Social Psychology

Sweetness668

SociallymotivatedprojectionNeedtobelongincreasesperceivedopinionconsensusonimportantissues.pdf

Home >Psychology homework help >Social Psychology

Research article

Socially motivated projection: Need to belong increases perceived opinion consensus on important issues

KIMBERLY RIOS MORRISON1* AND JÖRG MATTHES2

1Department of Psychology, University of Chicago, USA; 2Department of Communication, University of Zurich, Switzerland

Abstract

In three studies, we tested whether the need to belong would motivate people to perceive consensus for theiropinions on important

social issues. In Study 1, a nationally representative telephone survey, participants with a high dispositional need to belong

perceived greater consensus for their opinions on immigrant naturalization than did those with a low need to belong. However,

this relationship was strongest among participants who reported that the issue was personally important to them. In Study

2, participants primed with rejection-related (versus acceptance-related) words, and who reported high levels of issue

importance, demonstrated greater false consensus for their opinions on a proposed alcohol tax increase. In Study 3, participants

who received random feedback that they held a common (versus uncommon) opinion had a lower subsequent need to belong

when the issue was important to them, suggesting that consensus perceptions can in fact mitigate belongingness needs.

Michael Scott, protagonist of the popular American television

series The Office, 1 desperately wants to be accepted by others.

As the regional manager of a paper company, he considers

himself a ‘‘friend first and boss second’’ and is determined

to have close personal relationships with his employees.

However, one consequence of this is that he tends to believe

his employees have the same priorities, sense of humor, and

opinions about workplace issues (e.g., diversity, sexual

harassment) as he does. When these delusions are shattered,

he inevitably faces shock and disappointment.

Michael Scott is a prototypical example of someone with a

high need to belong, defined as the motivation to have both

frequent social contact and stable relationships with others

(Baumeister & Leary, 1995). Research has shown that people

who are high in need to belong, or whose belongingness needs

have been obstructed (e.g., who have been socially excluded),

monitor their social environments more closely than do those

low in need to belong. For example, having a high need to belong

causes individuals to selectively recall and attend to social

information (Gardner, Pickett, & Brewer, 2000; Gardner, Pickett,

Jefferis, & Knowles, 2005), accurately decode nonverbal cues

(Pickett, Gardner, & Knowles, 2004), and mimic their interaction

partners (Lakin, Chartrand, & Arkin, 2008). Yet it is unclear

whether and how the need to belong might affect people’s

perceptions of others’ opinions. In the present paper, we test the

prediction that like Michael Scott, individuals high in need to

belong will exhibit inflated estimates of consensus for their

opinions, and that they will be most motivated to do so on

issues that are personally important to them.

NEED TO BELONG AND PERCEPTIONS OF

CONSENSUS

The question of what prompts individuals to believe that

others’ opinions are congruent with their own has piqued the

interest of many psychologists (Ross, Greene, & House, 1977;

for reviews, see Krueger, 1998; Marks & Miller, 1987). One

possible cause that has been hinted at but not fully investigated

in the literature involves people’s motives for social support

(i.e., their belongingness needs). For example, adolescent

smokers tend to overestimate the percentage of their peers who

smoke, perhaps in an attempt to convince themselves that such

deviant behaviors are common (Sherman, Presson, Chassin,

Corty, & Olshavsky, 1983). Similarly, opinion minorities—

who may be unsure of whether they fit in with others—have

been found to estimate higher levels of consensus for their

attitudes than do opinion majorities (Mullen & Hu, 1988;

Robbins & Krueger, 2005). Also consistent with the idea that

perceptions of consensus might be amplified among those who

have a high need to belong, people project their attitudes more

onto those with whom they anticipate interacting socially

(Miller & Marks, 1982). However, given that belongingness

*Correspondence to: Kimberly Rios Morrison, Department of Psychology, University of Chicago, 5848 S. University Avenue, Chicago, IL 60637, USA. E-mail: [email protected]

1 The Office is based on an English television series of the same name, which debuted in 2001 and starred Ricky Gervais as David Brent (the Michael Scott-like character).

European Journal of Social Psychology, Eur. J. Soc. Psychol.

Published online 14 March in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/ejsp.797 4 , 707 719 (2011)–1

needs have only been inferred and not assessed in these studies,

the evidence on this point is inconclusive.

Additional evidence that the need to belong might increase

consensus perceptions comes from research on attachment

styles, whereby anxious-ambivalent individuals (i.e., who

worry that their romantic partners will not reciprocate their

affection) are more prone to overestimate the popularity of

their opinions than avoidant (i.e., who find it difficult to trust

their romantic partners) or securely attached (i.e., who have

stable relationships with their romantic partners) individuals

(Mikulincer, Orbach, & Iavnieli, 1998). If people with an

anxious-ambivalent attachment style, relative to their avoidant

and securely attached counterparts, are more concerned about

being socially excluded, this would again suggest that the need

to belong increases perceptions of consensus. But having an

anxious-ambivalent attachment style is not the same as having

a high need to belong (see Carvallo & Gabriel, 2006). For

instance, it is conceivable that someone would be preoccupied

with maintaining a particular romantic relationship, but not

with being accepted by other people in general (Baumeister &

Leary, 1995). It, therefore, remains an open question whether

the need to belong would cause people to misperceive others’

opinions as similar to their own.

Although there is a lack of direct evidence for a link

between belongingness needs and consensus estimates,

recent research indicates that people who have a chronic or

experimentally induced need to belong strive to affiliate with

their peers in other ways. For instance, they are more interested

in forming new interpersonal connections (Maner, DeWall,

& Baumeister, 2007), more likely to conform to others’

judgments (Williams, Cheung, & Choi, 2000), and less

inclined to believe that they have encountered personal

discrimination (Carvallo & Pelham, 2006). Furthermore,

people with a high need to belong tend to perceive their

ingroup as more homogeneous (i.e., with its members

possessing similar characteristics; Pickett & Brewer, 2001)

and entitative (i.e.; with its members being bonded together

into a coherent entity; Knowles & Gardner, 2008). Group

homogeneity and entitativity can be distinguished concep-

tually from opinion consensus, however. For instance, a

group or organization may consist of individuals with similar

demographic backgrounds (e.g., young college graduates) and

be perceived as a tight-knit entity, yet at the same time foster

individualism and welcome the expression of unique opinions

(see Hornsey & Jetten, 2004).

