University of Southern California Dissertations and Theses

Effects of threat and self-focus on consensual bias in majorities

(USC Thesis Other)

Effects of threat and self-focus on consensual bias in majorities

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zeeb Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
EFFECTS OF THREAT AND SELF-FOCUS ON CONSENSUAL BIAS IN
MAJORITIES
by
Jared Bowden Kenworthy
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
Master of Arts
Psychology
December 1998
Copyright 1998 Jared Bowden Kenworthy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI N u m b e r : 1394792
UMI Microform 1394792
Copyright 1999, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY O F S O U T H E R N CA LIFO RN IA
TH E GRADUATE SC H O O L
U N IV ER SITY FARK
LOS A N G ELES. C A LIFO R N IA 9 0 0 0 7
This thesis, written by
Jared_ Bowden^ Kenworthy __ _ _
under the direction of Aii?----Thesis Committee,
and approved by all its members, has been pre
sented to and accepted by the Dean of The
Graduate School, in partial fulfillment of the
requirements for the degree of
* *.
D»§m
- ____________
D afe N ovem ber 3 0 , 1998
THESIS' COMM!
t
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Page
LIST OF TABLES AND FIGURES........................................................... iii
Section
1. INTRODUCTION
Overview of the False Consensus Effect............... 1
Threat as a Variable................................................. 3
Self-Focus as a Variable.......................................... 3
Experimental Predictions........................................ 4
2. METHOD
Participants............................................................... 9
Procedure
Phases of the Study...................................... 10
Dependent Measures.................................... 13
3. RESULTS
Introduction/ Overview...............................................18
Manipulation Checks...................................................18
Effects of Variables of Estimates of Consensus 20
Co-variation Analyses.................................................23
Consensual Bias for Other Issues Assessed...............23
Ancillary D ata............................................................. 25
4. DISCUSSION
Choice of Issue to be Assessed...................................25
Discussion of Findings............................................ 28
Discussion of “Change” Scores..................................31
Further Research...................................................... 32
5. REFERENCES / BIBLIOGRAPHY........................................34
6. APPENDIX / Dependent Measures................................. 39
7. AUTHOR’S N O TE.................................................................. 43
8. FIGURES AND TABLES........................................................44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES AND FIGURES
Table / Figure Page
Figure 1. Effects of Threat and Self-Focus on First Estimate of Social
Consensus..............................................................................44
Figure 2. Interaction between Threat and Self-Focus on Second Estimate
of Consensus.......................................................................... 45
Figure 3. Mean Change from First Estimation to Second Estimation.
Numbers reflect means of raw differences between first
and second percentages given...............................................46
Table 1. Actual group sizes, consensus estimate averages, and degree
of consensual bias................................................................. 47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
Participants estimated consensus on several issues, one of which was the central
issue for the study. Two key variables (Threat and Self-Focus) were examined as to
their effects on estimates of consensus. Threatened participants were told that their
written justification for holding their opinion would be judged by others advocating the
opposite position. Self-Focused participants were seated in front of a mirror and a
recording video camera for the duration of the study. There was a main effect for
Threat, with threatened participants giving higher estimates of consensus. There was
also, more importantly, a significant interaction between the two experimental variables.
There was no main effect for Self-Focus. All participants in the final analysis were in
the majority on the central issue. Empirical support for the “Golden Section” (Gross &
Miller, 1997) was also found.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Effects of Threat and Self-Focus on Consensual Bias in Majorities
In mid-May of 1998, the government of India performed a series o f underground
nuclear weapons tests after numerous public threats to do so. In addition to mild global
distress and worry over the incident, this prompted India’s rival neighbor, Pakistan, to
announce several nuclear tests of its own. But surprising in this news story is not the
fact that India’s nuclear test threatened to begin an arms race, but rather that no other
country attempted to dissuade the Indian Bharatiya Janata Party (BJP) from executing
the tests in light of their very public warnings that they intended to do so.
In the United States, the CIA conducted a probe to find out why American
Intelligence operations and spy tactics failed to correctly monitor and pursue a
prevention to the nuclear crisis. The consensus within the American government was
that this was indeed a serious Intelligence blunder, perhaps the worst gaffe in the past
ten years (Wolf, 1998). But the main question asked why this bungle occurred. CIA
officials attributed the foul-up to one, bottom-line issue: American Intelligence agents
had an “everyone-thinks-like-us” mindset (Wolf, 1998). Thus, when BJP officials
announced nuclear tests, American Intelligence agents, as well as policy-makers
“misread the BJP as U.S.-style politicians unlikely to follow through on campaign
promises when confronted with all the costs” (Wolf, 1998).
This “everyone-thinks-like-me” mindset is the layperson’s term for what is
referred to as the False Consensus Effect (hereafter, FCE; Ross, Greene & House, 1977;
see also Marks & Miller, 1987), or, more generally, consensual bias. The FCE can be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
described as misrepresented (overestimated and/or underestimated) perceptions of
social consensus (Gross & Miller, 1997; Mullen & Hu, 1988), as in the case of
American Intelligence thinking that all politicians are like their own (American)
politicians.
But, more specifically, the FCE is a “comparison between the estimates of the
prevalence of a single choice option, made respectively by those endorsing and those
rejecting an opinion” (Gross & Miller, 1997; italics added). So, the FCE, per se, is
measured by taking both majority and minority opinions and asking for their estimates
of consensus. The degree to which all persons polled err in their judgments, whether by
overestimation or by underestimation, is the degree of false consensus. The FCE can be
said to be a phenomenon comprising a subset of general consensual bias.
Consensual bias comes in different forms. The two main outfits that consensual
bias may be seen wearing are minority overestimation and majority underestimation.
Mullen and Hu (1988) found that the degree or magnitude of minority (< 50%)
overestimation is generally greater than the degree or magnitude of majority (> 50%)
underestimation (see also Krueger & Clement, 1997). Gross and Miller (1997)
demonstrated that there is a “Golden Section” or cut-point at which overestimation
ceases and underestimation begins. Contrary to the assertion of Mullen and Hu (1988)
that overestimation ceases and underestimation begins at a group size of 50%, a meta-
analytic review (Gross & Miller, 1997) of 128 false consensus effects showed that cut
point, or “Golden Section,” to be at a real group size of 61.8%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
Rather than continue the empirical verification of minority overestimation,
majority underestimation, or even the “Golden Section,” the aim of the current study
was to search for explanatory variables within a consensual bias paradigm. It was the
purpose of this study to examine two variables as to their effects on estimates of
consensus. There are many theoretical explanations for consensual bias (see Dawes,
1989; Gross & Miller, 1997; Krueger & Clement, 1997; Marks & Miller, 1987; Mullen
& Hu, 1988; Ross, Greene & House, 1977), including cognitive and motivational, and
both complex and simple, accounts for the phenomenon. This study seeks to examine
the effects of social Threat and Self-Focus on consensual bias.
As a variable, “Threat” is operationalized here as the imposition of the
anticipation of an unpleasant and potentially hostile social interaction. Specifically,
participants were told that they would fill out a questionnaire asking for their opinions
(yes or no) on several different issues. They were told that, for one of the issues in
particular (#5 on the questionnaire), they would later write out a defense of the position
they took. That defense would be presented to, reviewed, examined and judged by a
group of students holding the opposite position. That group of students, they were told,
would also make some personality judgments about them. They were also told that all
of this information would be compiled and available for them to see at a later date.
The Self-Focus manipulation used in this study consisted of participants being
seated in front of a large mirror with a recording video camera aimed at them from
slightly above head-level and to their right.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The opinion issue used to assess consensus estimates (with respect to the
experimental manipulations) in the current study was the question, “Are you in favor of
raising the current voting age to 25 ?’1
Predictions
Threat. It bears repeating, at this point, that the purpose o f this study was not to
find out whether consensual bias exists, per se, since there is ample evidence of that
phenomenon (Gross & Miller, 1997; Krueger & Clement, 1997; Marks & Miller, 1988;
Mullen & Hu, 1988), but rather to find out under what conditions that bias may be
changed or influenced in one direction or the other. What conditions, besides
considerations of the “Golden Section,” might lead to minority underestimation or
majority overestimation? Is such an extreme phenomenon (minority underestimation or
majority overestimation) even possible? In a more conservative query, one might ask
what conditions might lead to a decrease in minority overestimation or a decrease in
majority underestimation. In fact, this study seeks to answer the latter question;
specifically, when, i f ever, might majority underestimation bias be decreased?
lu the literature on intergroup behavior, there is evidence that majorities may feel
threatened by minorities (Baron, Kerr & Miller, 1992; Bettencourt, Charlton &
Kemahan, 1997; Ng & Cram, 1988; Tajfel, 1978), thus possibly leading to a motivation
to assuage that anxiety. Perhaps one way to ease that anxiety would be to estimate
greater consensus for one’s own position. In theory, by increasing the subjective,
1 It was decided that this particular issue would be used after pilot studies indicated that another issue
(“Do you consider yourself to be pro-choice?”) used in the same experimental paradigm was impervious
