Professional Documents
Culture Documents
Aronson Marihuana
Aronson Marihuana
In 1959 Festinger and Carlsmith pub- ing. The greater attitude change in the low
lished their then startling findings that in- money condition produces a negative relation-
dividuals who told a person that a dull task ship between attitude change and financial in-
was actually interesting for a $1 bribe were ducement to comply.
more likely to believe it than those who told A great deal of data has flowed over this
the same lie for a $20 bribe. The investigators particular research dam since then; the most
derived this prediction from the theory of accurate general conclusion one can draw
cognitive dissonance in the following way: A from all of these findings is that the cogni-
subject's cognition "I said X" was dissonant tive world is a far more complicated place
with his cognition "I believe not-X." The than Festinger and Carlsmith believed it to
more money he received for lying, the greater be in the good old days of 19S9. Some in-
the justification and hence the less the dis- vestigators replicate the dissonance predicted,
sonance. Thus, subjects in the $1 condition negative relationship (Cohen, 1962; Lepen-
experienced more dissonance than those in dorf, 1964), some find a positive or direct
the $20 condition. In order to reduce dis- relationship (Collins, 1968a, 1968b; Elms &
sonance, these $1 subjects tended to convince Janis, 1965; Hornbeck, 1968; Rosenberg,
themselves that the task really was interest- 1965), and some report no relationship (Ash-
1 more & Collins, 1968; Ashmore, Collins,
The authors wish to thank Barry E. Collins
of the University of California, Los Angeles, for Hornbeck, & Whitney, 1968; Collins & Ash-
his collaboration and support during all phases of more, 1968; Collins & Helmreich, in press).
the research. The research was supported by Con- More recent experiments have helped to
tract N00014-67-A-0126-0001 of the Office of Naval clarify the situation by demonstrating a dis-
Research, Group Psychology Branch, Robert Helm-
reich, principal investigator; Elizabeth Nel's par- sonance effect under one set of conditions and
ticipation was supported by National Science Founda- a positive effect under a different set of con-
tion Grant GS1194, Barry E. Collins, principal ditions all within the same factorial design
investigator. The authors wish to thank Mrs.
Frederick George who ably served as Experimenter ]. (Carlsmith, Collins, & Helmreich, 1966;
2
Now at the University of Stellenbosch, Re- Helmreich & Collins, 1968). In the Carlsmith,
public of South Africa. Collins, and Helmreich (1966) experiment,
3
Requests for reprints should be sent to Robert
Helmreich, Department of Psychology, Mezes Hall subjects either made a 2-minute counter-
211, University of Texas, Austin, Texas 78712. attitudinal statement to a confederate in a
117
118 E. NEL, R. HELMREICH, AND E. AEONSON
direct face-to-face situation or wrote a IS- —rather, dissonance arose between a cognition
minute anonymous, counterattitudinal essay. about the self and a cognition about a be-
In each of these conditions, subjects were havior which violated this self-concept. Thus,
paid either $.50, $1.50, or $5.00 for making Aronson suggested that in the original
the counterattitudinal statement. Where sub- Festinger-Carlsmith experiment, it was inac-
jects wrote anonymous essays there was a curate to state that dissonance occurred be-
positive or direct relation between financial tween the cognitions (a) "I believe the task
inducement and attitude change—the more is dull" and (b) "I said the task is interest-
money subjects were paid, the greater the ing." What was dissonant for most people
opinion change. In the face-to-face situation, (i.e., those with a high self-concept) is the
the opposite effect was obtained—the less cognition (a) "I am a good and decent
money the subjects were paid, the greater human being" and (b) "I have committed
the opinion change. an indecent act; I have misled another
Helmreich and Collins (1968) subsequently person."
produced a similar interaction with a simpler Collins (1969) made a similar point in
manipulation which did not allow such wide suggesting that dissonance occurs in situations
between-condition variation in the nature of which threaten self-esteem or produce ob-
the counterattitudinal task. In all conditions jective negative consequences for the self or
the subjects made an oral, counterattitudinal the audience.
