j
after each slide was shown they were to check off on a j
checklist all the items that they remembered having seen
in the slide. The ten slides were slides of famous paint- |
i
ings which depicted numerous portraits, landscapes and I
; ■ i
water-related scenes. Each slide was shown for 20 seconds.|
Subjects were then allowed 60 seconds after the slide was |
removed in which to complete the checklist. A separate ;
checklist was employed for each slide. Following the
completion of the last checklist the subjects were instruc-
i
: ted that the only remaining task was for them to once
again rate their feelings of thirst. As before, they were i
; told that it would be necessary to wait approximately two j
minutes before completing the thirst scales. !
; j
s |
Counterattitudinal commitments. While the subjects j
;were waiting for the two minutes to pass before completing
; j -
ithe thirst scales, the experimenter mentioned that he could!
i
; i
now provide them with more information about the experimentj
He stated that the present study was the third in a series
of four experiments that he was performing in order to i
! learn more about how biological drives affected psycho- j
; logical variables. The first two studies, they were
told, had examined the effects of hunger. The fourth
experiment, like the present one, would examine the
effects of thirst. The experimenter then stated that he
would like to describe the next experiment so that the
subjects could decide whether or not they would like to
participate. He explained that it would be very helpful
if they did participate since he already had some data
from them, but that it was entirely up to them as to ;
i
whether or not they agreed to do so. At this point he
handed to each subject a copy of the booklet to be used !
and asked them to look it over while he' described what ;
they would be doing (see Appendices C and D for the
; "future experiment" booklets). The experimenter then
i stated that in this study, which would start immediately j
after the completion of the "present" experiment, they
would be required to go without liquids for 24 hours and j
i
I then to return and answer a series of questions about 1
i their experiences during that period. Since it would be
1 unpleasant, subjects were told that they would be paid j
; $5 ($20 depending on the condition) for participating but j
: that no additional experimental credit would be offered. I
; The experimenter then asked each subject to sign and date ;
I
; a commitment slip if she was willing to participate, re- j
i !
! stated that it was entirely up to her, and then paused. j
26 !
Any questions about the future experiment were briefly
answered but no further encouragement or prompting to
sign up were given. Although many subjects appeared
somewhat hesitant to volunteer, all except the one with i
the medical problem agreed to participate in the future
l
study and signed a small signup slip which read "I ________I
agree to participate in the thirst deprivation study for !
the sum of $5 ($20). ITfurther agree to go without liquids
for twenty-four hours, at which time I will return to this j
room to complete the study, have the special litmus *band-^
t
aid' removed, and receive my money." The signing of this j
slip constituted the commitment to the discrepant behav- !
ior. After signing the commitment slips, the subjects
i
were told that they would first finish the "present" per- I
ceptual judgement experiment by completing the thirst |
scales and then they would be told more about the water-
deprivation study.
i
Control against false compliance. Included in the :
above instructions concerning the "future" water depriva- ;
tion experiment was a procedure designed to make the sub- j
jects feel that they would not be able to cheat on the ■
water-deprivation restrictions without being detected.
j
During the description of the study the experimenter j
j
stated that each subject would have attached to her arm i
i
"a small band-aid like piece of tape which has on it a I
27
: special type of litmus paper." This was said to be for
; two purposes— a check on liquid intake during the 24 hours,
: and a measure of the body's reaction to the water depriva
tion. Subjects were warned not to get this tape wet while
washing or bathing.
"Control" groups. Subjects in the two "control"
conditions performed identical tasks up to the point where
; the counterattitudinal commitment was introduced to the
; experimental groups. At this time "control" subjects were
told only that, as before, it would be necessary to wait
approximately two minutes before completing the thirst
scales. Following this statement, the experimenter re
mained silent and the subjects sat quietly until two min-
• utes had passed. At this time, the experimenter asked the
subjects to turn their booklet to the next page of instruc
tions. The instructions, which were the same as those used
for the experimental groups, stated that the last task for
' the study was to once again indicate how thirsty they felt
;at that very moment. The two scales were the same as they
:had used earlier. After all subjects had completed both
;scales, the experimenter collected their booklets, signed
their experimental participation cards, and debriefed them.
Postmeasures. After the signing of the commitment
slips, the subjects were instructed to put aside the book
let for the future experiment and to return to the
28 j
instructions for the present study. The instructions j
stated that the last task for this study was to once
again indicate how thirsty they felt at that very moment.
j
The two scales were the same as they had used earlier. !
i
After all subjects had completed both scales the experi- |
menter collected their booklets, signed their experimental i
participation cards, and stated that he would then explain I
in detail what they were to do for the next experiment. j
They were told that before having the litmus paper attached
they would be allowed to have one last drink of water. j
Each subject was then allowed to drink as much water as ;
she desired. This was done by placing a pitcher of water, j
a large paper cup, and a straw in front of each subject. ;
The straws were employed in an attempt to help standardize ;
the conditions under which the drinking occurred. No one i
could see how much water the others drank. j
After all subjects drank as much as they desired,
l
they were debriefed and informed of the true purpose of j
the study. None of the subjects showed any evidence of
: suspicion or knowledge that they had been in an "attitude" !
change study. The subjects were then allowed to leave. ;
:At this time the experimenter recorded the number of |
mililiters of water consumed by each subject. This 1
amount was used as an additional dependent measure. I
CHAPTER III
RESULTS
The main data from the experiment consist of the
subject's change in ratings of thirst on the bipolar
adjective and the adjective checklist scales and the
subject's actual water consumption following commitment
to the 24-hour deprivation. A change score for each of
the two adjective scales was computed by subtracting the
post commitment rating from the precommitment rating for
each scale. This procedure yielded two attitude change
scores for each subject. For simplicity, the scores were
constructed such that changes in the positive direction
indicated a more "favorable" attitude toward the discrep
ant behavior. This procedure was used so that larger posi
tive scores would indicate more change in the predicted
direction of a more "favorable," i.e., less thirsty, atti
tude following the commitment. For the behavioral measure
of actual water consumption, the precise amount of water
drunk by the subject was measured and recorded in millilit
ers. This amount was subtracted from the amount of water
that was initially present, 414 milliliters, so as to give
a measure of how much water remained for each subject.
30 i
i
i
This procedure was used so that the more water that re- |
mained, the more it should indicate a change in the pre- j
dieted direction of less thirst following the commitment «
to the future deprivation. j
Validity of the Premeasures
!
In order to correctly use the various measures it
is first necessary to evaluate their validity. If, in
j
fact, the two self-report scales are both measuring per
ceived feelings of thirst, then one would expect them to
i
be positively correlated. Table 1 shows that the first
adjective checklist and bipolar scales are, indeed, highly1
i
correlated (jo < .005).
|
Success of the Initial Thirst Manipulation
Before examining the data concerning the change in !
the subject’s thirst ratings (attitudes), it is necessary I
to evaluate whether the low and high initial thirst levels :
were induced in the appropriate subjects. Means for the ;
j
ratings on the first bipolar adjective scale and the ad-
I
jective checklist scale are presented in Table 2 and
Table 3, respectively. Analyses of variance on these j