THE EFFECT OF PERSONAL IMPORTANCE

Though several studies have provided preliminary support for

a link between need to belong and perceived consensus,

none have directly demonstrated this relationship. Equally

important, research has yet to determine the conditions under

which belongingness needs might amplify consensus esti-

mates. In particular, if the need to belong motivates individuals

to perceive high levels of consensus for their opinions, then the

relationship between need to belong and consensus estimates

might be most pronounced among those who view the issue as

important to their self-concept (Thomsen, Borgida, & Lavine,

1995). Even on the most controversial of issues, people vary in

how important a given issue is to them personally (e.g.,

Boninger, Krosnick, & Berent, 1995; Pomerantz, Chaiken, &

Tordesillas, 1995). Individuals whose sense of self is closely

tied to the issue may be especially likely to care about whether

their opinions are held by others. In contrast, individuals

who attribute less personal importance to the issue may be

relatively indifferent to information about the commonness of

their opinions, as such opinions bear little consequence for

‘‘who they are’’ (see Katz, 1960).

Personal importance has indeed been shown to influence

consensus estimates on social issues, although evidence for

the extent and direction of this effect is mixed. In one

experiment, college students who were led to believe that a

tuition surcharge plan would affect them perceived greater

consensus for their opinions on the plan than did those who

were not led to believe they would be affected (Crano, 1983).

Presumably, because the issue was more important to the

former group of students, they were more motivated to seek

social support by thinking that their peers agreed with them.

However, additional studies have shown no relation between

personal importance and consensus estimates (Fabrigar

& Krosnick, 1995). Still other work has suggested that

importance makes individuals more accurate in their assess-

ment of others’ opinions (Campbell, 1986), perhaps due to

the tendency for importance to foster more effortful processing

of issue-relevant information (Petty & Cacioppo, 1984).

One potential reason for these discrepant results is that

need to belong was not measured or manipulated in any of

the studies. It may be that personal importance increases

consensus estimates especially among people who are

concerned about affiliating with others (i.e., who have a high

need to belong). By contrast, people who have a relatively low

need to belong may be less motivated to perceive agreement

among their peers, even on important issues. Looking at

the role of need to belong may, therefore, shed light on

when personal importance does and does not increase

consensus perceptions.

OVERVIEW OF STUDIES

The present studies used both survey and experimental

methodology to test the relationship between need to belong,

personal importance, and consensus perceptions. In Study 1,

a telephone survey conducted with a nationally representative

sample in Switzerland, participants completed a need to

belong measure and answered several questions about their

perceptions of public opinion on the naturalization of

immigrants. They also indicated how important the issue

was to them personally. In Study 2, a Web-based experiment

conducted in the United States, participants completed a

manipulation of either high or low need to belong, prior to

answering questions about their own and others’ opinions on

a proposed alcohol tax increase. Across both studies, we

hypothesized that people with a high (relative to low) need to

belong would perceive greater consensus for their opinions,

particularly when the issue was important to them.

Study 3 sought to provide evidence that perceiving high

levels of consensus for one’s opinion can in fact reduce the

need to belong. In this study, American participants completed

¨rg Matthes708 Kimberly Rios Morrison and Jo

, 707 719 (2011)–1

a Web-based experiment in which they received false feedback

that they held either a common or uncommon opinion on the

alcohol tax. They then completed a state measure of need to

belong. It was predicted that participants who were told that

their opinions were common would have a lower subsequent

need to belong than those who were told that their opinions

were uncommon, and that this relationship would be most

pronounced among those for whom the alcohol tax was

personally important.

STUDY 1

Method

The data from this study were collected as part of a survey

about a referendum for the naturalization of immigrants in

Switzerland. The referendum, which ultimately did not pass

but was under debate at the time of the study, would have

entailed stricter application and decision procedures for

immigrants who wanted to attain Swiss citizenship. Partici-

pants (n ¼ 1251) were recruited in June 2008 by a random digit dialing procedure. All were residents of Switzerland who

spoke either German or French. The questionnaire was

programmed for an application of Computer Assisted

Telephone Interview (CATI). To ensure high data quality,

the study was conducted by an internationally established

polling company. The measures were administered in the

order reported below, and they were interspersed with several

other items about the referendum that did not relate to the

present research. (See Table 1 for the correlations between all

measured variables in this study.)

Demographics

The following demographic variables were included as

controls in this study: Age (M ¼ 48.29, SD ¼ 16.71), sex (51% female), and language (78% German, 22% French).

Personal Importance

Participants responded to the items, ‘‘The issue of natural-

ization is very important to me personally’’ and ‘‘The issue of

naturalization is important to Switzerland’’ (1 ¼ do not agree at all, 5 ¼ fully agree). These items were averaged together to form a personal importance composite (M ¼ 4.20, SD ¼ 0.91; a ¼ .65).

Perceived Similarity

Participants indicated their agreement with the following three

items: ‘‘My opinion on the issue of naturalization equals most

opinions I hear in my surroundings’’; ‘‘In the place where I

live, most people think the same way about the issue of

naturalization as I do’’; and ‘‘The Swiss nation has the same

opinion about the issue of naturalization as I do’’ (1 ¼ do not agree at all, 5 ¼ fully agree). These items were averaged together to create an index of perceived similarity of opinion

(M ¼ 3.02, SD ¼ 0.82; a ¼ .59).2 Measures such as this one, which ask people for their subjective estimates of whether others

have the same opinions or characteristics as they do, have been

used in previous research on the antecedents and consequences

of consensus effects (Marks & Miller, 1987; Matthes,

Morrison, & Schemer, 2010; Miyake & Zuckerman, 1993).

Opinion Congruence

Participants responded to the statement, ‘‘I approve stricter

naturalization rules in Switzerland’’ (1 ¼ strongly disagree, 10 ¼ strongly agree; M ¼ 5.75, SD ¼ 3.01). They also indi- cated how they thought the majority of Swiss citizens felt

about the referendum (1 ¼ opposed, 2 ¼ neutral, 3 ¼ in favor; M ¼ 2.05, SD ¼ 0.84). Participants who opposed the refer- endum (i.e., who scored between 1 and 5) and thought the

Swiss majority opposed it were coded as perceiving opinion

congruence, as were those who favored the referendum (i.e.,

who scored between 6 and 10) and who thought the Swiss

majority favored it (total n ¼ 420). Participants who opposed the referendum but thought the Swiss majority favored it, and

vice versa, were coded as not perceiving opinion congruence

(total n ¼ 345) (for a similar measure, see Miyake & Zuckerman, 1993). The 486 participants who thought the

Swiss majority was neutral toward the referendum were not

coded, as the item assessing their own attitudes did not have a

scale midpoint. The opinion congruence measure served as an

additional, more implicit measure of consensus perceptions.