to the current manipulations. Please see the discussion section for further discourse on this matter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
perceived size of one’s (majority) ingroup, the (minority) outgroup becomes smaller,
and thus less threatening. Strumpfer (1970) found that people who are afraid strive for
affiliation with others (see also Schacter, 1959).
In the current experiment, because affiliation with others was not made possible,
it is a logical conclusion to expect participants to perceive that their group size is bigger
than they would normally perceive it to be, when made to feel anxious or threatened.
Janis (1968) suggests that the threat of external danger elicits a need for reassurance and
group cohesiveness, which in turn will assuage fears. In their 1987 study, Suls and Wan
collected estimates of consensus from participants who classified themselves as high or
low fear people (fear as a trait, not as a manipulation), and found that high-fear
participants tended to overestimate the degree of consensus on fear-relevant statements.
Sherman, Presson, Chassin, Corty, and Olshavsky (1983) discovered that
adolescents who smoke (and who consider smoking to be a deviant behavior)
overestimate the degree of peer consensus more than adults (who do not consider their
smoking to be deviant) do. This suggests that those individuals concerned with social
approval, vis k vis the behavior, are more motivated to perceive more social support for
their behavior.
Similarly, according to Thompson, Kohles, Otsuki and Kent (1997), individuals
faced with a threatening attitude statement about their ingroup membership perceived
their ingroup to be more homogeneous, thus satisfying the need to feel supported. Crano
(1983) asserted, based on his research, that the consequences (positive or negative) of
holding an opinion are a key factor in the motivation for assumed consensus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
Another theoretical position to be considered is Brewer’s (1991) Optimal
Distinctiveness Theory (hereafter, ODT). The ODT suggests that perceptions of the self
come about byway of two opposing processes: assimilation and differentiation (Brewer,
1991). Assimilation can be described as the blending in, of the self, with a group.
Differentiation can be described as just the opposite; namely, the process of making
oneself distinct from a group. At opposing ends of a continuum are complete de
individuation at one extreme, and complete distinctiveness at the other extreme. Any
move toward an imbalance of these opposing states and pressures will motivate the
individual to move in an equal and opposite direction on the scale of social being
(Brewer, 1993).
In terms of the current manipulations, if Brewer’s (1991, 1993) theory holds,
Threat should constitute a move toward distinctiveness (with negative focus on the self
as an individual), and thus should motivate the individual to assimilate. As previously
mentioned, no physical persons are present with the participant in this study, and so
assimilation can only be carried out by a subjective perception of the size of the
participant’s positional ingroup. ODT should predict that majority members, when
threatened, will decrease the degree to which they underestimate consensus.
Presumably, ODT explains majority underestimation per se by calling attention to the
fact that when majority members voice their opinions, they must somehow think or
realize that “everyone-thinks-like-me” and this constitutes a psychological move toward
de-individuation. A move toward de-individuation will motivate individuals to strive for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
distinctiveness. Perceiving the group size to be smaller than reality (underestimation) is,
in essence, saying, “Many people think like me, but not that many.”
With these considerations, I postulate that the Threat manipulations will decrease
the degree to which majority participants underestimate consensus for their position. In
essence, the Threat manipulation is designed to be a motivational factor moderating
consensual bias. As has been suggested by Marks and Miller (1988), perceived
similarity may buttress and reinforce perceived social support, validate the correctness
of a position, maintain self-esteem, or reduce the tension associated with anticipated
social interaction. The current Threat manipulation is predicted to motivate participants
to perceive increased similarity, and thus decrease underestimation.
Self-Focus. It has been asserted and empirically demonstrated that Self-Focus is
an aversive cognitive state (Carver & Scheier, 1981; Duval, Duval & Mulilis, 1992;
Duval & Wicklund, 1972). Failure can both increase the desire of both depressed and
non-depressed patients to Self-Focus (Greenberg and Pyszczynski,1986; Pyszczynski &
Greenberg, 1987). Depressed patients also persist an abnormally long time in Self-
Focusing behavior (Greenberg and Pyszczynski, 1986), suggesting the negative nature
and valence of Self-Focus. Wegner and Giuliano (1980) go so far as to suggest that
although it has been found that Self-Focus induces arousal, arousal may also induce
Self-Focus.
Geller and Shaver (1976) showed that self-referent words used in a Stroop-like
color-naming task produced substantially longer latencies in color naming. This suggests
that self-referent information is important enough to grab and slow the cognitive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
attentional processes. In these Stroop-like tasks, self-referent words behave much like
negatively valenced words (which produce long latencies as well; see Pratto & John,
1991), suggesting a link between Self-Focus and a negative cognitive state.
Duval, Duval and Mulilis (1992) demonstrated that high levels of Self-Focus led
participants to withdraw and avoid a cognitive task (see also Duval & Wicklund, 1972)
relevant to that state of Objective Self-Awareness (OSA).
Since the default focusing (self versus other) response of majority members is to
focus on the smaller of the two subgroups (i.e., the minority, or othei~, see Bettencourt,
Charlton & Kemahan, 1997; Brewer, 1979; Brewer & Miller, 1984, Duval & Duval,
1983; Duval and Wicklund, 1972; Mullen, 1987; Noel, Wann, &Branscombe, 1995), it
is interesting to wonder what would happen when majority members were made to focus
on themselves instead of the salient minority in an intergroup context. This is, in fact,
precisely the question asked in this study. Of interest is the effect o f Self-Focus on
participants’ (majority) estimates of consensus.
It is predicted that because Self-Focus produces an aversive cognitive state, it
should logically and accordingly elicit a response that would alleviate that aversive state.
Within the experimental context of this study, one behavioral response they have
available is making public consensus estimates. And a way to alleviate the aversiveness
of Self-Focus, in terms of consensus estimates, is to try to, in Brewer’s (1991, 1993)
terms, assimilate. In other words, Self-Focus should produce a decrease in the amount
of majority underestimation, leading to a greater (percentage) estimate.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
It is also predicted that coupling Threat and Self-Focus together will produce an
additive effect, with respect to the main dependent measure; namely, perceptions of
consensus estimates.
Thus, in this 2 X 2 (Threat / No Threat versus Self-Focus / No Self-Focus)
between-subjects experiment, (a) the control condition (No Threat / No Self-Focus)
should produce the lowest mean estimates of consensus (the greatest overall majority
underestimation); (b) the Threat / Self-Focus condition should produce the highest
estimates of consensus (the least overall majority underestimation); (c) the Threat-alone
and Self-Focus alone conditions should both be greater than the control condition, but
not significantly different from each other.
Method
Participants
Ninety-seven (97) volunteers from various undergraduate psychology classes, at
the University of Southern California, were used in this study and were given partial
course credit for their participation. Participants’ ages ranged from sixteen to twenty-
seven (M = 19.1), and less than 2% of participants were above the age of twenty-five.
Participants were eliminated from the analyses of this study because (a) they
were in the minority as to the central issue, or (b) because they were given a central
issue that was later dropped from the study (see discussion), (c) they failed to follow
instructions, or (d) they were suspicious of the methods of the study. In the end, the data
from sixty participants were analyzed and reported here.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
All participants were randomly assigned to experimental conditions before
arriving at the experiment, and each condition of the experiment included only one
participant. There were sixty participants (47 female and 13 male) in the final analysis.1
Procedure
Before beginning replications of this experiment, questionnaires were distributed
to approximately 150 students, on two occasions (total of approximately 300 students) in
two introductory psychology courses. The questionnaires assessed real population
distributions of how many people were in favor of, or against, several issues. From
these data, the central issue of the current study was determined by choosing an issue
with a large majority/minority split in the observed population. Raising the voting age
to 25 had an average of 87.3% opposed and 12.7% (n = 302) in favor of it. Thus, this
was the issue chosen for the current study. Notably, these results are also in accord with
the distribution of the population opposing and agreeing with this issue (87% and 13%,
respectively), as found by Travers (1941) and reported in Gross and Miller (1997).
Upon arrival to the study, called “Personality and Impression Formation,”
participants were seated at a table in a small room (approximately 6' X 91 ). Participants
in the Self-Focus conditions were facing a mirror (a reflective two-way window
common to psychology departments). Self-Focus participants also had a running
(recording) video camera aimed at them for the duration of the study. The location of
the camera was to the right of participants and slightly above head-level, approximately
1 15 participants dropped for (a) above; 15 participants dropped for (b); 1 participants dropped for (c); 6
participants dropped for (d). 8 participants were dropped from the Threat / Self-Focus condition; 10
participants were dropped from the Threat / No Self-Focus condition; 9 participants were dropped from
the No Threat / Self-Focus condition; 10 participants were dropped from the No Threat / No Self-Focus
condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
four feet away. In the non-Self-Focus conditions, a large, white sheet of poster-board
covered up the mirror and there was no video camera present.
After signing in (to receive participation credit), participants were informed of
the (alleged) purposes of the experiment. They were told that the study concerned how
personality and impression formation are affected by the opinions we hold.
They were told that they were going to first fill out some forms, one of which
would be a questionnaire asking whether they were in favor of, or against, several issues
such as abortion, socialized medicine, and raising the voting age to 25, one of which