statement. But in one condition, it was in Bramel (1968) argued that
the form of an audio tape recording in Dissonance is a feeling of personal unworthiness (a
which they were allowed to remain anon- type of anxiety) traceable to rejection of oneself
ymous ; in other conditions, the statement was by other people either in the present or in the past.
recorded on video tape and subjects were Any information which implies that one is in-
clearly identified by name. In the anonymous competent or immoral arouses dissonance. The reason
dissonance is greatest when the person feels per-
condition, there was a nonsignificant positive sonally responsible for his behavior is that re-
trend, while in the nonanonyinous conditions, jection by other people is usually greatest when they
there was a dissonance effect. believe the person voluntarily acted in an inap-
What is the conceptual significance of these propriate way [p. 365].
manipulations? In other words, what is the Each of these three interpretations of dis-
process involved in making a public, counter- sonance theory stresses the dissonance-arous-
attitudinal statement that results in a dis- ing characteristics of discrepant cognitions
sonance phenomenon? Subsequent investiga- about the self. Support for the importance
tion fails to support the notion that public of the self-concept in dissonance induction
statements produce dissonance while private can be found in the results of several recent
ones do not. In several studies public, high- studies.
commitment essays produce significant anti- In the previously mentioned Helmreich and
dissonance, positive relationships between Collins (1968) study, there were two con-
financial inducement and attitude change ditions in which subjects made public video
(Collins, 1968a, 1968b; Hornbeck, 1968). tapes advocating a counterattitudinal posi-
The public-anonymous variable, then, is not tion. In the condition designed to maximize
in and of itself sufficient to explain the data. dissonance, the subject not only took the
In a recent attempt to refine and clarify counterattitudinal position publicly, but
dissonance theory, Aronson (1968, 1969) has agreed to stand by his position for a period of
suggested that the crucial determinant is several months. The second video condition
whether or not the counterattitudinal be- was designed to minimize dissonance; in this
havior threatens to diminish the individual's "take-back" condition subjects made the
self-concept. More specifically, according to identical video tape but also made a second
Aronson, in the clearest experiments per- tape explaining that their counterattitudinal
formed to test dissonance theory, dissonance advocacy was only for the purpose of an
did not arise between just any two cognitions attitude-change experiment and could then
OPINION CHANGE AS A FUNCTION OF AUDIENCE PERSUASIBILITY 119
express their true opinions on the issue. showed more attitude change when given low
The authors predicted an inverse (dissonance) financial incentive for their compliance than
relationship between incentive and opinion those given larger incentives. On the other
change for the first condition. In the "take- hand, when decision freedom was reduced, a
back" condition they expected either a highly positive relationship between incentive and
attenuated dissonance effect or an incentive attitude change was found. It is easy to see
effect. how the self is not really involved unless the
The results, however, showed significant subject himself has chosen to participate. If
dissonance effects in both conditions. How the subject tells a lie because the experi-
can one interpret the dissonance effects in the menter implies that these are the rules of
take-back condition? It is possible that com- the game, the act probably does not threaten
mitting the negative (immoral) act of publicly his self-esteem. But if he himself chooses to
adopting a counterattitudinal position for in- tell the lie, then, again, dissonance arises be-
sufficient justification lowers the self-concept tween the cognition of a generally high self-
and that taking it back is not sufficient concept and the cognition that he chose to lie
to restore a positive self-concept because the to another person.
act itself cannot be undone. It is also pos- Although a number of alternative explana-
sible that the "harm had already been done" tions exist for all of these data, the self-con-
and the subjects felt they could not undo cept notion seems to merit more direct in-
the negative consequences for others that they vestigation. How can we further test the
produced with their initial counterattitudinal efficacy of this explanatory concept? One
act. reasonable strategy would be to use the self-
Two conditions in a study by Davis and concept formulation to predict a totally differ-
Jones (1960) are relevant to the discussion ent type of outcome within the framework of
of "taking back" a negative act. In this the counterattitudinal advocacy paradigm.