That is, participants were asked about their own and others’

opinions separately, rather than directly reporting how similar

they believed their opinions were to the majority.

Need to Belong

After answering 50–60 filler questions on unrelated topics,

participants responded to five items assessing their need to

belong, developed by Hayes and colleagues (Hayes, Matthes,

& Eveland, 2010; see Appendix for the list of items).

Participants indicated their agreement with each item with a

number from 1 (do not agree at all) to 5 (fully agree). Their

responses were averaged into a composite, with higher scores

reflecting a higher need to belong (M ¼ 2.95, SD ¼ 0.89; a ¼ .69). This measure has been validated in multiple samples from two different countries (the U.S. and Switzerland), and it

correlates only weakly with dispositional shyness. 3

Participants’ need to belong scores were positively

correlated with their support for the referendum (r ¼.13, p < .001). However, because the filler questions before the need to belong scale took approximately 20 minutes to

2 The a coefficient for this perceived similarity composite was quite low, most likely because each of the items referred to a different group (i.e., ‘‘my surroundings,’’ ‘‘the place where I live,’’ and ‘‘the Swiss nation’’). Study 2 addressed this issue by referring to the same group (i.e., ‘‘other study participants’’) in each item, which was expected to produce a more reliable measure of perceived similarity. 3 A separate sample of 112 individuals from the same online population as in Study 2 completed both this measure (a ¼ .88) and the 10-item Need to Belong scale (Leary, Kelly, Cottrell, & Schreindorfer, 2007) (a ¼ .77). The two measures, which were presented in randomized order and interspersed with several other individual difference questionnaires, correlated strongly with one another (r ¼.69, p < .001).

Need to belong and perceived consensus 709

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

complete, it is unlikely that participants’ attitudes toward the

referendum primed them to have a higher or lower subsequent

need to belong.

Results

It was hypothesized that participants high in need to belong

would perceive greater consensus for their opinions on the

naturalization of immigrants, particularly if the issue was

important to them personally. To test this hypothesis, the

results were submitted to a multiple regression with perceived

similarity as the outcome variable (Aiken & West, 1991), and

to a logistic regression with the opinion congruence measure

as the outcome variable. Prior to analyzing the data, all

continuous predictor variables were mean-centered to aid in

interpretation. The demographic controls (i.e., age, sex, and

language) were entered into the first block of the equation.

Need to belong and personal importance scores were

entered into the second block. The need to belong � personal personal importance interaction term was entered into the third

and final block. Each predictor variable was interpreted in the

first block that it appeared.

Perceived Similarity

Participants perceived greater similarity between their own

and others’ opinions the higher they were in need to belong

(b ¼ .18), t(1232) ¼ 6.46, p < .001, and the more personally important they felt the issue was (b ¼ .06), t(1232) ¼ 1.95, p ¼ .05. The only other significant effect was the hypothesized interaction between need to belong and personal importance

(b ¼ .06), t(1231) ¼ 2.05, p ¼ .04, model R2 ¼ .04 (see Figure 1). Decomposition of this interaction at one standard

deviation above and below the mean personal importance

score revealed that the relationship between need to belong and

perceived similarity of opinion was stronger for participants

high in issue importance (b ¼ .23), t(1231) ¼ 6.15, p < .001, than for participants low in issue importance (b ¼ .12), t(1231) ¼ 3.02, p < .005. Additionally, importance predicted greater perceived similarity for participants high in need to

belong (i.e., 1 SD above the mean), (b ¼ .09), t(1231) ¼ 2.75, p < .01, but not for participants low in need to belong (i.e., 1 SD below the mean), (b ¼ �.02), t(1231) < 1, ns.

Opinion Congruence

On this measure, the only result to emerge was a marginally

significant interaction between need to belong and issue

importance (B ¼ 0.15, SE ¼ 0.09), x2(1) ¼ 2.98, p ¼ .08, model R

2 ¼ .01 (see Figure 2). Simple slopes analyses indicated that need to belong predicted opinion congruence

among participants high in issue importance, (B ¼ 0.21, SE ¼ 0.12), x

2 (1) ¼ 3.44, p ¼ .06, but not among those low in issue

importance, (B ¼ �0.06, SE ¼ 0.12), x2(1) ¼ .25, ns. More- over, issue importance predicted opinion congruence among

participants high in need to belong, (B ¼ 0.21, SE ¼ 0.10),

Table 1. Correlations between measured variables, Study 1

Attitude Need to belong

Personal importance

Perceived similarity

Opinion congruence Age Sex Language

Attitude .13 ��

.04 .25 ��

.05 .07 � �.03 �.11��

Need to belong .13 ��

.01 .18 ��

.03 �.01 .01 �.07� Personal importance .04 .01 .06

� .06 .04 �.01 �.003

Perceived similarity .25 ��

.18 ��

.06 �

.08 �

.04 �.03 �.02 Opinion congruence .05 .03 .06 .08

� .04 �.05 �.03

Age .07 � �.01 .04 .04 .04 �.03 .003

Sex �.03 .01 �.01 �.03 �.05 �.03 �.01 Language �.11�� .07� �.003 �.02 �.03 .003 �.01 � p < .05, ��p < .01.

Figure 1. Perceived similarity as a function of need to belong and personal importance (þ/� 1 SD), Study 1

Figure 2. Opinion congruence as a function of need to belong and personal importance (þ/� 1 SD), Study 1

rg Matthes710 Kimberly Rios Morrison and Jö

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

x 2 (1) ¼ 4.64, p ¼ .03, but not among those low in need to

belong, (B ¼ �0.06, SE ¼ 0.13), x2(1) ¼ .20, ns.