(raising the voting age to 25, they were correctly informed) was the central issue of
interest in the study. After filling out the questionnaire, they were told that there was
going to be a second part to the experiment.
Participants in the Threat conditions were told that the second part of the
experiment involved their writing down a justification for holding the opinion that they
held with respect to the central issue, raising the voting age. They were told that their
justification would be read, reviewed and critiqued by a group of students, in another
experiment, who held the opposite position.
This group of students, they were told, was also going to form impressions about
the participant and make some general personality judgments about them. Participants
were told that this review of their justification, along with the personality judgment,
would be available to them at a later date (upon completion of the other study) if they so
desired it. They were told that, if they did want a copy of the review and personality
judgment, they could write down their name and address after the experiment was over.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
Most participants expressed interest in receiving these items.2 They were told that, after
their written defense was completed, the experiment would be over and they would be
able to leave.
Participants in the non-Threat conditions were told that the second part of the
experiment entailed their watching a five-minute video clip portraying a group of
students informally discussing the main issue (raising the voting age to 25). They were
told that, after the video was finished, they would fill out a short questionnaire regarding
their impressions of the students on the video. After the short questionnaire was
completed, they would be able to leave.
Once the procedure of the experiment had been satisfactorily explained to
participants, the experimenter said to those in the Self-Focus conditions, “You might be
wondering what the video camera is for. Well, the department has a policy of
monitoring the experiments that are run here. They want to see just what we are doing
in our studies. Since they can’t monitor all of them, they have chosen to randomly
select a few to monitor by way of video recording. Has this happened to you already in
another experiment? (answer: usually ‘no’) Well, once they finish their evaluation, they
erase all these tapes, so you don’t have to worry about being on file somewhere.
Anyway, the entire session will be recorded.”
All participants agreed to the procedure and had no further questions at this
point. The experimenter told each participant that he would return shortly with the
materials needed for the study, and then he left the room. Before leaving, he asked
2 At the aid of the experiment, and during the de-briefing session, participants were informed that, for
obvious reasons, their address would no longer be needed as the personality judgment was a fiction.
Most participants expressed disappointment at this announcement
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
participants to put away all materials and keep them out of sight for the remainder of the
session, to avoid distractions. All participants agreed. The experimenter then left for
approximately one minute, to allow the Self-Focus manipulation to begin to have its
predicted effect.
Participants in the non-Self-Focus conditions were told nothing about a video
camera (since no camera was present) after the procedures were explained. The
experimenter told them that he would return shortly with the materials needed for the
study, and then he left the room. Before leaving, he asked them to put away all
materials and keep them out of sight for the remainder of the session, to avoid
distractions. All participants agreed. The experimenter left for approximately one
minute, to parallel his time of absence in the Self-Focus condition.
When the experimenter returned he gave the participant a packet of forms, and
explained the alleged purpose of each one. The first two forms (see forms #1 & #2,
Appendix) were in reality measures to assess the effects of the Self-Focus
manipulations. However, participants were told that the first two forms were to assess
their cognitive activities, and that this is an important indicator of personality styles. For
the first form participants were instructed to write anything and everything that came to
their minds in two minutes, without regard to any subject matter. The second form had
fifteen sentences on it (see Exner, 1973), all of which were missing a word. They were
given three word choices for each sentence, and were told that all three choices were
equally correct but they were instructed to pick the word they thought was most
appropriate. The third form was the main questionnaire (see form #3, Appendix), asking
for their opinions (yes or no) on several different issues. The experimenter circled the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
number in front of the central issue (raising the voting age) and reminded the participant
that it would be the issue on which they would be defending themselves to opposing
students (Threat conditions), or about which they would be watching the short video
(non-Threat conditions). After each question on this form, there was a question asking
what percentage of other people they thought would answer the question the same way
they did. This was the first estimation of consensus assessed in this experiment.
The fourth and final form at this stage assessed the importance of each issue to
the participants and the extent to which they had thought about each issue (see form #4,
Appendix). These measures were assessed as possible co-variates in the data analyses.
Also included on this form was a check on the Threat manipulation. Threatened
participants were asked how they felt, on five adjective scales, about being evaluated on
their opinions by those who disagree with them. Non-threatened participants were given
the same adjective scales, but were asked how they felt about watching the video.
When all four forms had been adequately explained, the experimenter briefly re
explained each form, and the order in which they were to be filled out. Then the
experimenter reminded the participant that he or she had two minutes to complete the
first form, and that none of the other forms had time limits. There was an electronic
timer on the table that the experimenter set for two minutes. He explained to each
participant that when they were finished with all four forms, they were to ring the bell,
which was also located on the table. The bell was for the purpose of allowing the
experimenter to be absent while the participant completed the forms, but also to know
when they were finished so the experiment could resume in a timely fashion.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
When the participant had no more questions about the process, the experimenter
started the two-minute timer and left the room to allow the participant to begin his or her
work.
Upon hearing the bell ring from the experimental room, the experimenter
returned with two more forms required for the experiment. He explained to participants
that the next brief part of the experiment concerned how holding an opinion differs, and
affects personality in different ways depending on whether you are in the majority or the
minority fo r that issue. He then explained to participants that those who are in favor of
raising the voting age are in the minority, and that those who oppose raising the voting
age are in the majority. He emphasized that the definition of majority is between 50%
and 100% of the population, and that the definition of minority is between 0% and 50%
of the population. He then told each participant whether they were in the majority or
minority3 by asking them how they had answered (in favor of, or against) the central
question.
He then explained to participants that he was interested in how thinking about an
issue may lead people to change their minds about it in some respects. He told them that
half of the participants that come in will be asked to think about the issue for a moment
more then make another estimation of how many people they thought agreed with them,
and that half of the participants will not think about it any further but will make another
estimation straightaway. He then told all participants that they would be in the group
that would be thinking about the issue for a moment more before making another
3 The data from participants who were in the minority (they were in favor of raising the voting age) were
not used in the present data analyses, but they were allowed to complete the experiment. Their data were
reserved for other purposes and are not presented in this paper.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
estimation. In reality, this was a cover story, constant across all conditions, for the
second estimation of consensus. The reason for assessing a second estimation is that a
pilot study revealed a high range of responses in participants’ estimations, presumably
because they had little experience or exposure to the issue. This procedure of
ascertaining a second estimation of consensus was designed to reduce the range of
responses (and thus the variance of the dependent measures) by informing participants
of where they stood (majority or minority) with respect to the issue and giving them
another chance to estimate consensus. Also of interest in the current study is the effect
of the described manipulations of estimates of consensus by majority members, and so
participants had to be told whether they were in the numerical minority or majority with
respect to the central issue. As majority and minority status was not experimentally
manipulated in this study, participants were simply informed of their status according to
their self-reported responses to the opinion measures.
On this form, participants read (after it being explained to them as well) that they
were to think briefly about the issue in terms of (a) when they first came to hold their
opinion position, (b) whether they have ever held the opposite opinion, and (c) why they
hold the current opinion. They were instructed that, when they had finished thinking
about the issue, they were to make another estimation of how many people they thought
agreed with them, if necessary. They were also instructed not to spend a lot of time
thinking about it at this point.
Along with this form there was one more form. It was a form probing for
suspicion, questions regarding the experiment, and a check to assess whether
participants were cognizant of the videotaping process, or absence of it. The rationale
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
for giving this form before the end of the experiment, participants were told, was that the
experimenter wanted the participants to be able to leave right away after writing out
their defense (or watching the short video), and not have to worry about any more of
these forms to fill out. It was explained as a matter of convenience. All participants
agreed to this in anticipation of the last part of the experiment.
When participants had understood the instructions for these two final forms, the
experimenter explained, “Please hit this bell again when you are finished and I will
return so we can move on to the defense of your position (or, video presentation).” The
experimenter then left the room and awaited the belL