i
l
measures (Tables 4 and 5) indicated that the manipulation j
of thirst by eating the hot sauce cracker or the peanut I
i
butter cracker was quite successful. Subjects in the !
High Thirst conditions rated themselves as being signif- j
!
icantly more thirsty than did subjects in the Low Thirst j
31
TABLE 1
Correlations Between the First Two
Thirst Scales8,
Scales N Means df R
y
Adjective Check- 90 830.51 12.11 88 0.76*
list Scale and
Bipolar Scale
^he two "control" groups are included.
*£< .005
TABLE 2
Mean Ratings of Thirst on the
3>
First Bipolar Adjective Scale
Initial Thirst Incentive for
Level Commitment Control
$5 $20
High Thirst
22.27. (15) 20.47 (15) 21.93 (15)
Los Thirst 2.47 (15) 3.60 (15) 1.93 (15)
aLarger number indicates more thirst was reported.
The cell n’s are presented in the parentheses.
33
TABLE 3
Mean Ratings of Thirst on the
First Adjective Checklist Scalea
Initial Thirst Incentive for
Level Commitment Control
$5 $20
High Thirst 9.20 9.18 9.30
Low Thirst
7.12 7.83 7.22
Larger numbers indicate more thirst was reported.
The verbal descriptions relating to these values
can be approximated by examining Appendix B.
34
TABLE 4
Analysis of Variance of Mean Ratings of Thirst
on the First Bipolar Adjective Scale
Source SS df MS F
Initial Thirst
Level (A)
8027.78
1
8027.78 73.28*
Incentive (B) 3.10 2 1.60 0.01
A x B
46.15
2
23.07
0.21
Within (error)
9021.87
84
109.55
*£ < . 001
35
TABLE 5
Analysis of Variance of Mean Ratings of Thirst
on the First Adjective Checklist Scale
Source SS df MS F
Initial Thirst
Level (A)
' 760,305.13
1
760,305.13
48.60
Incentive (B) 19.005.00 2 9.502.50 0.60
A x B 2 6,620.00 2 13,110.00
0.83
Within (Error)
1.314,048.25
84
15,643.43
#
£ < . 001
36
conditions on both the bipolar adjective scale (£ < .001)
and the adjective checklist (jd < .001). There were no
other significant main effects or interactions. Further
more, there were no significant differences among the High
Thirst means nor among the Low Thirst means for either
scale as shown in the contrast analyses (Winer, 1962) pre
sented in Tables 6 and 7.
Validity of the Change Scores
As with the premeasures, it is necessary to estab
lish the validity of the change scores. To be valid, the
change scores on the Bipolar Adjective Scale and in the
Adjective Checklist scale should be positively correlated.
Table 8 shows that, indeed, the change scores on the two
scales are highly correlated (jd < .005). Likewise, if
these two scales are valid, they should correlate posi
tively with the behavioral measure of how much water re
mained after consumption. As shown in Table 8, the change
scores on both the Bipolar Adjective Scale and the Adjec
tive Checklist Scale are significantly correlated with the
behavioral measures of residual water following consump
tion (_p < .005 and £ < .023j respectively). Consequently,
all three measures appear to be of acceptable validity.
37
TABLE 6
Contrast Analysis of Mean Thirst Ratings
on the First Bipolar Adjective Scale8 ,
HT/LI LT/LI HT/HI LT/HI HTControl LTControl
HT/LI 26.84* 0.22 2 3.86* 0.01
28.31*;
LT/LI 2 2.18* 0.09 25.94* 0.02 ;
HT/HI 19.48* 0.15 23.52*!
LT/HI 23.01*
0.19
HTControl
27.39*
LTControl
High and low are abbreviated H and L, respec
tively, and thirst and incentive are abbreviated
T and I, respectively.
*£ < . 001
38
TABLE 7
Contrast Analysis of Mean Thirst Ratings on the
First Adjective Checklist Scalea
HT/LI LT/LI HT/HI LT/HI HTControl LTControl
HT/LI 20.65*** 0.01 9.00** 0.05 18.79***i
LT/Li 20.38*** 2.39 22.78*** 0.04
1
HT/HI 8.82* 0.07 18.53*** |
HTControl 10.42** 1.78
LTControl 20.83***
High and low are abbreviated H and L, respectively,j
and thirst and incentive are abbreviated T and I,
respectively.
*£ < .05
**£<..01
***£ < .001
39
TABLE 8
Correlations Between the Two Change Scores
cl
and the Behavioral Measure
Scales N Means df
Rx,y
Adjective Check
list Scale and Bi
polar Adjective
Scale
90 54.28 3.27 88 +0.65**
Residual Water and
Bipolar Adjective
Scale
60
268.05
4.92 58 +0.42**
Residual Water and
Adjective Check
list Scale
60
268.05 89.16 58 +0.29*
aThe two self-report scales include "control"
group subjects.
*£ < .025
**£ < .005
4o
Change in the Thirst Ratings
Table 9 presents the mean change from the first to
the second bipolar adjective scale for each experimental
and control condition. Analysis of variance of these
change scores (Table 10) shows significant main effects
for thirst level and incentive, as well as a significant
interaction between these two variables (£ < .001, £ < .01
and .001, respectively). The contrast analysis
(Table 11) revealed that, following the commitment to the
future deprivation, subjects in the High Thirst/Low Incen
tive condition reported themselves as being significantly
less thirsty than did subjects in any of the other exper
imental conditions. There was no significant difference
between the two low thirst conditions for this change
score. In fact, these two low thirst groups displayed
no significant changes whatsoever, as indicated in
Table 11 by the nonsignificant differences between these
two low thirst groups and the two control groups.
Data for the mean scores on the adjective checklist
scale are presented in Table 12. Analysis of variance
(Table 13) of these change scores shows the same signif
icant main effects and interactions as those for the
bipolar adjective scale (Table 10). The contrast analysis
for this scale (Table 14) also shows the same results.
The High Thirst/Low Incentive subjects again reported a
significantly larger decrease in thirst rating than did
41
Mean
TABLE 9
Change in Thirst Ratings on
Bipolar Adjective Scale8 ,
the
Initial Thirst Incentive for
Level Commitment Control
$5
$20
High Thirst
20.87 7.13 1.47
Low Thirst -5.00
-3.33 -1.47
Larger numbers, plus or minus, indicate greater
change. Positive numbers indicate that the post-
commitment rating was more "favorable," i.e.,
the subject reported herself as being less
thirsty.
42
TABLE 10
Analysis of Variance of Mean Change in Ratings
of Thirst on the Bipolar Adjective Scale
Source SS df MS F
Initial Thirst
Level (A)
3855
1
3855 46.93**
Aatbi^hlsh thirst)
5018 1 5018 61.19**
. Aatb2^low thlrst)
828 1 828 10.10**
Aatb3(oontro1)
97
1
97
1.12
Incentive B
1029
2
515
6.26*
Batax ($5)
5501 1 5501 6 7.08**
%ta2' ($20)
338 1 338 4.12*
A x B 2050 2
1025 12.47**
Within (error) 6898 84 82
*£ < .01
**£ < .001
43
TABLE 11
Contrast Analysis of Mean Change in Thirst Ratings
on the Bipolar Adjective Scale8 ,
HT/LI LT/LI HT/HI LT/HI HTControl LTControl
HT/LI 6 1.11*** 17.22*** 53.48***
34.37*** 45.55***
LT/LI 13.44** 0.25 3.82 1.14
HT/HI 10.00**
2.93 6.75*
LT/HI 2 .1 0 0.32
HTControl
0.79
LTControl
a
High and low and thirst and incentive are abbre
viated as in previous contrast analysis tables.
*2 < .05
**£ < . 01
***£< .001
44
TABLE 12
Mean Change in Thirst Ratings on the
Adjective Checklist Scale8 ,
Initial Thirst
Level
Incentive for
Commitment Control
$20
High Thirst 2.66 1.14 0.02
Low Thirst -0.24 0.04 -0.36
cl
Larger numbers, plus or minus, indicate greater
change. Positive numbers indicate that the post-
commitment rating was more "favorable," i.e., the
subject reported herself as being less thirsty.
45
TABLE 13
Analysis of Variance of Mean Change in Ratings
of Thirst on the Adjective Checklist Scale
Source SS df MS
F
Initial Thirst
Level (A)
48o;632 1 480,632
19.41**
Aatt) (high thirst)
1
715,695
1
715,695
28.92**
Aatb ^low )
88,238 1 88,238
3.57
Aatb (0°ntro1)
8,708 1 8 ,7 0 8
0.35
Incentive (B) 286,250 2
143,125
5.78*
Bata;L (*5)
683,213 2 341,607 13.80**;
Bata2 (*20)
6,794 2
3,397
0.14
A x B
255,215 2 127,608
5.15*
Within (error)
2,079,283
84
24,753
*
£ < .01
**£ < .001
46
TABLE 14
Contrast Analysis of Mean Change in Thirst Ratings
Si
on the Adjective Checklist Scale
HT/LI LT/LI HT/HI LT/HI HTControl LTControl
HT/LI 2 5.60*** 6. 99** 2 0.88*** 2 1.22*** 2 7.60***
LT/LI 5.84* 0.24 0 .2 0 0.04
HT/HI
3.71
3 .86 6.81*
LT/HI 0.00
0.47
HTControl 0.42
LTControl
aHigh and low and thirst and incentive are abbre
viated as previously.
*P < .05
***£ < •001
47
any of the other subjects. Also, as before, the two low
Incentive groups did not display significantly different
changes in thirst ratings, nor were they significantly
different from the two control groups, indicating no sig
nificant change whatsoever.
Water Consumption
The mean amounts of residual water following con
sumption for the experimental groups are presented in
Table 15. The analysis of variance (Table 16) indicated
the results were similar to those obtained using the two
self-report measures. That is, the actual amount of water
drunk varied as a function of the initial thirst level
(jd < .01). However, there were no significant effects
for the level of incentive or the interaction. The con
trast analysis (Table 17) revealed that subjects in the
High Thirst/Low Incentive condition consumed significantly
less water than did subjects in the Low Thirst/Low Incen
tive group (jd < .01) and than subjects in the Low Thirst/