Discussion

Using a representative sample of Swiss citizens, Study 1

provided evidence that need to belong predicts higher

estimates of consensus on a controversial issue (i.e., stricter

naturalization rules for immigrants), especially among those

who consider the issue personally important. Participants high

in need to belong and issue importance perceived greater

similarity between their own and others’ opinions on both

direct measures of consensus (i.e., being asked how similar

their opinions are to others’ opinions) and more indirect

measures (i.e., being asked to provide their own opinions and

estimates of the majority opinion separately). Notably, the

correlation between these two consensus measures, though

significant, was quite weak. This may have been due to the

differences in reference groups across the two measures

(e.g., ‘‘one’s surroundings’’ on perceived similarity, ‘‘Swiss

citizens’’ on opinion congruence). 4

Unfortunately, it was not possible to assess participants’

accuracy in their perceptions of consensus, as there was no

clear majority position on the issue of immigrant naturaliz-

ation. Because equal numbers of participants were opposed to

(i.e., scored between 1 and 5, n ¼ 626) and in favor of (i.e., scored between 6 and 10, n ¼ 615) stricter naturalization rules, x 2 (1) ¼ .10, ns, we could not determine whether participants

correctly guessed the opinion of the majority. To more

definitively test the relationship between need to belong and

overestimates (i.e., false perceptions) of consensus, in Study 2

we incorporated measures of false consensus from previous

research (e.g., Campbell, 1986; Mikulincer et al., 1998), as well

as the measures used in Study 1. In addition, we used a different

social issue and population—a proposed alcohol tax in the

United States—to increase the generality of our findings.

By using a random digit dial procedure and a real-life social

issue, Study 1 provides compelling evidence for the external

validity of our paradigm. Even though the effect sizes were

smaller than those typically observed in controlled experi-

mental settings (see Hovland, 1959; Peterson, 2001), the fact

that need to belong and consensus perceptions were related at

all is quite surprising, given that the need to belong scale was

administered at a different, much later point in time than the

consensus measures. However, one limitation of Study 1 is that

the correlational design did not allow for a test of causality.

To this end, a final objective of Study 2 was to ascertain that

the need to belong leads people—at least those who view the

issue at hand as personally important—to perceive greater

consensus for their opinions. Need to belong was experimen-

tally manipulated by priming participants with words related to

either rejection or acceptance (see Sommer & Baumeister,

2002). According to theory and research on the need to belong,

rejection primes heighten belongingness needs, whereas

acceptance primes reduce or satisfy them (Baumeister &

Leary, 1995; Leary, Tambor, Terdal, & Downs, 1995).

STUDY 2

Method

Sixty-eight individuals (23 men, 44 women, 1 unspecified;

Mage ¼ 35.50, SD ¼ 11.78) from all areas of the United States were recruited through an email announcement to participate

in this study. The announcement was sent to people who had

signed up to receive notifications of paid online studies,

sponsored by social science faculty at The Ohio State

University. In exchange for completing the study, participants

were entered into a drawing to win one of several $25 gift

certificates to a major online retailer.

Three participants did not complete the need to belong

manipulation described below and thus were excluded from

analyses. The remaining 65 participants were retained in the final

sample. Of these, 30 were randomly assigned to the high need to

belong condition, and 35 to the low need to belong condition.

All experimental materials were presented online. The

study was described as an investigation of the relationship

between language usage and social attitudes. Participants first

completed the need to belong manipulation, described as a

language judgment task. On the next screen, they read that

policy makers in several states had proposed a 10-cent per

drink alcohol tax increase. At the time that the study was run,

this issue was actually under debate among politicians and

the general public in the United States. Participants then

reported their own attitude toward the proposed tax increase

and completed several measures of their perceptions of others’

attitudes (i.e., perceived consensus for opinion). Finally, they

completed a mood measure, a demographic survey, a measure

of personal importance, and a suspicion probe. None of the

participants correctly guessed the purpose of the study.

Need to Belong Manipulation

The need to belong manipulation, taken from Sommer and

Baumeister (2002), consisted of 20 sets of four words. Partici-

pants were instructed to form a meaningful three-word phrase

from each set by leaving out the ‘‘filler’’ word (e.g., cake could

ate she would become she ate cake). In the high need to belong

condition, 10 of the sets contained a rejection-related word

(e.g., from isolated on others, feels she top rejected). In the low

need to belong condition, 10 of the sets contained an acceptance-

related word (e.g., by included on others, feels she top accepted).

Previous research has shown that being primed with negative,

rejection-unrelated words does not produce the same results

as being primed with rejection-related words (Sommer &

Baumeister, 2002). Thus, any effects of the need to belong

manipulation cannot be attributed to differences in mood alone. 5

4 Indeed, the correlation between the perceived similarity and opinion con- gruence measures in Study 2 (which used the same reference group—‘‘other participants’’) was much stronger, r(63) ¼ .54, p < .001. It should be noted, however, that direct and indirect measures of self-other similarity tend to exhibit inconsistent relationships (see Douglas & Sutton, 2004).

5 To ensure that being primed with rejection-related words (versus acceptance- related words) activates belongingness needs over and beyond any effects of mood, a pre-test was conducted with a separate group of 109 individuals from the same population as in this study. Participants completed the need to belong manipulation, followed by the five need to belong items from Study 1 (a ¼ .79) and a measure of their current mood (1 ¼ very negative, 7 ¼ very positive). As expected, participants primed with rejection words (n ¼ 48, M ¼ 3.31, SD ¼ 0.75 on a five-point scale) had higher need to belong scores than did those primed with acceptance words (n ¼ 61, M ¼ 3.02, SD ¼ 0.93), F(1, 107) ¼ 3.09, p ¼ .04 (one-tailed). However, participants in the rejection and acceptance conditions showed no differences in mood, F(1, 107) ¼ 0.05, p ¼ .42 (one-tailed). Furthermore, the effect of condition on need to belong scores held after controlling for mood, F(1, 105) ¼ 3.42, p ¼ .03 (one-tailed).

Need to belong and perceived consensus 711

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

Attitude Toward Alcohol Tax

Participants reported their own attitude toward the proposed

alcohol tax increase on a five-point scale (1 ¼ strongly oppose, 2 ¼ somewhat oppose, 3 ¼ neither oppose nor favor, 4 ¼ somewhat favor, 5 ¼ strongly favor). The sample mean was 3.52 (SD ¼ 1.28).

The need to belong manipulation did not significantly

affect participants’ attitudes toward the alcohol tax, F(1, 63) ¼ 1.53, p ¼ .22.