When the bell was heard, the experimenter returned to the experimental room
and asked if the participant was ready to move on to the defense (or, video). All
participants agreed. At this point, the experimenter informed participants that the
experiment was actually over, and that there would be no defense (or, video). He
debriefed participants and probed once more for suspicion regarding the purposes of the
experiment. When all participants had been instructed about the true nature and
purposes of the experiment and the deception involved, they were asked not to speak of
the experiment with other potential participants. They were then thanked and excused.
Results
Although males tended to give, on the whole, slightly higher estimates of
consensus, there were no significant differences between males and females in any of
the ensuing reported results, and so no sex differences are reported.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Manipulation Checks.
An index of anticipation was computed by aggregating the five separate bi-polar
adjective scales. The five bi-polar adjective scales were completed by participants
circling the number from 1 to 7 that corresponded most closely to how they felt (about
anticipating either the threatening situation or the video) according to the adjectives
given on each scale. The scales were (a) overjoyed to unpleasant, (b) very comfortable
to uncomfortable, (c) thrilled to annoyed, (d) loved to threatened, and (e) peaceful to
anxious (Cronbach’s alpha for these five items, V = .68). Each scale had a “neutral”
label at the mid-point (4) of the scale. The anticipation scale was not implemented, in
its final form, until roughly 40% of participants had completed the experiment. Thus,
the anticipation scores for the remaining 60% of participants were standardized. Those
respective z-scores (z-scores for all five items pooled within subjects to yield a mean z-
score for each participant) are used in the present analysis. Participants in the Threat
condition (anticipation and expectation of a negative event) scored significantly higher
on the aggregate anticipation scale, t(34) = 2.083, p < .04. These results suggest that the
manipulation of Threat was successful in producing its desired effect; namely, a state of
anxiety.
One index of self-focus was computed by averaging scores from the sentence
completion task. Within the fifteen sentences, there were seven that allowed for a
choice of self-referent words (Exner, 1973) versus non-self-referent words. For these
sentences, a first-person self-referent word (e.g., I, me, etc.) was given a score of five
points. A collective self-referent word (e.g., us, ours, etc.) was given a score of three
points, and a non-self-referent (e.g., his, her, theirs, etc.) possessive or pronoun word
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
was given a score of one point. The scores for the seven sentences were summed and
averaged for each participant. Self-Focused participants attained higher average scores
on this index than non-Self-Focused participants. The resulting statistic was, however,
only marginally significant, t(58) = 1.756, p < .085.
The other index of Self-Focus was the “free-thought” task, in which participants
wrote whatever they were thinking about, for two minutes. Those responses were coded
for their self-referent content as well Instead of giving each pronoun or possessive a
score (of 1, 3, or 5), it was decided that, since the number of sentences and types of
word could not be controlled for, it would be better to assess a raw number of self
referent words used, compared between participants. But also of interest were the non-
self-referent words. So, the index from the “free thought” task was computed by
subtracting the number of non-self-referent words from the number of self-referent
words. Self-focused participants attained a higher score than non-Self-Focused
participants on this task as w ell But again, the result was only marginally significant,
1(57) = 1.817, p < .075.
When the scores for both of these tasks (sentence completion and “free thought”)
were standardized, added together and averaged into a composite index o f Self-Focus,
the result was significant, t(57) = 2.576, p < .015.
Effects of Variables on Estimates of Consensus
The main dependent measures of this study were participants’ estimates of
consensus concerning the central issue. They were asked to give their response in the
form of a percentage. Given the nature of the response type, it was likely to produce a
wide range of responses. This increases the likelihood of obtaining outliers (extreme
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
values) in the data set (and thus Type I and Type II errors; see Tabachnick & Fidell,
1996; Wilcox, 1995), and so it was deemed appropriate to winsorize (a process of
replacement of extreme values; see Wilcox, 1995, pp. 15-17, for details) the dependent
measures’ values. The purpose of winsorizing data is to more accurately estimate
central tendency without the detrimental influence and obscuring effects of extreme data
points (Tabachnick & Fidell, 1996; Wilcox, 1990; Wilcox, 1995;).
The first estimate of consensus was given in the first packet of forms, just
following the “free thought” task and the sentence completion task. In this first
estimation of consensus, there were no main effects, but a marginally significant
interaction emerged, F(l, 59) = 3.11, p < .083. As can be seen in figure 1, the most
conservative estimates were given by those in the Threat / Self-Focus condition, and by
those in the control condition. This will be mentioned and considered later, in the
discussion section.
As mentioned before, pilot studies determined that participants’ responses to the
first estimate had too great a range (the issue was such that most students were
uninformed as to the distribution in the population), and so it was considered necessary
to try to restrict the range of responses to reduce the variance. This was done by
informing participants that their responses to (in favor of or against) the central issue put
them either in the majority or in the minority on that issue. They were also informed
that they could change their estimate if they wanted to.1 The results for the second
estimate of consensus can be seen in Figure 2. In the analyses of the second estimate of
1 A direction or magnitude of change was not suggested. The direction of change was obvious and even
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
consensus, the interaction emerged to be highly significant, F(l, 59) = 8.902, £ < .005,
and a main effect for Threat also resulted, F(l, 59) = 5.711, p < .02. When the analyses
were executed without winsorizing the data set, the interaction remained significant, F(l,
59) = 4.837, £ < .04, and the overall main effect for Threat was marginally significant,
F(l, 59) =4.325, fi< .07.
Post-Hoc analyses revealed significant differences between the non-Self-Focused
f Threat condition and the control condition, £ < .001, as well as between the non-Self-
Focused / Threat condition and the Self-Focused/ Threat condition, p i < .016. No other
between-groups differences were significant.
These results suggest that the induced threat was sufficient to reduce the level of
normal majority underestimation and bring levels of consensus estimation up (in the
second estimation) to almost the true population size of the majority (approximately
90%).
However, this main effect is importantly qualified by the more interesting and
notable interaction between the two variables. In the control condition, consensus
estimates reflected an average 13% underestimation. In the absence of Threat, Self-
Focus reduced underestimation by approximately 7% compared to the control condition.
And in the absence of Self-Focus, Threat reduced underestimation by almost 15%. But
in the condition in which Threat and Self-Focus are present, participants responded
conservatively and underestimated the true population of agreeing others even more than
their non-threatened, Self-Focused counterparts did (even after change scores were
implicitly required for those participants who incorrectly estimated their position to be majority or
minority.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
computed; see next paragraph for details). Possible explanations are given for this
finding in the discussion. The grand mean of majority underestimation (second
estimate, winsorized) was M = 79.7% (M = 74.01% for first estimation; M = 79.3%,
second estimation, non-winsorized), reflecting a 10.3% overall degree of bias from
reality (90% real group size) for all participants.
A measure of the direction and magnitude of participants’ change in estimates
from the first estimate to the second estimate was computed by subtracting each
participants’ first estimate from the second (see Figure 3). Just before making their
second estimation of consensus, participants were reminded that they would move on to
the last part of the experiment (either the written defense “to be reviewed and examined
by those who oppose you,” or the video) after completing the second estimation form.
An ANOVA was computed for the change scores and it was found that only a main
effect for Threat emerged, F(l, 59) = 4.419, p < .04. These results suggest that the
anxiety produced by the Threat manipulation led participants to underestimate less,
making a more extreme (higher) projection of social consensus.
Covariation Analyses
It was thought that both the importance of the issues and how much a person had
thought about an issue might be potential moderators of the dependent effects. This
prediction was partially confirmed.
As co-variates in the ANCOVA procedure, the participants’ perceptions of the
subjective importance of the issue (on a seven-point scale; M = 4.667, mode = 5),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
raising the voting age, had no effect on either the main effect for Threat or the obtained
interaction.
However, the reported amount o f thought (previous to the experiment) put into
the issue, as a co-variate in the ANCOVA procedure, eliminated the main effect for
Threat, while the obtained interaction was left virtually unaffected in terms of
significance, F (1,35) = 4.3, £ < .040.
Consensual Bias for Other Issues Assessed
Of interest to the general literature on the FCE are the responses to the non-
central issues asked on the questionnaire. The other issues were, “Do you think birth
control clinics should be legalized?’; “Do you watch more than 3 hours of TV per
day?”; “Do you think a woman will be elected president of the U.S. in the next 20
years?” ; “Are you in favor of socialized medicine?”; and “Do you sleep more than 8
hours per night?”
The actual sample group sizes, their estimates of group consensus, and the
computed biases from (sampled) reality are presented in Table 1. Estimates of opposite-
position percentages (Le., “What percentage of people do you think hold the opposite
position?’) were not assessed and are consequently not presented. Of interest in this
data set are two features.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
The first is that further confirmation has been obtained for the finding (Mullen &
Hu, 1988) that the degree of minority overestimation (M = 22%; see Table 1) exceeds
the degree of majority underestimation (M = -11%).2
Second is that further (albeit modest) support for the “Golden Section” (see