High Incentive group (p < .05).
48
TABLE 15
Mean Number of Milliliters of Residual Water
Following Consumption8 ,
Initial Thirst Level Incentive for
Commitment
$5
$20
High Thirst
312.47
276.20
Low Thirst
231.73
251.80
cl
The larger the number, the more water that
remained after consumption. Larger numbers,
thus, indicate more "attitude change."
4 9
TABLE 16
Analysis of Variance of the Mean Number of
Milliliters of Residual Water
Following Consumption
Source SS df MS F
Initial Thirst
Level (A)
41,448.76 1 41,448.76 7. 21*
Incentive (B) 984.14 1 984.14
0.17
A x B
11,900.63
1
11,900.63 2.07
Within (error) 321,637.86 56
5,743.53
* .
£ < .01
50
TABLE 17
Contrast Analysis of Mean Number of Milliliters
S i
of Residual Water Following Consumption
HT/LI LT/LI HT/HI LT/HI
HT/LI
8.51** 1.72 4.81*
LT/LI
2 .5 8
0.53
HT/HI 0.78
LT/HI
aHigh and low and thirst and incentives are
abbreviated as previously.
*E < .05
**E < .01
CHAPTER IV
DISCUSSION i
j
I
The general purpose of the research reported in this!
dissertation was to test the competing predictions made j
from the dissonance and self-perception analyses of the !
effects of the extremity of the initial attitude and of
the incentive used to induce commitment within a "forced j
compliance" paradigm. First, it was necessary to manipu- j
late the extremity of the initial attitude. That this j
attempt was successful can be seen from Tables 2 and 3,
i
which show that subjects in the High Initial Thirst condi- j
!
tions did report significantly more thirst before the |
!
counterattitudinal manipulation than did the subjects in j
i
the Low Initial Thirst conditions, and from Tables 6 and 7 j
which show, respectively, that subjects within the High j
Initial Thirst conditions were not significantly different j
j
from one another at that time and that subjects within the I
;Low Initial Thirst conditions were not significantly dif- j
! . i
: !
ferent from one another. Thus, the Initial thirst manipu- j
• !
lation was highly effective. Table 1 shows that the first j
; i
Bipolar Adjective scale and the first Adjective Checklist j
: . i
scale were highly correlated. This high correlation j
strengthens the confidence of the success of the initial !
52
thirst manipulation in that it offers support for the fact
that the two self-report measures appear to be of accept
able validity.
Before examining the main results it is necessary
to also evaluate the validity of the various dependent
measures. Table 8 shows that the validity of these mea
sures was also quite acceptable. The change scores on the
two self-report thirst scales were highly correlated and
the scores on those two scales were significantly corre
lated with the behavioral measure of the residual water
following consumption; all correlations were in the pre
dicted directions. The fact that both paper and pencil
self-report and behavioral criteria were used further
strengthens the case for the validity of these measures
in that it helped to reduce the possibility of spuriously
high correlations due to method variance (Campbell and
Fiske, 1959). Method variance might also explain the fact
that the correlation between the two self-report measures
(same method) were higher than the correlations between
the behavioral measure and the self-report measures (dif
ferent methods). Also, it is worthy to note that it ap
pears that when a person convinces herself that she has
little thirst, she actually consumes less water. Thus
even our basic motivations appear to be not simply a
function of the "basic drive levels," but they also
depend at least in part on the cognitions associated with !
I
the motivational states (Mannsson, 1969). i
As can be seen from Figures 3* 4, and 5, the pre
dictions derived from dissonance theory were strongly
upheld while those derived from self-perception theory '
found little support. The dissonance hypothesis that the
greatest attitude change would occur when discrepancy was ;
high and justification was minimal was strongly supported !
i
on all three of the dependent measures. Tables 9 and 12
show a sharp decrease in self-reported thirst for the High ;
Thirst/Low incentive subjects and Table 15 shows that these
subjects actually did consume less water when given the
opportunity. Tables 11, 14, and 17 show that these changes
occurred significantly more for subjects in this condition j
than for subjects in any of the other conditions. Self- j
{
perception theorists would be hard put to explain these i
findings. j
j
Likewise, the dissonance hypothesis of a main ef
fect for the extremity of the initial attitude was strongly
j
supported. Tables 9 and 12 show decreases in thirst on |
1
both of the self-report measures for subjects in the two j
I
High Initial Thirst conditions and no significant changes !
for subjects in the two Low Initial Thirst conditions.
Although the results obtained on the behavioral measure
I
were not as strongly supportive, the decrease was signif- j
leant for the High Thirst/Low Incentive subjects and the j
MEAN CHANGE I N RATING O F THIRST
25
20
High Thirst
15
10
5
0
iw Thirst
5
-10
$20 Control
LEVELS OF INCENTIVE FOR COMPLIANCE
Figure 3. Mean change in thirst on the bipolar
adjective scale for the two levels of
thirst, the two levels of incentive,
and the two thirst "control" groups
VJI
4 =-
MEAN CHANGE I N RATINGS O P THIRST
+3.5
+3.0
+2.5
High Thirst
+2.0
+1.5
+1.0
+0.5
0.0
Low Thirst
-0.5
- 1.0
$20 Control
LEVELS OP INCENTIVE FOR COMPLIANCE
Figure 4. Mean change in thirst on adjective checklist
scale for the two levels of thirst, the two
levels of incentive, and the two "control"
groups.
u i
U T
MEAN MILLILITERS O F RESIDUAL WATER
320
310
300
290
280
270
260
250
240
230
220
High Thirst
Low Thirst
$3 $20
LEVELS OF INCENTIVE FOR COMPLIANCE
Figure 3.
Mean milliliters of residual water.; following
consumption for the two levels of thirst
and the two levels of incentive.
\j i
o\
trend was in the predicted direction for the other High j
Initial Thirst subjects. Again, the self-perception j
analysis was not upheld since it predicted no significant j
effect for the discrepancy manipulation.
Furthermore, the other predictions derived from j
dissonance theory were strongly supported. It was hypoth-j
esized that the extremity of the initial attitude and the !
level of incentive for commitment would interact such that j
more attitude change would occur for High Thirst subjects
i
who were offered $5 than for High Thirst subjects who were j
offered $20. This prediction was upheld for both of the
i
self-report measures as is shown in Tables 11 and 14. i
Although the result for the behavioral measure was non
significant, Table 15 shows that the trend was in the j
direction predicted by dissonance theory.
The predictions derived from the self-perception I
analysis received little support. Tables 10 and 13 show \
i
that the self-perception hypothesis of no significant main :
]
effect for initial thirst level was not supported; sub
jects in the two High Initial Thirst conditions did dis- j
i
play significantly more attitude change than did subjects j
i
in the two Low Initial Thirst conditions. Furthermore, !
. i
the self-perception prediction of a main effect for lncen- j
tive received relatively little support. The hypothesis j
i
was that subjects in the two Low Incentive conditions j
would display significant and similar attitude change, j
j whereas subjects in the two High Incentive conditions j
i
■ would display no significant attitude change. There was j
: a significant main effect for incentive level, as seen
: in Tables 10 and 13. However, examination of the simple
j
main effects reported in these two tables reveals that the ;
' i
■ High Thirst/Low Incentive condition was primarily respon- j
sible for the significant main effect of incentive and that!
i
!
: the Low Thirst/Low Incentive condition contributed rela- !
tively little to the overall significant effect of incen- |
tive. Further evidence of the differences between the two !
Low Incentive groups is seen in Tables 12 and 14 which j
i
i
show that the High Thirst/Low Incentive subjects displayed j
; significantly more attitude change than did the Low Thirst/
Low Incentive subjects. Self-perception theory would find |
these results difficult to explain. I
i
i
The above discussion shows that self-perception !
theory would have a very difficult time accounting for :
I
I the obtained results. A weak attempt to do so might !
: I
; center upon the fact that the results obtained from the
; i
; behavioral measure did not as strongly support the dls- j
i sonance predictions as did the results obtained from the
!two self-report measures. Since self-attributions are !
1 said to occur primarily from observations of overt behav- j
|
ior rather than from internal states, the argument might j
; i
: be offered that only the behavioral measure was a valid
test of the theory. Such an argument would be difficult J
59
to support, especially since self-perception theorists