Consensus Measures

Participants estimated the percentages of ‘‘other participants in

the study’’ who held each of the following attitudes toward the

alcohol tax: Strongly oppose, somewhat oppose, neither

oppose nor favor, somewhat favor, and strongly favor. 6

Participants were instructed that their percentages needed to

add up to 100; they had no problems following these instructions.

Their estimates were used to compute measures of false con-

sensus and overestimation–underestimation (see Results sec-

tion). This procedure has been used in prior work on consensus

estimates (e.g., Campbell, 1986; Mikulincer et al., 1998).

Additionally, participants used the same five-point scale

(1 ¼ strongly oppose, 5 ¼ strongly favor) to report what they thought the attitude of ‘‘most other participants’’ was toward

the alcohol tax. As in Study 1, an opinion congruence measure

was computed by coding whether participants’ own attitudes

matched their perceptions of most other participants’ attitudes

(0 ¼ did not perceive opinion congruence, n ¼ 51; 1 ¼ perceived opinion congruence, n ¼ 14).

Participants also responded to four perceived similarity

items, similar to those used in Study 1 (e.g., ‘‘Most participants

in this study have the same opinion on the proposed alcohol tax

increase as I do’’). Their responses were averaged to form a

composite, with higher scores reflecting greater perceived

similarity between one’s own and others’ opinions (a ¼ .78; M ¼ 2.71, SD ¼ 0.83 on a five-point scale).

Mood

Participants completed a shortened version of the Positive and

Negative Affect Schedule (PANAS; Thompson, 2007). This

version consisted of five positive emotions (active, alert,

attentive, determined, inspired; a ¼ .90) and five negative emotions (afraid, ashamed, hostile, nervous, upset; a ¼ .90), presented in random order. For each emotion, participants

indicated the extent to which they currently felt that way (1 ¼ not at all, 5¼ extremely). Higher scores on the positive [negative] mood index corresponded to more positive [negative] affect.

Personal Importance

At the end of the study, participants responded to two questions

designed to assess issue importance, on a scale from 1 (not at

all) to 5 (very much): ‘‘How important is the proposed alcohol

tax increase to you?’’ and ‘‘How relevant is the proposed

alcohol tax increase to you personally?’’ Their responses were

averaged into a composite (a ¼ .84; M ¼ 2.29, SD ¼ 1.22). Participants’ scores on this composite did not differ by

condition, F(1, 63) ¼ 0.12, p ¼ .72, suggesting that the need to belong manipulation did not affect how important the issue

was to them.

Results

Prior to analyzing the results, two scores were computed

for each participant: False consensus and overestimation–

underestimation. False consensus was calculated by taking the

difference between each participant’s estimated percentage of

others who held his/her own attitude, and the estimated

percentage for that attitude given by all participants who did

not hold the attitude. A positive score indicated a false

consensus effect (M ¼ 6.11, SD ¼ 14.41). Unlike the false consensus measure, which involves

comparing two estimates of the prevalence of an attitude

(i.e., from those who do and do not hold that attitude),

the measure of overestimation–underestimation is based on

comparisons with the actual distribution of others’ attitudes.

As such, it provides a truer test of whether people are correct

in their perceptions of consensus (Marks & Miller, 1987).

Overestimation–underestimation was calculated by taking the

difference between the participant’s estimated percentage of

others who held his/her own attitude, and the percentage of all

participants in the sample who held that attitude. A positive

score indicated overestimation, and a negative score indicated

underestimation, of the prevalence of one’s attitude

(M ¼ �0.22, SD ¼ 19.65). It was hypothesized that participants primed with rejection-

related words (i.e., in the high need to belong condition) would

exhibit higher scores on the consensus measures than

participants primed with acceptance-related words (i.e., in

the low need to belong condition). However, this effect was

expected to be strongest among participants who considered

the proposed alcohol tax increase important to them

personally. To test the hypothesis, participants’ scores on

the consensus measures were submitted to a need to belong

condition (0 ¼ low, 1 ¼ high) � personal importance (mean- centered continuous variable) regression (Aiken & West,

1991). As in Study 1, the opinion congruence measure was

submitted to a logistic regression analysis. The main effects were

entered and interpreted in the first block, and the interaction term

in the second block, of the analyses. To decompose the

interactions, the relationship between need to belong condition

and each dependent measure was assessed at one standard

deviation above and below the mean personal importance score.

Furthermore, the relationship between personal importance

and each dependent measure was assessed in the high and low

need to belong conditions separately. (See Table 2 for

correlations between the measured variables in this study.)

False Consensus

Participants in the high need to belong condition exhibited

greater false consensus for their attitudes toward the proposed

6 ‘‘Other participants’’ were used as the comparison group because they provided the most precise test of the false consensus effect. That is, partici- pants’ estimates of the attitudes of other participants were compared with the actual attitudes of the sample.

rg Matthes712 Kimberly Rios Morrison and Jö

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

alcohol tax increase than did participants in the low need

to belong condition (b ¼ .27), t(62) ¼ 2.27, p < .03. Most important, the hypothesized interaction between need to

belong condition and personal importance was significant

(b ¼ .33), t(61) ¼ 2.30, p < .03, model R2 ¼ .17 (see Figure 3). Simple slopes analyses revealed that need to belong predicted

greater false consensus among participants high in personal

importance (b ¼ .57), t(61) ¼ 3.28, p < .005, but not among participants low in personal importance (b ¼ �.01), t(61) < 1, ns. Moreover, personal importance was associated with greater

false consensus in the high need to belong condition (b ¼ .57), t(61) ¼ 2.77, p < .01, but not in the low need to belong condition (b ¼ �.01), t(61) < 1, ns. When the false consensus scores of participants in each cell were compared to zero, those

high in need to belong and personal importance exhibited a

positive, significant false consensus effect (t ¼ 4.76, p < .001). The false consensus scores of all three other groups of

participants did not differ from zero (ts < 1, ns).