experiment, all subjects read a negative Specifically, since this notion rests upon the
evaluation of a stimulus person over a micro- assumption that choosing to mislead is the
phone presumably broadcasting to this in- major source of dissonance, under what con-
dividual. In the two crucial conditions, where ditions would this be most disastrous to the
the subjects perceived having chosen to eval- self? One important consideration involves
uate the stimulus person negatively, those the consequences of this advocacy on the
who did not anticipate personal interaction audience. If an individual makes a counter-
with the stimulus person evaluated him more attitudinal statement to an audience which is
negatively than those who expected to meet not persuasible, he should experience less
with him. The authors interpret this finding dissonance than if he made the same state-
as indicating that less dissonance is aroused ment to a completely naive and gullible
where the person anticipates being able to audience. For example, suppose a person who
undo his negative act through personal en- is opposed to the use of marijuana were asked
counter. In the light of the Helmreich and to make a statement advocating its use. Sup-
Collins data, it seems reasonable to assume pose further that he makes the statement to
that individuals feel far more capable of an audience consisting of members of the
undoing a wrong if they have the opportunity vice squad who are irrevocably committed to
to chat with the victim in a face-to-face, a position opposing the use of marijuana.
give-and-take encounter than in a situation Although the speaker might anticipate some
which involves one recorded disclaimer of audience hostility, there are no negative con-
responsibility. sequences for the audience in counteratti-
Another recent study by Linder, Cooper, tudinal advocacy in this kind of situation as
and Jones (1967) emphasized the importance (hey are unlikely to be changed by the com-
of choice in dissonance induction. Subjects in munication. Similarly, if the individual is
this study who felt free not to comply with a asked to make the same statement to a group
request for counterattitudinal advocacy of Haight-Ashbury potheads, there will not
120 E. NEL, R. HELMREICH, AND E. ARONSON
tion of marijuana on the second page. you make in an effort to persuade them even more
Because it was vital to collect the posttest after that there should be no legal restrictions on the
(he experimental manipulation, a second deception use of marijuana for people over 21.
was introduced at this point. The opinion poll was
collated in such a way that subjects received two 3. Subjects in the con audience condition were given
copies of the first page and no second page. Most the same information as those in the uncommitted
subjects realized that the poll was miscollated as audience with the exception of the second sentence
soon as they turned the page, and called this to which read: "We've already asked them once what
the experimenter's attention. If subjects filled out their opinions are and our results show that they
the first page again, Experimenter 1 allowed them to are against the legalization of marijuana for people
complete it and "noticed" the mistake while check- over 21."
ing the form. As soon as Experimenter 1 "noticed" Just before the video tape was made, subjects
the miscollation, she asked the subject to excuse her were again reminded of the position of the in-
while she went to the other psychology laboratory tended audience on the issue. They were asked to
building to get a second page. When going out the begin the recording by identifying themselves by
door, Experimenter 1 turned back and suggested that name and giving their home town, class, and major
the subject could save some time by helping Experi- before presenting the arguments. The points on the
menter 2 while Experimenter 1 was getting a new outline page given subjects were:
test. This seemed like a natural suggestion and a few 1. Should be no legal restrictions on the use of
subjects spontaneously suggested this procedure. marijuana for people over 21—mature in the eyes
Experimental manipulations. On the subject's ar- of the law—can take responsibility for own actions.
rival at the second experimental room, Experi- Restricting rights of mature people violation of
menter 2 explained that the research she was plan- concept of individual freedom in democratic
ning to conduct was on attitude change. The pro- society.
cedure then varied according to condition. 2. Marijuana minor problem compared with
1. Subjects in the uncommitted-audience condition alcohol. Has received disproportionate share of
were told: publicity regarding crime.