often employ self-report techniques in their work.
Furthermore, the results are difficult to explain by
self-perception theory even if this argument were accept
able. Self-perception theory could not account for the
fact that even on the behavioral measure the most attitude
change occurred when discrepancy was largest.
The greatest difficulty encountered in designing a
study to evaluate the relative merits of the dissonance
and self-perception analyses of the "forced compliance"
paradigm was the fact that self-perception theory (Bern,
1965, 1967b) does exactly what its proponents have so
often criticized dissonance theory of doing. That is, it
is extremely difficult to derive precise empirical predic
tions from the two basic postulates of self-perception
theory. The present study is a case in point. Self-
perception theory was not explicit enough to easily derive
predictions about the effects of an extremity of initial
attitude manipulation. Based upon the Bern and McConnell
(1970) study, the conclusion was reached that, according
to self-perception theory, extremity of initial attitude
would have little or no effect on final attitudes. This
conclusion was reached by logical interpretation of the
theory and the results supporting it rather than from the
theoretical implications postulated by the theory itself.
No such implications could be derived. However, after the
60 j
]
present study was completed, an extensive discussion of j
i
self-perception theory (Bern, in press) has become avail- ,
able which is precise enough to accurately assess the role i
of initial attitudes in the "forced compliance" situation :
as postulated by a self-perception analysis. Therefore,
it is of value to reexamine the present study in light j
of this recent presentation of self-perception theory. i
i
t
According to Bern (in press), the crucial aspect of j
the initial attitude is the degree to which it is salient !
i
to the subject. Self-perception theory postulates that an|
i
individual will rely upon external cues, e.g., the in- j
t
duced behavior, to infer his attitudes only when internal j
i
cues are weak, ambiguous, or uninterpretable. Conse- !
quently, one would predict that making the subject's j
i
initial attitude highly salient would diminish the extent j
to which the final attribution would be based upon the
induced behavior and, thus, would diminish the subject's !
attempt to change his attitude to make it consistent with j
this behavior. Thus, for the first time self-perception
theory is precise as to what parameter is crucial in de-
j
termining the effects of the initial attitude on the finalj
j
attitude; high salience of the initial attitude will re-
j
suit in the induced behavior having little or no effect
upon the subject's final attitude. Quite obviously dis- j
sonance theory does not make the same assumption. In the
forced compliance situation increasing the salience of the
61 |
: initial attitude provides cognitive support for non- i
; i -
compliance. Therefore, dissonance theory would predict
; just the opposite of self-perception theory: other things j
being equal, the more salient the initial attitude, the j
greater the postcommitment dissonance and subsequent change!
I
in attitude to make it consistent with the induced behav- |
' ior. I
i
The two theories do agree, however, on the conse- j
quences of making the behavior more salient. Both dis- j
i
sonance theory and the "new" self-perception theory would !
predict that if the induced behavior were made more salient^
the subject would experience more attitude change in the
direction of making it consistent with the induced behav- j
1
! ior. Bor dissonance theory increasing the salience of
I
the induced behavior would simply make the dissonant act I
more salient and, thus, result in more dissonance and sub- I
sequent attitude change. For self-perception theory this j
:would increase the salience of "the very source of evidencej
upon which the final attribution is based (Bern, in press, !
! ■ i
ip. 32)." That is, the subject would view his behavior as j
; i
a "tact" and use it as the basis for inferring his atti- j
|tudes. As with dissonance theory, the self-perception j
; j
prediction would be that the subject’s attitudes would
:"change" so as to be more consistent with the induced
behavior. i
The possibility also exists that both the initial
j
attitude and the behavior could be salient for the subject;
Once again the two theories make competing predictions. !
i
The "new" self-perception theory would predict that making ;
both salient would be equivalent to making the initial j
attitude alone salient. According to the self-perception !
analysis, if any element present in the situation implies !
that the behavior is not relevant, then the presence of
j
other cues implying relevance do not help. Bern (in press) j
states that "if any source implies irrelevance or low
j
credibility, then self-attributions will not occur (p. 1 9)"
In this case, the subject would view his behavior as a |
"mand" and would experience little or no attitude change, j
|
Dissonance theory, however, would predict that the great- j
est amount of attitude change would occur when both the
I
initial attitude and the induced behavior were salient. '
The result of making both salient would be maximal dis- ;
i
sonance arousal since the impact discrepancy between pri- I
vate opinion and compliance is maximal in this situation.
i
Thus, one would predict that the greatest attitude change j
i
would occur under these circumstances.
The logical test of these competing predictions !
j
would be to create a "forced compliance" situation in j
which the salience of both the initial attitude and of i
i
the induced behavior would be manipulated. However, before!
63 I
!
this is discussed further, it is of value to reevaluate j
t
the present study in terms of the "new'1 self-perception j
theory. Neither the salience of the initial attitude nor j
the salience of the induced behavior were manipulated in i
the present study. Yet, it does appear that one crucial j
element for testing the "new" self-perception analysis was;
present. It seems logical to assume that subjects in all ;
of the experimental conditions were well aware of their
initial attitude when they committed themselves to the !
future water deprivation. In full view of all experimental
!
subjects were the filled water pitchers, the paper cups, i
i
and the drinking straws. The presence of these materials j
should have made thirst level highly salient for all of j
these subjects. In fact, others have manipulated salience I
in this very manner; Brehm (1962) manipulated salience of j
i
initial thirst by either having a pitcher of water and j
i
cups in the subject's view (High Salience) or by not ■
' j
1
having these relevant cues present (Low Salience). There
fore, it seems safe to conclude that the salience of the
1
!
initial attitude was high for all the experimental sub- i
j
jects in the present study. If so, self-perception theory I
i
is not supported by the results. It would predict that j
1
1
subjects with highly salient initial attitudes would dls- j
|
play little or no attitude "change," regardless of the i
salience of the induced behavior. However, Tables 9, 12, !
64
and 15 show that the subjects in two of the groups did |
display significant attitude change. On the other hand,
as mentioned above, dissonance theory can readily account
for these results. The dissonance prediction would be that;
the more salient the initial attitude, the more the sup- !
port for noncompliance and, thus, the greater the post
commitment dissonance and subsequent change in attitude
to make it more consistent with the discrepant behavior. ;
It should not be concluded from the above discussion
that the value of the present study is decreased by the
information gained from Bern’s (in press) upcoming article. ;
The study does provide evidence that dissonance theory can j
account for certain patterns of results which self-percep- i
tion theory would find difficult to explain. Nonetheless, j
Bern's (in press) article does suggest an experiment which
could more clearly settle the dissonance— self-perception ;
controversy. By manipulating the salience of both the ;
initial attitude and of the induced behavior, the competingj
predictions of the two theories could be tested once and j
for all. In a study similar to the present one salience 1
!
of initial thirst level could be manipulated by having j
'thirst-related cues such as a full water pitcher, cups,
and paper straws either present (High Salience) or absent
;(Low Salience). The salience of the induced behavior couldj
{
be manipulated by the experimenter. For one-half of the |
subjects (High Salience) he could emphasize the unpleasant I
consequences of the commitment to future deprivation by-
pointing out such factors as the necessity to avoid liquid j
intake regardless of how strong the desire for water might j
i