Gross & Miller, 1997) in perceptions of social consensus was obtained, although this
was not a purpose of the study. According to Mullen and Hu (1988), overestimation is a
minority process (group size < 50%) and underestimation is a majority process (group
size > 50%), but these data suggest that overestimation is not a minority process, but a
process of members of a group with a size of less than, not 50%, but 61.8%, or the
“Golden Section.” This can be seen by looking at the actual group size for those who
believe a woman will be elected president of the U.S. in the next twenty years (Table 1).
While the actual group size is 59% (a majority), they have actually overestimated their
group size. And by sheer serendipity, the sample of participants in this study produced a
group size of 61% (just below and adjacent to the “Golden Section”) for those in favor
of socialized medicine (Table 1). This majority underestimated their group size by only
1%! Clearly there is something peculiar and extraordinary (if not altogether spooky)
about the “Golden Section.”
2 The degree to which minority overestimation exceeds majority underestimation is computed comparing
absolute values, regardless of the mathematical sign. Thus, minority overestimation, in this example,
exceeds majority underestimation not because the forma: is positive and the latter negative, but rather it
is simply because the absolute value I22% l is greater than the absolute value lll% l.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Ancillary Data
On the probing form, given at the end of the experiment, many of the Self-Focus
participants reported feeling uncomfortable and / or anxious with the process of
videotaping the experimental session. Thus, there is further, albeit indirect, evidence in
favor of the Self-Focus manipulation producing its desired state (e.g., a form of anxiety,
see Carver and Scheier, 1981; Duval and Wicklund, 1972) in participants.
Discussion
Choice of issue to be used
In the pilot testing of this experimental paradigm, there were initially two issues
used as between-subjects dependent measures (the issues relevant to the manipulations).
Two issues were chosen to increase the external validity of the results. That is, if the
manipulations are successful in producing results for one issue, it was reasoned, then at
least another issue must be used to ensure that there is nothing particular to the one issue
used that produced the results. The two issues to be used, as mentioned in the
introduction, were, initially, “Are you in favor of raising the current voting age to 25?”
and “Do you consider yourself to be pro-choice regarding the abortion issue?” The
latter issue was dropped for several reasons.
The main reason was that responses to this issue were not susceptible to the
experimental manipulations. At first notice of this observation, one might argue that
there is in fact something particular and special about the other (raising the voting age)
issue that produced the desired results. But upon further examination of the results, it
can be seen that the issue-bound-particularity rests, rather, with the pro-choice issue
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
itself. In Table 1 it can be seen that the pro-choice issue responded neither to the
experimental manipulations, as did the voting issue (not presented in Table 1; see results
section for details), nor to typical, and in fact ubiquitous, social projection biases
exhibited by the other issues measured here (Table 1) as well as elsewhere (for an
analysis of 128 False Consensus effect sizes see Gross & Miller, 1997). In fact,
participants estimated their group size quite accurately, regardless of actual majority /
minority status or experimental manipulation. But, that this issue was perhaps a poor
choice doesn’t vitiate the results of the main issue, especially in light of the results
presented in Table 1.
In hindsight, it seems obvious that this issue should not have been used, for two
reasons. First, given the public knowledge and popular familiarity of the abortion issue,
in general, this issue should have been avoided from the outset A simple availability
explanation (Tversky & Kahneman, 1973) can account for the resistance of this issue to
manipulations. When a heuristic is available, and information primed, the heuristic
and/or information will be cognitively preferred to extra mental processing
(Bodenhausen, 1993; Bodenhausen & Lichtenstein, 1987; Gilbert & Hixon, 1991;
Kahneman & Tversky, 1979; Macrae, Milne, & Bodenhausen, 1994; Tversky &
Kahneman, 1973). It has also been found that predictive accuracy in consensus
estimates is limited by participants’ lack of knowledge and information regarding the
target population (Hoch, 1987).
Second, a review of the “Golden Section” (Gross & Miller, 1997) would have
precluded the use of this issue because all data, in contemporary assessments, show that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the abortion issue is approximately a 60% / 40% split in terms of the percentages of pro-
choice and pro-life advocates, respectively. And being a 60% majority, pro-choicers
would not, if this were like all other issue, underestimate the degree to which others
believe as they do. The “Golden Section” predicts that this group would actually
overestimate consensus, instead of underestimate. And moreover, since the majority
group size is near the “Golden Section” inflection point of 61.8%, very little bias a t all
would be expected. In light of that reflection, the experimental manipulations should
not have had the same effects on this issue as it did with the voting issue, anyway. So, it
turns out that the abortion issue would have been a bad choice to use with these
manipulations. Hence, it was dropped from the study as an issue relevant to the main
dependent measures.
Also, with considerations of the scant availability of participants, it was deemed
appropriate to continue the study with just one issue; namely, the voting issue. If it
turned out that the manipulations were successful and results were obtained, other issues
would be tested in the future to rule out possible issue-bound-particularity and issue
artifact. And since, in the current study, the manipulations were successful and results
were obtained, the next step is to test the paradigm with several other issues, as well as
exploring further implications of the independent variables themselves.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Discussion o f Findings
Overall, the predictions were confirmed. All experimental conditions produced
higher estimates of consensus than the control group. The Self-Focus-only condition
produced a difference from the control group that was in the predicted direction, but not
reliably so (p < .36).
Threat alone (without Self-Focus) produced a reliable difference from the control
group, in consensus estimates, in the predicted direction (p < .001). Evolution has
manufactured, presumably through the process and pressures of Natural Selection, the
cobra. This serpent’s key to survival, over the eons of phylogenetic development,
resided (apart from the gift of venom with which Nature has also endowed this creature)
in its selective advantage of appearing larger when threatened. Perhaps an analogy may
be made to the participants in this study. Is it not reasonable to suggest that the threat
imposed on these participants led to their higher estimates of consensus in order to bring
about the cognitive and, more importantly, socially advantageous consequences of
belonging to a larger group rather than a smaller one (Bettencourt, Charlton &
Kemahan, 1997; Tajfel, 1978)? Would not the defensive projection of a larger in-group
size help to ward off the “attackers” in this paradigm? Certainly this explanation is not
inconceivable.
Counter to the original predictions, however, were the results of the Threat /
Self-Focus condition. The mean estimate of that condition was not only not
significantly higher than the control group (p < .76), it was reliably lower (p < .016) than
the Threat-alone condition (see Figure 2). Where an additive effect was expected,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
producing the highest estimates of all conditions, relatively conservative estimates were
obtained.
These findings suggest that there may be something unique about combining the
two experimental manipulations within one condition. That combination of social
Threat and Self-Focus led to results which countered the predictions. What is it about
that combination of variables that led to such results in this study?
Perhaps the overload of aversive circumstances led participants to shut-down
cognitively for the estimation task. Duval and Wicklund (1972) found that high levels
of Self-Focus led participants to perform badly (substantially worse than low levels of
Self-Focus) on cognitive tasks. Duval, Duval and Mulilis (1992) found that high levels
of Self-Focus led participants to withdraw completely from a cognitive task.
Kahneman and Tversky (1979) suggested that under high risk or high stress,
people will make conservative judgments and decisions. Likewise, Lanzetta (1955)
found participants to be less aggressive, less confident and less concerned with
individual problems under high stress.
In the current study, when only the Threat conditions were considered (first set
of columns, Figure 1 & Figure 2), the two conditions did not differ overall with respect
to the aggregate anticipation / anxiety score average. However, the Threat / Self-Focus
participants obtained a reliably higher score for the individual items Threatened (p <
.032) and Unpleasant (p < .022) than the Threat / No Self-Focus participants. Thus,
participants in the Threat / Self-Focus condition reported being more threatened and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
having a more unpleasant experience. For findings concerning the main effect o f Threat
on this index, see results section.
Given that participants in the Threat / Self-Focus condition reported being more
threatened and unpleasant than all other participants, could this not account for their
conservative estimates of consensus? If stress produces participants that are less
aggressive and less confident (Lanzetta, 1955), would they also be less likely to make an
extreme judgment of consensus? If high levels of Self-Focus (Self-Focus p lu s social
threat with the focus of attention on the participant, in this experiment) lead to a
cognitive shut-down, avoidance or withdrawal, would it be inconsistent or unreasonable
to predict that consensus estimates would not differ from control participants? If high
stress leads participants to be less concerned with individual problems (Lanzetta, 1955),
does this constitute a move toward de-individuation (Brewer, 1991, 1993), thus
prompting participants to differentiate themselves and give lower estimates of
consensus?
With respect to the Self-Focus manipulation in this study, it seems that this
variable did not have the independent effect it was expected to have (see Gross & Miller,
1997), although it did interact with the other variable, which is an interesting and notable
finding in and of itself.
Perhaps a future examination of Self-Focus, within the context of perceptions of
social consensus, should include a manipulation of Objective Self-Awareness p er se
(Duval & Wicklund, 1972), rather than mere Self-Focus. Self-Focus alone does not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