Overestimation–underestimation

Participants high in personal importance overestimated the

prevalence of their attitudes toward the proposed alcohol tax

increase to a greater extent than did participants low in

personal importance (b ¼ .37), t(62) ¼ 3.21, p < .01. The only other significant effect to emerge was the hypothesized need to

belong condition � personal importance interaction (b ¼ .32), t(61) ¼ 2.38, p ¼ .02, model R2 ¼ .24 (see Figure 4). Decompo- sition of the interaction indicated that need to belong was

associated with greater overestimation among participants

high in importance (b ¼ .47), t(61) ¼ 2.82, p < .01, but not among participants low in importance (b ¼ �.11), t(61) < 1, ns. Furthermore, the relationship between importance and

overestimation–underestimation was significant in the high

need to belong condition (b ¼ .76), t(61) ¼ 3.85, p < .001, but not in the low need to belong condition (b ¼ .19), t(61) ¼ 1.62, p ¼ .17.

Comparing the scores of participants in each cell to zero

revealed that participants high in both need to belong and

personal importance significantly overestimated the popularity

of their opinions (t ¼ 3.62, p ¼ .001). By contrast, participants high in need to belong and low in personal importance

(t ¼ �2.26, p < .03), as well as participants low in both need to belong and personal importance (t ¼ �1.70, p ¼ .09) tended to underestimate the popularity of their opinions. Participants

low in need to belong but high in personal importance were the

only group who correctly perceived how widespread their

opinions were (t ¼ .12, ns).

Perceived Similarity

Participants in the high need to belong condition had higher

scores on the measure of perceived similarity than did

participants in the low need to belong condition (b ¼ .24), t(62) ¼ 2.12, p < .04. Additionally, participants high in personal importance had higher perceived similarity scores

than did participants low in personal importance (b ¼ .43), t(62) ¼ 3.85, p < .001. Most germane to the hypothesis, the

Table 2. Correlations between measured variables, Study 2

Attitude Personal

importance False

consensus Overestimation– underestimation

Perceived similarity

Opinion congruence

Attitude �.36�� .16 �.67�� .17 �.27� Personal importance �.36�� .20 .37� .40�� .37�� False consensus �.16 .20 .82�� .58�� .46�� Overestimation–underestimation �.67�� .37� .82�� .53�� .50�� Perceived similarity �.17 .40�� .58�� .53�� .54�� Opinion congruence �.27� .37�� .46�� .50�� .54�� p < .05,

�� p < .01.

Figure 3. False consensus as a function of condition (high vs. low need to belong) and personal importance (þ/� 1 SD), Study 2

Figure 4. Overestimation–underestimation as a function of con- dition (high vs. low need to belong) and personal importance (þ/� 1 SD), Study 2

Need to belong and perceived consensus 713

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

need to belong condition � personal importance interaction was significant (b ¼ .45), t(61) ¼ 3.58, p ¼ .001, model R2 ¼ .36 (see Figure 5). Simple slopes analyses indicated that the

relationship between need to belong and perceived similarity

was present among participants high in importance (b ¼ .64), t(61) ¼ 4.20, p < .001, but not among participants low in importance (b ¼ �.15), t(61) ¼ �1.00, p ¼ .32. Furthermore, personal importance was associated with higher perceived

similarity scores in the high need to belong condition (b ¼ .96), t(61) ¼ 5.33, p < .001, but not in the low need to belong condition (b ¼ .18), t(61) ¼ 1.42, p ¼ .16.

Opinion Congruence

Participants in the high need to belong condition perceived

congruence between their own and others’ opinions marginally

more frequently than did those in the low need to belong

condition (B ¼ 1.42, SE ¼ 0.74), x2(1) ¼ 3.63, p < .06. Furthermore, participants perceived opinion congruence more

frequently to the extent that the issue was important to them

(B ¼ 0.91, SE ¼ 0.30), x2(1) ¼ 9.03, p < .005. The predicted interaction between need to belong condition and issue

importance was also significant (B ¼ 1.84, SE ¼ 0.88), x 2 (1) ¼ 4.36, p < .04 (see Figure 6), model R2 ¼ .31. Simple

slopes analyses indicated that participants high in issue

importance perceived opinion congruence more frequently

in the high than low need to belong condition (B ¼ 2.68, SE ¼ 0.97), x2(1) ¼ 7.62, p < .01, whereas participants low in issue importance did not differ in their perceptions of opinion

congruence as a function of need to belong condition

(B ¼ �1.82, SE ¼ 1.70), x2(1) ¼ 1.15, p ¼ .28. Additionally, personal importance predicted more frequent perceptions of

opinion congruence in the high need to belong condition

(B ¼ 2.23, SE ¼ 0.82), x2(1) ¼ 7.49, p < .01, but not in the low need to belong condition (B ¼ 0.38, SE ¼ 0.34), x 2 (1) ¼ 1.29, p ¼ .26.

Mood

To ascertain that the present results were not due to differences

in mood, the need to belong condition � personal importance regression was re-run with positive and negative mood as

dependent measures. There was an unexpected relationship

between personal importance and positive mood (b ¼ .25), t(62) ¼ 2.07, p ¼ .04, but neither the main effect of need to belong condition (b ¼ .19), t(62) ¼ 1.60, p ¼ .12, nor the condition � personal importance interaction (b ¼ .21), t(61) ¼ 1.40, p ¼ .16, was significant. No significant effects emerged on negative mood (ps > .29). Additionally, when both positive and negative mood were included in the primary analyses as

covariates, the condition � personal importance interaction (and the simple effect of need to belong condition on

consensus estimates at high levels of importance) remained

significant at p < .05 on each dependent measure.

Discussion

The results of Study 2 supported the hypothesis that

belongingness needs, combined with high personal import-

ance, would produce inflated estimates of consensus on a

social issue (i.e., a proposed alcohol tax increase). Participants

primed with rejection-related words exhibited greater false

consensus effects, overestimated the popularity of their

opinions, were less accurate in their estimates of others’

opinions, and perceived more similarity between their own and

others’ opinions than did participants primed with acceptance-

related words. These relationships were strongest among

participants who reported that the issue was important to

them personally, and they held after controlling for both

positive and negative mood.

Study 2 provided the first demonstration that experimen-

tally manipulating the need to belong can influence consensus

perceptions on important issues. (See Table 3 for a summary

of which main effects and interactions were and were not

significant.) Moreover, by assessing participants’ accuracy in

their perceptions—specifically, by using standard measures of

false consensus and overestimation–underestimation as well as

comparing participants’ scores on these measures to zero—

Study 2 showed that those high in need to belong and personal

importance were the only group of participants who

erroneously thought their opinions were widely held.