become. For the other half (Low Salience) the experiments
j
er would not mention any such factors. On the basis of ;
the present experiment one would hypothesize that the !
I
results would support the dissonance analysis and, further-;
i
more, that the pattern of results would be similar to !
those obtained in the present study. j
In conclusion, then, the results strongly support
the hypotheses derived from dissonance theory; little sup-|
port for self-perception theory emerged. Thus, in spite ;
of Bern and McConnell's (1 9 7 0) statement that it was un- j
likely that a crucial experiment for discriminating be
tween dissonance theory and self-perception theory would j
ever be executed, the present study shows that the dis
sonance analysis of the "forced compliance" situation
appears to be far superior to that derived from self- j
perception theory. Furthermore, on the basis of the !
present experiment, it is expected that the results of j
the proposed study on the effects of various salience j
I
manipulations would provide additional support for the j
1
dissonance position. It is important, however, to put j
[
this conclusion in its proper perspective. This study j
66 .
should not be assessed only in terms of its support for
dissonance theory and its nonsupport for self-perception
theory. The true value of the study is in its contribu
tion to the scientific procedure of establishing the
limiting conditions of various theories; until the limits
are known, the use and value of any theory may be inappro
priately and/or unnecessarily applied or restricted. The
results of the present experiment suggest that dissonance
theory is appropriate for analyzing the "forced compli
ance" paradigm, whereas self-perception theory is inappro
priate. However, self-perception theory does appear to
offer explanations for other psychological phenomena and
it should be those circumstances, rather than the present
one, to which a self-perception analysis is applied and
evaluated. Dissonance theory remains the preferred ex
planation of the "forced compliance" situation.
REFERENCES
REFERENCES
Aronson, E. The psychology of insufficient justification:
An analysis of some conflicting data. In S. Feld
man (Ed.), Cognitive consistency: Motivational
antecedents and behavioral consequences. New York:
Academic Press, 1966.
Aronson, E. Dissonance theory: Progress and problems.
In R. P. Abelson et al. (Eds.), Theories of cogni
tive consistency: A sourcebook. Skokie, 111.:
Rand-McNally, I9 6 8.
Bern, D. J. An experimental analysis of self-persuasion.
Journal of Experimental Social Psychology, 1965, 1,
199-218.
Bern, D. J. Reply to Judson Mills. Psychological Review.
1967, 74, 536-537. (a)
Bern, D. J. Self-perception: An alternative interpreta
tion of cognitive dissonance phenomena. Psycholog
ical Review. 1967, 74, 183-200. (b)
Bern, D. J. The epistemological status of interpersonal
simulations: A reply to Jones, Linder, Kiesler,
Zanna, and Brehm. Journal of Experimental Social
Psychology. 1968, 4, 270-274.
68
Bern, D. J. Self-perception theory. In L. Berkowitz (Ed.)
Advances in Experimental Social Psychology. Vol. 6
New York: in press.
Bern, D. J., and McConnell, H. K. Testing the self-percep
tion explanation of dissonance phenomena: On the
salience of premanipulation attitudes. Journal of
Personality and Social Psychology, 1970* 14, 23-31.
Brehm, J. W. A dissonance analysis of attitude-discrepant
behavior. In M. J, Rosenberg and C. I. Hovland
(Eds.), Attitude Organization and Change. New
Haven: Yale University Press, i9 6 0.
Brehm, J. W. Motivational effects of cognitive dissonance
In M. Jones (Ed.), Nebraska Symposium on Motiva
tion. Lincoln: University of Nebraska Press, 1962
Brock, T. C., and Grant, L. D. Dissonance, awareness, and
thirst motivation. Journal of Abnormal and Social
Psychology. 1963, 67, 57-60.
Campbell, D. T., and Flske, D. W. Convergent and discrim
inant validation by the multitrait-multimethod
matrix. Psychological Bulletin. 1959* 56, 81-105.
Carlsmith, J. M., Collins, B. E., and Helmreich, R. L.
Studies in forced compliance: I. The effect of
pressure for compliance on attitude change pro
duced by face-to-face role playing and anonymous
essay writing. Journal of Personality and Social
Psychology. 1966, 4, 1-13.
70
Chapanls, N. A., and Chapanls, A. Cognitive dissonance:
Five years later. Psychological Bulletin. 1964, :
61, 1-22.
Cohen, A. R. An experiment on small rewards for discrep- i
ant compliance and attitude change. In J. W. Brehm ;
and A. R. Cohen (Eds.), Explorations in cognitive
dissonance. New York: Wiley, 1962.
Conolley, E. S. Attitudinal effects of commitment to dis
crepant behavior. Proceedings. 78th Annual Conven-j
tion, APA, 1970.
Conolley, E. S., Wilhelmy, R. A., and Gerard, H. B. The
effect of pre- vs. postdecisional justification on ;
subsequent task evaluations. Paper presented at
the meeting of the Western Psychological Associ
ation, San Diego, March, 1968.
Dillehay, R. C., Bruvold, W. H., and Siegel, J. P. Atti
tude, object label and stimulus factors in response
to an attitude object. Journal of Personality and
Social Psychology. 1969, 11, 220-223.
Elms, A. C. Role playing, incentive and dissonance.
Psychological Bulletin. 1967, 68, 132-148.
Festinger, L. An analysis of compliant behavior. In M.
Sherlf and M. Wilson (Eds.), Group behavior at the ;
crossroads. New York: Harper, 1953.
71
Festlnger, L. A theory of cognitive dissonance. Stanford:!
Stanford University Press, 1957.
Festinger, L., and Carlsmith, J. M. Cognitive consequences
of forced compliance. Journal of Abnormal and So
cial Psychology. 1959, 58, 203-210.
Gerard, H. B., Conolley, E. S., and Wilhelmy, R. A. Com
pliance, inconsistency, and attitude change. In
preparation for L. Berkowitz (Ed.), Advances in
Experimental social psychology.
Harvey, J., and Mills, J. Effect of an opportunity to
revoke a counterattitudinal action upon attitude
change. Journal of Personality and Social Psy
chology. 1971, 18, 201-209.
Janis, I. L., and Gilmore, J. B. The influence of incen
tive conditions on the success of role playing in
modifying attitudes. Journal of Personality and
Social Psychology. 1965, 1, 17-27.
Jones, A. R., Linder, D. E., Kiesler, C. A., Zanna, M.,
and Brehm, J. W. Internal states or external
stimuli: Observers' attitude judgments and the
dissonance theory— self-persuasion controversy.
Journal of Experimental Social Psychology. 1968,
4, 247-269.
Kiesler, C. A., Nisbett, R. E., and Zanna, M. On infer
ring one's beliefs from one's behavior. Journal
of Personality and Social Psychology, 1969, 11,
321-327.
Linder, D. E., Cooper, J., and Jones, E; E. Decision
freedom as a determinant of the role of Incentive
magnitude In attitude change. Journal of Person
ality and Social Psychology. 1967, 6, 245-254.
Mannsson, H. H. The relation of dissonance reduction to
cognitive, perceptual, consummatory, and learning
measures of thirst. In P. G. Zimbardo (Ed.),
The cognitive control of motivation. Glenview,
111.: Scott, Foresman, and Company, 1969.
McGuire, W. J. Attitudes and opinions. In P. R. Farns
worth (Ed.), Annual review of psychology. Vol. 1 7,
Palo Alto, California: Annual Reviews, 1966.
Pp. 475-514.
Mills, J. Comment on Bem»s "Self Perception: An .alterna- j
tive interpretation of cognitive dissonance phenom- j
ena." Psychological Bulletin. 1967, 74, 535.
Mills, J., and Harvey, J. Can self-perception theory ex
plain the findings of Harvey and Mills (1971)?
Journal of Personality and Social Psychology. 1972, I
22, 271-272.
Rosenberg, M, J. When dissonance falls: On eliminating
evaluation apprehension from attitude measurement.
Journal of Personality and Social Psychology. 1965, !
Rosenberg, M. J. Hedonism, inauthenticity and other goals
toward expansion of a consistency theory. In R. P.
Abelson et al. (Eds.), Theories of cognitive con
sistency: A sourcebook. Skokie, 111.: Rand-
McNally, 1968.
Winer, B. J. Statistical principles in experimental de
sign. New York: McGraw-Hill, 1962.
Zajonc, R. B. Cognitive theories of social psychology.
In G. Lindsey and E. Aronson (Eds.), Handbook of
social psychology. Vol. 1. (2nd ed.). Reading,
Mass.: Addlson-Wesley, I96 8.
APPENDIX A
The Complete Instructions Booklet for the
"Present" Experiment
74
THE INFLUENCE OF THIRST ON
PERCEPTUAL JUDGMENTS
Instructions Booklet
Please, DO NOT OPEN this booklet until
asked to do so.
75
General Instructions
As psychologists we are interested in studying how j
biological drives affect psychological variables. One very
: basic and important biological drive is that of thirst. ]
Today you will be helping by participating in an experi- j
! ment which is examining the relationship between level of j
thirst and perceptual judgments. Now, because this is an I
' experimental study, it is very important that you follow !
directions carefully and answer all questions accurately J
and seriously.
To assist us in this study you will be performing
a short perceptual task that will take approximately forty j