3. Relation of marijuana to psychosis—no scien-
We're preparing materials to be used in research tific proof.
on the attitudes of a large group of students about 4. Marijuana docs not lead to addiction, is less
the use of marijuana. We've already asked them habit-forming than ordinary tobacco, no with-
once what their opinions are and our results show drawal symptoms.
that they really don't have any opinion about it 5. Prohibition has negative effects—more exciting
and that they don't even know very much about for people, promotes blackmarketing, selective law
it. In this experiment we are going to use the enforcement.
video tape you make in an effort to persuade
them that there should be no legal restrictions on As soon as subjects had made the video tape, it
the use of marijuana for people over 21. After was played back to them and they were asked to
playing the tape to them, we're going to measure indicate how sincere and persuasive they seemed
their attitudes again to see if we could influence to be. After this, they were thanked for their help
them in any way and then we will explain the and sent back to Experimenter 1. Experimenter 1
true purpose of the experiment to them. As you then gave them the second page of the opinion poll
will see, there are five points here. [Experimenter 2 which contained the dependent variable embedded
then handed the subject an outline.] They are not among other questions.
full sentences and I want you to expand on them After the dependent variable had been collected,
in your own words. We could have hired an actress the subjects were questioned as to any suspicions
to do it, but we want to have it sound like an about the procedure and the true purpose of the
ordinary student. I was going to pay this girl experiment was explained to them.
anyway and I'm willing to pay you $.50 [$S] for
helping me out. Are you sure you really want to RESULTS AND DISCUSSION
do it? [Freedom of choice was stressed here and
at this point one subject refused to comply.] Means for attitude change are presented in
Thank you. Work on them for a few minutes, Figure 1. The associated analysis of variance
add any additional arguments you like and write
down your own comments if you want to. is shown in Table 1. The factorial analysis of
variance does not reach conventional levels of
2. Subjects in the pro audience condition were given significance, but some of the a priori com-
the same information with the exception of the
second and third sentences which read: parisons parallel to those computed in Helm-
reich and Collins (1968) are significant. The
We've already asked them once what their opinions
are and our results show that they are in favor of comparison between the $.50 and $5 uncom-
the legalization of marijuana for people over 21. mitted-audience conditions is significant at
In this experiment I'm going to use the tape the .05 level (F = 5,34, df = 1/36). Subjects
122 E. NEL, R. HELMREICH, AND E. AKONSON
TABLE 1
ANALYSIS OF VARIANCE FOR ATTITUDE CHANGE
Source MS di F
DAVIS, K. F., & JONES, E. E. Changes in interpersonal HOKNBECK, F. W. Studies in forced compliance: IX.
perception as a means of reducing cognitive dis- The effects of deception, commitment, and incen-
sonance. Journal oj Abnormal and Social Psy- live on attitude change produced by the writing
c.hology, 1960, 61, 402-410. of a contrattitudinal essay. Paper presented at
ELMS, A., & JANIS, I. L. Counter-norm attitudes meeting of Western Psychological Association, San
induced by consonant versus dissonant conditions Francisco, May 1967.
of role-playing. Journal of Experimental Research LEPENDORF, S. The effects of incentive value and
in Personality, 1965, 1, 50-60. expectancy on dissonance resulting from attitude-
FESTINGER, L., & CARLSMITH, J. M. Cognitive con- discrepant behavior and disconfirmation of ex-
sequences of forced compliance. Journal of Ab- pectancy. Unpublished doctoral dissertation, State
normal and Social Psychology, 1959, 203-210. University of New York at Buffalo, 1964.
HARVEY, 0. J. Conceptual systems and attitude LINDER, D. E., COOPER, J., & JONES, E. E. Decision
change. In M. Sherif & C. W. Sherif (Eds.), At- freedom as a determinant of the role of incentive
titude, ego-involvement, and change. New York: magnitude in attitude change. Journal of Per-
Wiley, 1967. sonalily and Social Psychology, 1967, 6, 245-254.
HELMREICH, R., & COLLINS, B. E. Studies in forced ROSENBERG, M. J. When dissonance fails: On
compliance: Commitment and magnitude of in- eliminating evaluation apprehension from attitude
ducement to comply as determinants of opinion measurement. Journal of Personality and Social
change. Journal of Personality and Social Psy- Psychology, 1965, 1, 28-42.
cliology, 1968, 10, 75-81. (Received August 12, 196S)