Both Studies 1 and 2, however, leave open the question of

whether high consensus estimates can actually reduce people’s

Figure 5. Perceived similarity as a function of condition (high vs. low need to belong) and personal importance (þ/� 1 SD), Study 2

Figure 6. Opinion congruence as a function of condition (high vs. low need to belong) and personal importance (þ/� 1 SD), Study 2

rg Matthes714 Kimberly Rios Morrison and Jö

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

belongingness needs. We propose that one way to test this is to

induce individuals to believe they hold a common opinion,

which should theoretically satisfy belongingness needs, and

then measure their need to belong. Supporting our reasoning,

previous studies have shown that making people feel

connected to others (e.g., by having them think of their group

memberships; Knowles & Gardner, 2008; by giving them a

‘‘nostalgic’’ product associated with their childhood relation-

ships; Loveland, Smeesters, & Mandel, 2010) can lower

belongingness needs and in some cases elevate mood and self-

esteem. Thus, to examine the ultimate effects of consensus

information on satisfaction of belongingness needs, in Study 3

we gave participants false information that their opinions

on the alcohol tax from Study 2 were either common or

uncommon, prior to assessing their need to belong. It was

predicted that participants who considered the alcohol tax

important to them would have a lower need to belong upon

learning that they held a common than uncommon opinion.

STUDY 3

Method

Participants

One hundred and twenty-three individuals (40 men, 82

women, 1 unspecified; Mage ¼ 35.93, SD ¼ 11.05), recruited from the same database as in Study 2, participated in this

experiment in exchange for a $5 gift certificate. Eight

participants who did not believe the information about the

commonness of their opinions, six participants who failed the

manipulation check (described below), and one participant

who did not report his/her attitude toward the alcohol tax were

omitted from all analyses. One additional participant, whose

data rendered him/her a statistical outlier (studentized deleted

residual ¼ 3.06, p < .001; McClelland, 2000) was excluded, leaving 107 individuals in the final sample. The excluded

participants were relatively evenly distributed across exper-

imental conditions.

Procedure and Materials

All stimulus materials were presented online. Upon logging

into the study, participants read that they would answer

questions about a proposed alcohol tax increase of 10 cents per

drink. First, they indicated on a two-point scale whether they

favored (n ¼ 67) or opposed (n ¼ 40) the alcohol tax. They then responded to two items assessing personal importance of the

alcohol tax (1 ¼ not at all, 5 ¼ very much): ‘‘How important is the proposed alcohol tax increase to you personally?’’ and

‘‘How relevant is the proposed alcohol tax increase to you?’’

These items were averaged into a composite (a ¼ .90; M ¼ 2.96, SD ¼ 1.22).

To separate the attitudes and issue importance measures

from the rest of the study, participants completed a filler task

that took approximately 5 minutes. Then they read a brief

paragraph containing (fictitious) ‘‘background information’’

about the alcohol tax. One group of participants (n ¼ 60) was randomly assigned to read that a majority of participants in a

previous survey (76%) favored the tax increase. Another group

of participants (n ¼ 47) was randomly assigned to read that a minority of participants (24%) favored it. Participants who

received information indicating that a majority of other survey-

takers held the same opinion as they did were coded as being in

the common opinion condition (n ¼ 51), whereas those who received information indicating that a majority of other survey-

takers held a different opinion than they did were coded as

being in the uncommon opinion condition (n ¼ 56). (For a similar procedure, see Morrison & Wheeler, 2010.)

Next, as a manipulation check, participants indicated

whether a majority or minority of other survey-takers favored

the alcohol tax. Only those participants who answered this

question correctly were retained in the final sample.

On the following screen, participants completed the five-

item need to belong measure used in Study 1 (a ¼ .83; M ¼ 3.24, SD ¼ 0.92). Participants’ need to belong scores were not significantly related to their attitudes toward the

alcohol tax, which they had reported earlier in the study F(1,

105) ¼ 0.99, p ¼ .32. Finally, participants were probed for suspicion and filled out

a demographic survey, prior to being fully debriefed. 7

Participants who suspected that the ‘‘background information’’

about the alcohol tax was not real or who thought that it was

meant to influence their need to belong scores were excluded

from analyses.

Results

It was hypothesized that among participants who considered

the alcohol tax personally important, those who received

Table 3. Summary of significant main effects and interactions (Studies 1 and 2)

Need to belong Personal importance Need to belong � personal importance Perceived similarity (Study 1)

� � �

Opinion congruence (Study 1) ��

False consensus (Study 2) � �

Overestimation–underestimation (Study 2) � �

Perceived similarity (Study 2) � � �

Opinion congruence (Study 2) ��

� p < .05, ��p < .10.

7 Within the demographic survey, participants reported their current mood (1 ¼ very negative, 7 ¼ very positive). There were no effects of opinion condition or personal importance on this measure, nor was there an opinion condition � importance interaction (ps > .47). Moreover, when mood was entered into the primary analysis as a covariate, the opinion condition � impor- importance interaction on need to belong held (b ¼ .26), t(102) ¼ 1.93, p ¼ .056.

Need to belong and perceived consensus 715

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

feedback that their opinions were common would have lower

subsequent need to belong scores than those who received

feedback that their opinions were uncommon, as the former

type of information should mitigate their belongingness

needs. To test this hypothesis, participants’ need to belong

scores were submitted to an opinion condition (0 ¼ common, 1 ¼ uncommon) � personal importance (centered continuous variable) regression analysis. Participants who had higher need

to belong scores considered the alcohol tax more important

to them (b ¼ .20), t(104) ¼ 2.10, p < .04. Additionally, the predicted interaction between opinion condition and personal

importance was significant (b ¼ .27), t(103) ¼ 2.03, p < .05, model R

2 ¼ .10 (see Figure 7). Simple slopes analyses revealed that participants high in

issue importance had lower need to belong scores when they

learned that their opinions were common than uncommon

(b ¼ .31), t(103) ¼ 2.26, p < .03. Conversely, the need to belong scores of participants low in issue importance did

not differ as a function of opinion condition (b ¼ �.08), t(103) < 1, ns. Looking at the interaction another way, among participants in the uncommon opinion condition, those who

considered the alcohol tax more important to them had higher

need to belong scores (b ¼ .39), t(103) ¼ 2.94, p < .005. Among participants in the common opinion condition, there

was no relationship between issue importance and need to

belong (b ¼ .01), t(103) < 1, ns.