minutes to complete. Also, several times throughout the j
experiment you will be asked to indicate on various scales ,
! how thirsty you feel. This is necessary because in past j
research we have found that a person's feelings of thirst
change rather rapidly and that people differ in their rates;
of change in thirst feelings. Each time you are asked to j
rate your thirst level, please be as accurate and honest j
as possible as to how thirsty you feel at that very moment.;
Also, remember that it is your individual feelings that we I
; want, so please do not discuss your ratings or judgments !
i with anyone else during the experiment. In other words, !
do not be influenced by nor try to influence anyone else
i who is around you.
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
77 !
As you now know, we are concerned with how thirst
level will affect your performance on the perceptual task ;
! which you will be doing in a moment. Consequently, it is :
extremely important that we have control of how thirsty .
people are when they perform this task. Therefore, we
; are going to ask you to help us by eating the cracker
which is under the paper plate in front of you. By j
doing this, we will be able to have all of you at approx- I
imately the same level of thirst when you begin. j
Now, please remove any gum, candy, or other sub-
: stances which you may have in your mouth. Place them on I
the empty paper plate on' your left. O.K. Now uncover thej
plate on your right and eat the entire cracker that is on j
the plate. ,
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
I
l
78
Ratings of Thirst Level
Now that you have eaten the cracker and we have
waited for your thirst level to somewhat stabilize, it
is necessary for you to rate your feelings of thirst.
There are two different scales on which you are to indi
cate how thirsty you feel at this very moment. One of
the scales consists of two adjectives connected by numbers
and dots from -30 to +30. You use the scale by circling
the number or dot which best expresses your thirst level.
For example, if you felt moderately thirsty, you would
circle a number or dot somewhere toward the "Extremely
Thirsty" end of the scale. However, if you felt fairly
strongly that you were not thirsty, you would circle a
number or dot toward the "Not at all Thirsty" end. The
other scale consists of several different adjectives
which could be used to describe a glass of water. Based
on your current level of thirst, circle the adjectives
which best describe how that glass of water would taste
to you right now. You may want to circle two or more of
the adjectives.
Do not worry over your choices for long periods.
It is your first impression that we want. Also, remember
that it is your own feelings that we want, so please do
not discuss your ratings during the experiment. Further-
more, notice that from time to time the scales will be
reversed such that the end of the scale which stands for
"Extremely Thirsty" will be assigned the number +30 part
of the time and the number -30 part of the time. Now,
if there are no questions you may turn the page and
circle your ratings. Then go on to the following page
and complete the second scale.
Third Rating Scale
Not at all Thirsty Extremely Thirsty
-30 -15 0 +15 +30
GO ON TO THE NEXT PAGE
Adjective Thirst Scale
80
Circle those adjectives which would best describe how a
glass of water would taste.
horrible
unfit
bad
undesirable
poor
inferior
below par
can be tolerated
passable
O.K.
likable
tasty-
desirable
enjoyable
delightful
delicious
DO NOT TURN THE PAGE .UNTIL ASKED TO DO SO
Perceptual Judgement Task
Now you are going to perform the perceptual judge
ment task. I am going to show you a series of slides of
famous paintings, one at a time. After each slide has
been shown, you will be asked to check off on a checklist
all the items that you saw in the painting which was just
shown. There will be ten paintings in allj please use a
separate checklist for each painting. Work quickly,
taking a minute or so to complete the checklist for each
painting. The ten checklists are on the next ten pages.
Remember, work quickly and use a separate checklist for
each painting.
I
I
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
PAINTING CHECKLIST
82
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon
trees
pitcher
musical notes
piano
cart
dress
flowers
fence
face
snow
barn
stars
chair
ladder
house
couch
haystack
glass
candle
mountains
boat
ribbon
dish
water
window bowl
river
fruit
hook
bridge
drapes rag
boat
hat
fish
sun
gloves
vase
man
cow
sunflower
woman
teepees
ink well
dog
roads
table
flag
ring
horns
necklace
book
trumpet
Painting Number
83
PAINTING CHECKLIST
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon
trees
pitcher
musical notes
piano
cart
dress
flowers
fence
face
snow
barn
stars
chair
ladder
house
couch
haystack
glass
candle
mountains
boat
ribbon
dish
water
window
bowl
river
fruit
hook
bridge
drapes
rag
boat
hat
fish
sun
gloves
vase
man
cow
sunflower
woman
teepees
ink well
dog
roads
table
flat
ring
horns
necklace
book
trumpet
Painting Number
84
PAINTING CHECKLIST
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon trees pitcher
iAusical notes piano cart
dress flowers fence
face snow barn
stars chair ladder
house couch haystack
glass candle mountains
boat ribbon dish
water
window bowl
river fruit hook
bridge drapes rag
boat hat fish
sun
gloves vase
man cow sunflower
woman
teepees ink well
dog
roads table
flat ring horns
necklace book trumpdt
Painting Number
PAINTING CHECKLIST
85
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon trees
pitcher
musical notes piano cart
dress flowers fence
face
snow barn
stars chair
ladder
house couch
haystack
glass candle
mountains
boat
ribbon dish
water
window bowl
river fruit
hook
bridge
drapes rag
boat
hat
fish
sun
gloves
vase
man
cow
sunflower
woman
teepees ink well
dog
roads table
flat
ring
horns
necklace book
trumpet
Painting Number
PAINTING CHECKLIST
86
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
.moon
jnusical notes
.dress
face
stars
house
glass
boat
• water
river
bridge
boat
__sun
_man
_woman
_dog
_flat
jiecklace
_trees
_piano
_flowers
_snow
.chair
.couch
_candle
__ribbon
_window
.fruit
.drapes
Jhat
.gloves
.cow
^teepees
.roads
Painting Number
.pitcher
cart
.fence
barn
gladder
_haystack
.mountains
dish
.bowl
.hook
.rag
fish
.vase
.sunflower
.ink well
.table
.horns
.trumpet
PAINTING CHECKLIST
87
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon trees pitcher
musical notes piano cart
dress flowers fence
face snow barn
stars chair ladder
house couch haystack
glass candle
mountains
boat ribbon dish
water window bowl
river fruit
hook
bridge drapes rag
boat hat fish
sun
gloves vase
man
cow sunflower
woman teepees ink well
dog roads table
flat ring
horns
necklace book trumpet
Painting Number
88
PAINTING CHECKLIST
Each painting will be shown briefly. After it is shown*
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
jnoon
_musical notes
dress
trees
_face
_stars
_house
_glass
__boat
_water
_river
_bridge
_boat
_sun
jman
_woman
_dog
_flat
necklace
piano
flowers
_snow
_chair
couch
candle
ribbon
window
_fruit
drapes
hat
gloves
cow
teepees
roads
jring
book
_pitcher
cart
fence
barn
ladder
haystack
jnountalns
_dish
bowl
hook
_rag
fish
_vase
sunflower
ink well
table
horns
trumpet
Painting Number
PAINTING CHECKLIST
89
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space,
_moon
musical notes
dress
trees
_face
_stars
_house
_glass
_boat
_water
jriver
_bridge
_boat
_sun
jman
_woman
_dog
_flat
necklace
_piano
flowers
snow
chair
_couch
candle
ribbon
window
_frult
drapes
hat
_gloves
cow
teepees
roads
_ring
book
joitcher
cart
fence
bam
ladder
haystack
jnountains
dish
bowl
hook
rag
fish
_vase
_s unflower
ink well
table
horns
trumpet
Painting Number
PAINTING CHECKLIST
90
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon
trees
pitcher
musical notes piano cart
dress flowers
fence
face
snow
barn
stars chair
ladder
house
couch
haystack
glass
candle
mountains
boat
ribbon
dish
water
window
bowl
river fruit
hook
bridge drapes
rag
boat hat
fish
sun
gloves
vase
man cow
sunflower
woman
teepees
ink well
dog roads
table
flat ring
horns
necklace book
trumpet
Painting Number
PAINTING CHECKLIST
Each painting will be shown briefly. After it is shown,
please make a checkmark next to each item listed below
that you remember seeing in the painting. Please use a
separate Checklist for each painting, and be certain to
write the number of each painting in the indicated space.
moon trees pitcher
musical notes piano cart
dress flowers fence
face snow barn
stars chair ladder
house couch haystack
glass candle
mountains
boat ribbon dish
water window bowl
river fruit hook
bridge drapes rag
boat hat fish
sun gloves vase
man
cow sunflower
woman
teepees ink well
dog roads table '
•flag ring horns
necklace book trumpet
Painting Number
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
92
Now that we have waited long enough for your thirst