explicitly involve the discrepancy between self and standard, or between actual behavior
and normative expectations. This added factor might yield interesting results as welL
A different, and practical, attempt to explain the results of the study might
employ a cost / reward analysis of consensus estimates. What are the costs and rewards
associated with publicly declaring one’s group to be a certain size? Taking up the
previous serpentine analogy, wouldn’t an individual be more likely to avoid costs, in a
potentially hostile situation, by “appearing” larger than reality? Isn’t it advantageous to
have as much social support as possible in such a situation, especially given an
evolutionary history of the “social animal” (Aristotle, c. 328 B.C.; Aronson, 1988)
called homo sapiens?
Discussion of “Change” scores
The change scores, presented in figure 3, represent the participants’ tendency
toward polarization of their opinion position. It may be argued that the opportunity
given each participant to think about the issue (see method section) before making
another consensus estimation, and the fact that they were given a relative majority
minority status, carries some explanatory burden in that polarization (Cotton & Baron,
1980; Gerard & Orive, 1987;Tesser, 1978, 1995). Tesser has repeatedly demonstrated
(1978, 1995) that simply thinking about one’s position will lead to attitude polarization.
Further, Gerard and Orive (1987; see also Cotton & Baron, 1980) argue that in the
presence of information about the prevailing opinion, attitudes of group members tend to
polarize. With that in mind, reconsider the results as presented in figure 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
If the participants’ thinking about their position were responsible for the upward
shift in consensus estimates, then the shift should be ubiquitous for all conditions, since
the thinking task was a constant for all participants. The same is true for the argument
that prevailing opinion information leads to attitude polarization. This, however, is not
the case. Although all conditions did, in fact, shift upward in their consensus estimates
given a second opportunity, repeated measures analyses reveal that only threatened
participants’ second estimations differ significantly from their first. The Threat / Self-
Focus condition differed with a mean change of 7.93, F (1, 14) = 9.683, p. < .008. The
Threat / No Self-Focus differed with a mean of 8.2, F (1,14) = 10.729, p < .006. The
No Threat / Self-Focus condition differed with a mean of 1.333, F (1,14) = .654, ns.
The No Threat / No Self-Focus condition differed with a mean of 3.13, F (1,14) = .9, ns.
It may be assumed from these analyses that the thinking task and /or the
prevailing opinion information had a weak polarizing effect, if any, and that the Threat
manipulation better accounts for the significant upward shift in the Threat conditions.
Further Research
It is assumed that the threat of a potentially hostile intergroup situation triggered
the decrease in majority underestimation in the current study. In order to find out the
nature of the responses to that threatening situation, more should be explored in this
area. Specifically, further studies should manipulate Threat again, but along with
another manipulation, that of affiliation. That is, if the same threatening situation is
imposed on participants, and some are given an opportunity to have some form of an
affiliative experience (k la Schacter, 1959) before facing the (assumed) intergroup
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
situation while some are not, what would happen to the estimates of social consensus?
If the decrease in underestimation (higher consensus estimates) is due to the
participants’ desire for social support, then would actual, physical social support assuage
the motivation for support in the form of higher consensus estimates? In Brewer’s
(1991) terms, would the affiliative experience satisfy the need for assimilation in the
face of a threatening (differentiating) situation? Would actual, physical social support
give participants a feeling of control over the stress, as has been suggested by Cobb
(1976)? This, of course, remains to be seen.
A fictional journey that I have mused about recently, regarding projection of
social consensus, is a make-believe trip into the folk-tale world of the 19th century
Danish author, Hans Christian Andersen. I would wander among the participants of the
notoriously (yet obliviously) exhibitionary Emperor, recently given brand-new “clothes”
by a pair of bamboozling tailors, and privately gather consensus estimates both from
those who saw the “clothes” and from those who dared to admit that they saw an
emperor au naturel. In such a judgmental situation, it would be intriguing and amusing
to see the real-world effects of extreme incertitude about one’s opinion on estimations of
social consensus. Within the current paradigm, this latter illustration might constitute
one of the future variables experimentally examined, yet presumably with less immodest
a target of judgmental accuracy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
References
Aristotle, (c. 328 B.C.). Politics.
Aronson, E. (1988). The Social Animal (5th ed.). New Y ork: W. H. Freeman and
Company.
Baron, R. S., Kerr, N. L., & Miller, N. (1992). Group Process. Group Decision.
Group Action. Pacific Grove, CA: Brooks / Cole Publishing Company.
Bettencourt, B. A., Charlton, K., & Kemahan, C. (1997). Numerical
representation of groups in cooperative settings: Social orientation effects in ingroup
bias. Journal of Experimental Social Psychology. 33. 630-659.
Bodenhausen, G. V. (1993) Emotions, arousal, and stereotypic judgments: A
heuristic model of affect and stereotyping. In D. M. Mackie and D. L. Hamilton (Eds.),
Affect. Cognition, and Stereotyping: Interactive Processes in Group Perception. San
Diego: Academic Press.
Bodenhausen, G. V., & Lichtenstein, M. (1987). Social stereotypes and
information-processing strategies: The impact of task complexity. Journal of
Personality and Social Psychology. 52. 871-880.
Brewer, M. B. (1979). Ingroup bias in the minimal intergroup situation: A
cognitive-motivational analysis. Psychological Bulletin. 86. 307-324.
Brewer, M. B. (1991). The social self: On being the same and different at the
same time. Personality and Social Psychology Bulletin. 17. 475-482.
Brewer, M. B. (1993). Social identity, distinctiveness, and in-group
homogeneity. Social Cognition. 11. 150-164.
Brewer, M. B., & Miller, N. (1984). Beyond the contact hypothesis: Theoretical
perspectives on desegregation. In N. Miller & M. B. Brewer (Eds.), Groups in Contact:
The Psychology of Desegregation. Orlando: Academic Press.
Carver, C., & Scheier, M. (1981). Attention and self-regulation. New York:
Springer-Verlag.
Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic
Medicine. 38. 300-314.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Cotton, J. L., & Baron, R. S. (1980). Anonymity, persuasive arguments, and
choice shifts. Social Psychology Quarterly. 43. 391,404.
Crano, W. D. (1983). Assumed consensus of attitudes: The effect of vested
interest. Personality and Social Psychology Bulletin. 9. 597-608.
Dawes, R. M. (1989). Statistical criteria for establishing a truly false consensus
effect. Journal of Experimental Social Psychology. 25. 1-17.
Duval, S., & Duval, V. H. (1983). Consistency and Cognition: A Theory of
Causal Attribution. New York: Erlbaum.
Duval, S., Duval, V., & Mulilis, J. (1992). Effects of Self-Focus, discrepancy
between self and standard, and outcome expectancy favorability on the tendency to
match self to standard or to withdraw. Journal of Personality and Social Psychology. 62.
340-348.
Duval, S., & Wicklund, R. (1972). A theory of objective self-awareness. New
York: Academic Press.
Exner, J. E. (1973). The self-focus sentence completion: A study of ego-
centricity. Journal of Personality Assessment. 37.437-455.
Geller, V., & Shaver, P. (1976). Cognitive consequences o f self-awareness.
Journal of Experimental Social Psychology. 12.99-108.
Gerard, H. B., & Orive, R. (1987). The dynamics of opinion formation. In L.
Berkowitz (Ed.), Advances in Experimental Social Psychology (VoL 20). New York:
Academic Press.
Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activation and
application of stereotypes. Journal of Personality and Social Psychology. 60. 509-517.
Gross, S. R., & Miller, N. (1997). The ‘Golden Section” and bias in
perceptions of social consensus. Personality and Social Psychology Review. 1. 241-271.
Greenberg, J., & Pyszczynski, T. (1986). Persistent high self-focus after failure
and low self-focus after success: The depressive self-focusing style. Journal of
Personality and Social Psychology. 50. 1039-1044.
Hoch, S. J. (1987). Perceived consensus and predictive accuracy. Journal of
Personality and Social Psychology. 53. 221-234.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Janis, I. L. (1968). Group identification under conditions of external danger. In
D. Cartwright & A. Zander (Eds.), Group Dynamics (pp. 80-90). New York: Harper &
Row.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision
under risk. Econometrica. 47. 263-291.
Krueger, J., & Clement, R. W. (1997). Estimates of social consensus by
majorities and minorities: The case for social projection. Personality and Social
Psychology Review. 1. 299-313.
Lanzetta, J. T. (1955). Group behavior under stress. Human Relations. 8. 29-53.
Lanzetta, J. T., Haefner, D., Langham, P., & Axebrod, H. (1954). Some effects
of situational threat on intergroup behavior. Journal of Abnormal and Social
Psychology. 4 9 .445-453.
Macrae, C. N., Milne, A. B., & Bodenhausen, G. V. (1994). Stereotypes as
energy-saving devices: A peek inside the cognitive toolbox. Journal o f Personality and
Social Psychology. 66. 37-47.
Marks, G., & Miller, N. (1987). Ten years of research on the false consensus
effect: An empirical and theoretical view. Psychological Bulletin. 102. 72-90.
Mullen, B. (1987). Self-attention theory: The effects of group composition on
the individual: A self-attention perspective. In B. Mullen, & G. R. Goethals (Eds.),
Theories of Group Behavior. New York: Springer-Verlag.
Mullen, B., & Baumeister, R. F. (1987). Group effects on self-attention and
performance: Social loafing, social facilitation and social impairment. In C. Hendrick
(Ed.) Group Processes and Intergroup Relations (pp. 197-198). Newbury Park, CA:
Sage.
Mullen, B., & Hu, L-L (1988). Social projection as a function of cognitive
mechanisms: Two meta-analytic integrations. British Journal of Social Psychology. 27.
333-356.
Ng, S. H., & Cram, F. (1988). Intergroup bias by defensive and offensive
groups in majority and minority conditions. Journal of Personality and Social
Psychology. 55. 749-757.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Noel, J. G., Wann, D. L., & Branscombe, N. R. (1995). Peripheral ingroup
membership status and public negativity toward outgroups. Journal o f Personality and
Social Psychology. 68. 127-137.
Pratto, F., & John, O. P. (1991). Automatic vigilance: The attention-grabbing
power of negative social information. Journal o f Personality and Social Psychology. 61.
380-391.
Pyszczynski, T., & Greenberg, J. (1987). Self-regulatory perseveration and the
depressive self-focusing style: A self-awareness theory of reactive depression.
Psychological Bulletin. 102. 122-138.
Ross, L., Greene, D., & House, P. (1977). The “false consensus effect”: An