level to truly stabilize, it is necessary to have another
rating of how thirsty you feel. As mentioned earlier, we
have found that people differ in terms of the rate at
which their thirst level changes and also that some people
change one way and some people change the other way.
Therefore, we want you to give us an estimate of how
thirsty you are at this moment. The scales are the same
as before. Remember, make as careful and accurate an es
timate as you can concerning how thirsty you are at this
very moment. The scales are on the next two pages. After
you have completed the scales I will give you a slip indi
cating that you have completed the experiment and that you
are to receive one hour of credit for participation.
PLEASE GO ON TO THE NEXT PAGE
93
ADJECTIVE THIRST SCALE
Circle those adjectives which would best describe how a
glass of water would taste.
delicious
delightful
enjoyable
desirable
tasty
likable
O.K.
passable
can be tolerated
below par
inferior
poor
undesirable
bad
unfit
horrible
PLEASE GO ON TO THE NEXT PAGE
Thirst Rating Scale
Extremely Thirsty
+30 +15
Not at all Thirsty
-15 -30
0
vo
-pr
I
APPENDIX B
Scale Values for the Adjective
Checklist Items
95
Scale Values for Adjective
Checklist Items
Adjective Scale Value
Horrible.................................. 1.24
Unfit . . . . ......................... 1.78
Bad .2.11
Undesirable............................. 2.46
Poor .................................. 3.14
Inferior ............................... 3.54
Below Par ............................. 4.25
Can be tolerated......................... 4.92
Passable ............................... 5-95
O.K. 6.26
Likable ............................... 7.94
Tasty 8.63
Desirable............................... 9.19
Enjoyable............................... 9.45
Delightful............................... 10.00
Delicious............................... 10.57
APPENDIX C
The Complete Instructions Booklet for the
"Future" Experiment for Subjects in
the Low Incentive Conditions
97
THE PSYCHOLOGICAL CONSEQUENCES OP
WATER DEPRIVATION
Please DO NOT OPEN this booklet until
asked to do so.
99
General Instructions
As psychologists we are interested in studying how
biological drives affect psychological variables. One
very basic and important biological drive is that of
thirst. Today you will be helping by participating in an
experiment which is examining the psychological conse
quences of being deprived of water. Now, because this
is an experimental study, it is very important that you
follow directions carefully and answer all questions
truthfully and accurately.
To assist us in this study you will be required to
go without drinking any fluids or eating any food which
contains large amounts of liquids, such as soups and
fruits, for a period of twenty-four hours, beginning as
soon as the instructions are completed. After the 24-
hour period you will return to this room and answer a
series of questions which I will give you at that time.
It is extremely important that you rigidly observe
the water restrictions I have mentioned. Therefore, in a
few minutes I will attach to your arm a small "band-aid"
like piece of tape which has on it a special type of
litmus paper. This litmus paper will tell us two things.
It will check on your liquid Intake by absorbing your
perspiration and it will give us a measure of your reac
tion to the water deprivation. After you take the ques
tionnaire tomorrow I will remove the "band-aid" from your
arm. However, in the mean time, be very careful not to
get it wet while washing your hands or anything else that
you may do near water.
PLEASE DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
100
We realize that it will not be pleasant to go
without water or fluids for twenty-four hours. There
fore, you are completely free to decide for yourself
whether or not you want to participate. I cannot offer
you experimental credit for this experiment. However,
since it is going to be uncomfortable I can offer you
five dollars for participating. Of course it is entirely
up to you. Below is a statement to sign if you are will
ing to help out. Please read it carefully and then sign
it if you are willing to participate.
I ____________________ agree to participate in
the thirst deprivation study for the sum of five dollars
($5). I further agree to go without any liquids for
twenty-four (24) hours, at which time I will return to
this room to complete the study, have the special litmus
"band-aid" removed, and receive my money.
Date
PLEASE DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
101
Before you’ begin your twenty-four hours of depri
vation, you are allowed one last drink of water. You may
drink as little or as much as you like. In front of you
is a pitcher of water and a cup. For sanitary reasons,
please use the straw that is provided. Now, go ahead
and drink all the water that you want.
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
102
Now, the last thing to do for today is to receive
your "band-aid" and then you may leave. I will attach
the band-aid and it is not to be removed for any reason
until I remove it at this time tomorrow. Be extremely
careful not to get it wet while washing your hands or
at any other time you are around any liquids. Also,
remember that you are to return to this room at the same
time tomorrow to complete the experiment and receive*
your money.
Please come up to the desk and I will attach your
band-aid.
APPENDIX D
The Complete Instructions Booklet for the
"Second" Experiment for Subjects in
the High Incentive Conditions
103
io4
THE PSYCHOLOGICAL CONSEQUENCES OP WATER DEPRIVATION
Please DO NOT OPEN this booklet -until
asked to do so.
General Instructions
As psychologists we are interested in studying how
biological drives affect psychological variables. One
very basic and important biological drive is that of
thirst. Today you will be helping by participating in
an experiment which is examining the psychological conse
quences of being deprived of water. Now, because this
is an experimental study, it is very important that you
follow directions carefully and answer all questions
truthfully and accurately.
To assist us in this study you will be required to
go without drinking any fluids or eating any food which
contains large amounts of liquids, such as soups and :
fruits, for a period of twenty-four hours, beginning as
soon as the instructions are completed. After the 24-
hour period you will return to this room and answer a
series of questions which I will give you at that time.
It is extremely important that you rigidly observe
the water restrictions I have mentioned. Therefore, in a
few minutes I will attach to your arm a small "band-aid"
like piece of tape which has on it a special type of litmus
paper. This litmus paper will tell us two things. It
will check on your liquid intake by absorbing your per
spiration and it will give us a measure of your reaction i
to the water deprivation. After you take the question
naire tomorrow I will remove the "band-aid" from your arm. '
However, in the mean time, be very careful not to get it
wet while washing your hands or anything else that you
may do near water.
PLEASE DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
106
We realize that it will not be pleasant to go
without water .or fluids for twenty-four hours. Therefore,
you are completely free to decide for yourself whether or
not you want to participate. I cannot offer you experi
mental credit for this experiment. However, since it is
going to be uncomfortable I can offer you twenty dollars
for participating. Of course it is entirely up to you.
Below is a statement to sign if you are willing to help
out. Please read it carefully and then sign it if you
are willing to participate.
I agree to participate
in the thirst deprivation study for the sum of twenty
dollars ($20). I further agree to go without any
liquids for twenty-four (24) hours, at which time I
will return to this room to complete the study, have
the special litmus "band-aid" removed, and receive my
money.
Date
PLEASE DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
Before you begin your twenty-four hours of depriva
tion, you are allowed one last drink of water. You may
drink as little or as much as you like. In front of you
is a pitcher of water and a cup. For sanitary reasons,
please use the straw that is provided. Now, go ahead and
drink all the water that you want.
DO NOT TURN THE PAGE UNTIL ASKED TO DO SO
108
Now, the last thing to do for today is to receive
your "band-aid" and then you may leave. I will attach
the band-aid and it is not to be removed for any reason
until I remove it at this time tomorrow. Be extremely
careful not to get it wet while washing your hands or at
any other time you are around any liquids. Also, remember
that you are to return to this room at the same time to
morrow to complete the experiment and receive your money.
Please come up to the desk and I will attach your
band-aid