egocentric bias in social perception and attribution processes. Journal of Experimental
Social Psychology. 13. 279-301.
Schacter, S. (1959). The Psychology of Affiliation. Stanford, CA: Stanford
University Press.
Sherman, S. J., Presson, C. C., Chassin, L., Corty, E., & Olshavsky, R. (1983).
The false consensus effect in estimates of smoking prevalence: Underlying mechanisms.
Personality and Social Psychology Bulletin. 9, 197-207.
Strumpfer, D. J. (1970). Fear and affiliation during a disaster. Journal of Social
Psychology. 82. 263-268.
Suls, J., & Wan, C. K. (1987). In search of the false-uniqueness phenomenon:
Fear and estimates of social consensus. Journal of Personality and Social Psychology.
52*211-217.
Tabachnick, B. G., & Fidell, L. S. (1996). Using Multivariate Statistics (3rd ed.).
New York: Harper & Row.
Tajfel, H. (1978). The social psychology of minorities. London: Minority Rights
Group.
Tesser, A. (1978). Self-generated attitude change. In L. Berkowitz (Ed.),
Advances in Experimental Social Psychology (VoL 11). New York: Academic Press.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Tesser, A. (1995). The impact of thought on attitude extremity and attitude-
behavior consistency. In J. A. Krosnick & R. E. Petty (Eds.), Attitude Strength:
Antecedents and Consequences (VoL 4). Mahwah, NJ: Lawrence Erlbaum Associates,
Inc.
Thompson, S. C., Kohles, J. C., Otsuki, T. A., & Kent, D. R. (1997). Perceptions
of attitudinal similarity in ethnic groups in the U.S.: ingroup and outgroup homogeneity
effects. European Journal of Social Psychology. 27. 209-220.
Travers, R. M. W. (1941). A study in judging the opinions of groups. Archives of
Psychology. (Whole no. 266).
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic forjudging
frequency and probability. Cognitive Psychology. 5. 207-232.
Wegner, D. M., & Giuliano, T. (1980). Arousal-induced attention to self.
Journal of Personality and Social Psychology. 38. 719-726.
Wilcox, R. (1990). Comparing the means of two independent groups.
Biometrical Journal. 32. 771-780.
Wilcox, R. (1995). Statistics for the Social Sciences. Orlando: Academic Press.
Wolf, J. (1998). CIA Inquest finds U.S. missed Indian‘mindset’. Yahoo!
News: Top Stories Headlines. Retrieved June 3,1998 from the World Wide Web:
http://dailynews.yahoo.com/headlines/top_...=z/reuters/980603/news/stories/cia_l.html
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix
(Form#l)
Please write down, on this page, the thoughts that come to your mind in the next two minutes. Do not
concern yourself with writing style or being logicaL
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
(Form #2)
♦♦READ INSTRUCTIONS CAREFULLY**
To collect some information on basic mental processes, we ask that you read the following fifteen sentences and
write the word (from the three choices provided in parentheses) that most appropriately completes the sentence in the
blank provided. All of the answers are technically correct, but a close analysis of each sentence might reveal that one
alternative is more likely to occur in that context than others.
1. Under certain circumstances, (she, we, I ) will almost always choose to lie rather than tell the truth,
choice:_________________________
2. The dogs waited (quietly, nervously, fretfully) for Sam to come home,
choice:____________________________
3. After spreading fertilizer liberally over the flower bed, (I, she, we) watered the flowers,
choice:____________________________
4. Walking down the street near his home, Stuart often found time to (hum, think, daydream),
choice:____________________________
5. Many critics accused the prime minister of (taunting, mocking, fighting) the royal family,
choice:____________________________
6. Please don't do this to (me, her, us), it is just not fair, choice:______________________
7. After the first rainfall of spring, the (animals, tourists, citizens) arrive in great numbers,
choice:____________________________
8. After closing the windows and curtains, (he. I, we) sat down to relax after the party,
choice:____________________________
9. Children often think it (silly, amusing, fun) to eat ice cream, choice:__________________
10. Although (their, our, my) personal library consists of only a few books, a good number o f them are classics,
choice:________________________
11. Once the airplane had landed, the passengers were (relieved, happy, anxious) to have arrived at their destination,
choice:_____________________
12. At first it didn't make any difference, but by later that night the noise from the party was entirely too loud to allow
(us, her, me) to sleep, choice:____________________________
13. The weather in Minnesota is very much (unpredictable, annoying, pleasant),
choice:_______________________________
14. During summer, there's hardly anything for (him, me, us) to do.
choice: ________________________________
15. There are twice as many good (meals, waiters, seats) as bad ones at this restaurant,
choice:________________________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
(form #3)
Please answ er the follow ing questions by checking the appropriate blank follow ing each question. W e are particularly
interested in your opinion about the item in bold w riting (item #5). Then, for each question, indicate, using a num ber
betw een 0 and 100, the percentage of others w ho you thinlr agree w ith you.
EXAMPLE: Do you like peanut butter with tuna fish sandwiches? yes________ no X_____
What percentage of other people do you think would answer this question the same way you
did?________ 95%___________
1) Do you think that birth control clinics should be legal? yes_______ n o __________
What percentage of other people do you think would answer this question the same way
you did?_________________________
2) Are you in favor of socialized medicine? yes____________ no_____________
What percentage of other people do you think would answer this question the same way
you did?_________________________
3) Do you watch more than 3 hours of television per day? yes_______ n o _________
What percentage of other people do you think would answer this question the same way
you did?_________________________
4) Do you sleep at least 8 hours per night? yes_____________n o _____________
What percentage of other people do you think would answer this question the same way
you did?_________________________
**5) Are you in favor of raising the current voting age to 25 years old?
yes____________ n o ____________
W hat percentage of other people do you think would answer this question the same
way you did?________________________
6) Do you think there will be woman elected as the US president in the next 20 years?
yes____________ n o ____________
What percentage of other people do you think would answer this question the same way
you did?_________________________
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
(form #4)
Please indicate, using the provided scale, how important each issue is to you, personally.
Not at all Indifferent Very Important
1-----------2--------- 3---------4 ---------- 5----------6--------- 7
_______ 1. Raising the voting age to 25.
_______ 2. Whether you are frequently depressed or not.
_______ 3. Whether you are Pro-choice versus Pro-life on the abortion issue.
_______ 4. Whether a woman will be elected to be US president in 20 years.
_______ 5. Whether you are for or against the death penalty.
_______ 6. Whether you are for or against homosexual marriages.
_______ 7. Whether or not we will discover life on other planets.
_______ 8. Whether or not women with young children should work outside the home.
Please indicate, using the provided scale, how much you have thought about each issue.
Not at all A Little Very Much
_______ 1. Raising the voting age to 25.
_______ 2. Whether you are frequently depressed or not.
_______ 3. Whether you are Pro-choice versus Pro-life on the abortion issue.
_______ 4. Whether a woman will be elected to be US president in 20 years.
_______ 5. Whether you are for or against the death penalty.
_______ 6. Whether you are for or against homosexual marriages.
_______ 7. Whether or not we will discover life on other planets.
_______ 8. Whether or not women with young children should work outside the home.
Please indicate how you feel about being evaluated on your opinions, by those who disagree with you,
below (circle the number):
I feel:
Overjoyed Neutral Unpleasant
1---------- .2----------3---------4 -------- 5-----------6---------- -7
Very Comfortable Neutral Uncomfortable
1-----------2--------- 3---------4 ---------- 5-------- -6----------- 7
Thrilled Neutral Annoyed
1----------- 2----------3---------4 -------- 5-----------6---------- -7
Loved Neutral Threatened
Peaceful Neutral Anxious
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Author’s Note
Jared B. Kenworthy, Department of Psychology, University of Southern California.
This study was conducted as a Master’s Thesis. I would like to gratefully acknowledge
Dr. Norman Miller for his shrewd insights and comments during the entire project and T.
Shelley Duval for his advice and direction concerning the Self-Focus manipulations during the
preparation stages of this study. Finally, I also warmly thank Dr. Jerald Jellison for always
bringing insightful and practical perspectives to my research and thinking in general.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M ean F irst Estimate o f Consensus
44
9 0-
88 >
86 ■
8 4-
8 2-
8 0.
78
76
74
72
70
68
Threat no Threat
Threat induction level
Self-Focus level
B |S elf-F ocused
| |not Self-Focused
Figure 1. Effects of Threat and Self-Focus on First Estimate of Social Consensus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M ean Second Estim ation o f Consensus
45
Self-Focus level
^Hself-Focused
I Inot Self-Focused
Threat no Threat
Threat induction level
Figure 2. Interaction between Threat and Self-Focus on Second Estimate of Consensus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
d)
C B
E
to
L U
a >
co
c
c o
sz
O
c
co
c u
Self-Focus level
IIISelf-Focused
| |not Self-Focused
Threat no Threat
Threat induction level
Figure 3. Mean Change from First Estimation to Second Estimation. Numbers reflect
means of raw differences between first and second percentages given.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
Bias from actual group size
Issue Choice Actual
size%
Own group
estimate
Majority
bias
Minority
bias
In favor of legalizing Yes 92% 68% -24%
birth control clinics? No 8% 24% 16%
Sleep > 8 hours per night? No 85% 59% -16%
Yes 15% 69% 44%
Watch > 3 hours of TV No 71% 47% -24%
per day? Yes 29% 74% 45%
=Are you pro-choice? Yes 58% 58% 0%
No 42% 45% 3%
Woman elected president? Yes 59% 69% *+ 10%
No 41% 51% 10%
In favor of socialized Yes 61% 60% -1%
medicine? No 39% 53% 14%
M = -11 % 22%
=This issue was assessed as a “ central” issue with regard to the manipulations. See discussion.
*For a discussion of majority overestimation and the “Golden Section” see Gross & Miller (1997).
Table 1. Actual group sizes, consensus estimate averages, and degree of consensual
bias.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IMAGE EVALUATION
TEST TARGET (Q A -3 )
A
✓
& A s ?
/,
A
1 . 0
l.l
1.25
( £
| 2.8
m
I S
m
L - .
u i
ys
i l l 2 - 0
1 . 4
1.8
1 . 6
150m m
I I W 1 G E . I n c
1653 East Main Street
Rochester, NY 14609 USA
Phone: 716/482-0300
Fax: 716/288-5989
0 1993. Applied Im age. Inc.. All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Ideological threat, social identity, and consensus estimation