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The effects of physical proximity and body boundary size on the self-disclosure interview

PDF

Attention, retention, and incentive processes in observational learning

PDF

Hypnotic Susceptibility, Achievement Motivation, And The Treatment Of Obesity

PDF

Age, Sex, And Task Difficulty As Predictors Of Social Conformity: A Search For General Tendencies Of Conformity Behavior

PDF

Modification Of Low Self-Confidence In Elementary-School Children By Reinforcement And Modeling

PDF

Changes In Memory As A Function Of Age

PDF

The Effects Of Nonreward Dissonance And Secondary Reward On Extinction And Attractiveness

PDF

The Effects Of Repression-Sensitization, Race, And Levels Of Threat On Extensions Of Personal Space

PDF

The Effects Of Generation, Religion, And Sex On The Relationship Of Family Vertical Solidarity And Mental Health In Lebanon

PDF

Psychomotor Performance And Change In Cardiac Rate In Subjects Behaviorally Predisposed To Coronary Heart Disease

PDF

Weight Reduction As A Function Of The Timing Of Reinforcement In A Covertaversive Conditioning Paradigm

PDF

An Examination Of Positive And Negative Reinforcement In Classical And Operant Conditioning Paradigms In The Primary Psychopath

PDF

Self-Sacrifice, Cooperation, And Aggression In Women Of Varying Sex-Role Orientations

PDF

Prognostic Expectancy Effects In The Desensitization Of Anxiety Over Invasion Of Body Buffer Zones

PDF

Strategies Of Marital Communication

PDF

Instrumental Escape Behavior Applied To Remedial Reading

PDF

The Relationship Of Multidimensional Scaling Spaces Of Trait Adjectives For Different Reference Persons

PDF

A Factor Analysis Of The Semantic-Evaluation Abilities

PDF

Looking Back After Coming Down: Conformity And Commitment In Campus Protest

PDF

Cue Dependent Active Memory And Ecs Produced Memory Occlusion

Asset Metadata

Creator Green, Duane (author)

Core Title Dissonance and Self-Perception Analysis of "Forced Compliance": When Two Theories Make Competing Predictions

Degree Doctor of Philosophy

Degree Program Psychology

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

Tag OAI-PMH Harvest,psychology, experimental

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Conolley, Edward S. (committee chair), Bengtson, Vern L. (committee member), Frankel, Andrew Steven (committee member), Miller, Norman (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-885049

Unique identifier UC11363341

Identifier 7331349.pdf (filename),usctheses-c18-885049 (legacy record id)

Legacy Identifier 7331349

Dmrecord 885049

Document Type Dissertation

Rights Green, Duane

Type texts

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

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

Repository Name University of Southern California Digital Library

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