PDF

Asymmetries in the bidirectional associative strengths between events in cue competition for causes and effects

PDF

Effects of personal resource sufficiency on perceived difficulty and desirability of earthquake preparedness

PDF

Intragroup evaluations, attitude source, and in-group member derogation

PDF

Depression and suicidality in Latino adolescents: A study of acculturation and gender role beliefs

PDF

Explanations, the availability heuristic, and biases of subjective likelihood in economics

PDF

Hedonic aspects of conditioned taste aversion in rats and humans

PDF

Affection and conflict in family relationships

PDF

Adolescents' social attitudes: Genes and culture?

PDF

Effect of affective reactions by an outgroup on preferences for crossed categorization discussion partners

PDF

Five quantile test: A comparison of independent groups

PDF

Head injury and dementia: A co-twin control study of Swedish twins

PDF

Effects of bilateral stimulation and stimulus redundancy on performance in processing nonword letter trigrams

PDF

Expectancies for alternative behaviors predict drinking problems: A test of a cognitive reformulation of the matching law

PDF

Articulated thoughts about intentions to commit anti-gay hate crimes

PDF

Ethnic identity, acculturation, self-esteem and perceived discrimination: A comparison study of Asian American adolescents

PDF

Adaptation to sine-wave gratings selectively reduces the sensory gain of the adapted stimuli

PDF

Cognitive functioning and dementia following cancer: A Swedish twin study

PDF

Alcohol expectancies and consumption: Age and sex differences

PDF

Insights into the nature of phonological and surface dyslexia: Evidence from a novel word learning task

Asset Metadata

Creator Kenworthy, Jared Bowden (author)

Core Title Effects of threat and self-focus on consensual bias in majorities

School Graduate School

Degree Master of Arts

Degree Program Psychology

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

Tag OAI-PMH Harvest,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-29324

Unique identifier UC11337565

Identifier 1394792.pdf (filename),usctheses-c16-29324 (legacy record id)

Legacy Identifier 1394792.pdf

Dmrecord 29324

Document Type Thesis

Rights Kenworthy, Jared Bowden

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