Discussion

The results of Study 3 provide evidence that consensus

information can indeed reduce belongingness needs, especi-

ally on important issues. Participants who received random

feedback that their opinions on a personally important issue

were common had less of a need to belong, relative to

participants who received random feedback that their opinions

were uncommon. Reading that a majority of survey-takers

shared one’s opinion presumably reassured participants that

they fit in with others. However, it is possible that learning of

one’s majority opinion status not only reduced participants’

belongingness needs, but also heightened their distinctiveness

needs. According to optimal distinctiveness theory (Brewer,

1991), people are motivated to be different from as well as

similar to others. These competing motives are manifest both

at the group level (e.g., whereby people seek to differentiate

their own group from other groups; Jetten, Spears, & Postmes,

2004) and at the individual level (e.g., whereby people seek to

establish their uniqueness from other members of their group;

Hornsey & Jetten, 2004; Imhoff & Erb, 2009; Morrison &

Wheeler, 2010). Thus, perhaps participants in Study 3 who

learned that their opinions were shared by most other people

felt that their sense of uniqueness was threatened, and they

downplayed their need to belong. This possibility should be

more thoroughly investigated in future research.

GENERAL DISCUSSION

Under what conditions will individuals perceive higher levels

of consensus for their opinions than is warranted? Although

some research has indicated that people’s beliefs about the

popularity of their opinions are largely accurate (e.g., Nisbett

& Kunda, 1985), other work suggests that error can ensue

(e.g., Ross et al., 1977). The present studies introduce one

potential source of misperception. Specifically, individuals

who have a high need to belong are motivated to believe that

most others agree with them, in order to feel as though they

fit into their peer group. These studies are thus the first to

directly demonstrate that need to belong can produce inflated

consensus estimates.

The present research is also the first to show that people

high in need to belong are most prone to overestimate

consensus on issues that are important to them personally.

As such, perceiving high levels of consensus for one’s

opinions, like conforming to the majority group opinion

(Williams et al., 2000) and forming new social connections

(Maner et al., 2007), is another means through which people

can attempt to satisfy their motives to fit in with others. On the

surface, the fact that having a high need to belong can lead to

inflated consensus estimates as well as to other-oriented,

prosocial behaviors (e.g., conformity, seeking social connec-

tions) may seem inconsistent with some prior research, which

has portrayed false consensus as a largely egocentric

phenomenon (e.g., Hoch, 1987; Krueger & Clement, 1994).

However, the present work suggests that somewhat para-

doxically, egocentrism—as manifested here in perceptions

of opinion consensus—is one strategy that people may

sometimes use to satisfy their motives to relate to other people

(see Hess & Pickett, 2010).

The effects of need to belong and personal importance

on perceived consensus were replicated in two different

countries, using multiple methodologies (nationally represen-

tative survey in Study 1, experiment in Study 2) and issues

(naturalization of immigrants in Study 1, proposed alcohol

tax increase in Study 2). As such, these results demonstrate

both external and internal validity of the relationship between

need to belong, importance, and consensus estimates. Further-

more, the key finding from Study 3—that people who

considered the alcohol tax important to them had a lower

need to belong upon learning of the supposed commonness of

their opinion—suggests that perceiving high levels of con-

sensus can in fact quell people’s fundamental motives to

connect with others.

Figure 7. Need to belong as a function of opinion condition (com- mon vs. uncommon) and personal importance (þ/� 1 SD), Study 3

rg Matthes716 Kimberly Rios Morrison and Jö

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

Notably, the present findings demonstrate that belonging-

ness needs and personal importance trigger biased estimates

of consensus. In other words, participants with a high need to

belong—and who considered the issue a central part of their

self-concept—did not simply believe that their opinions were

more widely held than did participants with a low need to

belong or who did not consider the issue a central part of their

self-concept. Rather, as Study 2 demonstrated by comparing

participants’ consensus estimates to the real distribution of

others’ opinions, these individuals were so motivated to seek

social support for their opinions that they perceived support

even when it was not present. That is, people high in need to

belong and issue importance perceived their opinions to be

more common than was actually the case.

Future studies should explore the process by which people

high in need to belong come to believe that their opinions are

widely held. By showing that consensus information reduces

people’s belongingness needs, Study 3 provides evidence that

the need to belong can motivate (false) perceptions of opinion

consensus. How, though, do these misperceptions come

about in the first place? One possibility is that when people’s

belongingness needs are heightened, they selectively recall

information suggesting that others’ opinions are congruent

with their own, just as they become more prone to remember

social information in general (see Gardner et al., 2000). This

possibility is not incompatible with the present findings, as

selective memory could mediate the relationship between need

to belong, personal importance, and consensus estimates.

Another potential account, supported by some previous

research (e.g., Bosveld, Koomen, & van der Pligt, 1994;

Wojcieszak & Price, 2009), is that people high in the need to

belong tend to selectively affiliate with like-minded others.

However, the need to belong was experimentally manipulated

in Study 2, thereby rendering it impossible for participants

to spontaneously decide to affiliate with others who shared

their opinions. Furthermore, the consensus measures in all

studies involved rather obscure reference groups with which

participants were unlikely to have high levels of affiliation

(e.g., ‘‘my surroundings’’ in Study 1, ‘‘other participants’’ in

Studies 2 and 3). It is, therefore, unlikely that selective

exposure to similar opinions was responsible for our results.

In addition to having implications for research on the need

to belong, the present studies lend insight into when personal

importance does and does not inflate estimates of consensus.

Prior research on the relationship between importance and

perceived consensus has produced conflicting results. Whereas

some studies have shown that overestimates of consensus

are more likely to emerge on highly important relative to

unimportant issues (Crano, 1983), other studies have shown

either the opposite (Campbell, 1986) or no relationship at all

(Fabrigar & Krosnick, 1995). By looking at the moderating

role of need to belong, the current studies help to resolve these

apparently contradictory findings. People high in need to

belong, who care about fitting in with their peers, are motivated

to believe others share their opinions to the extent that the issue is

important to them personally. But people low in need to belong,

who are unperturbed by the prospect of social exclusion, instead

appear motivated to correctly perceive others’ opinions on highly

important issues. Indeed, the overestimation–underestimation

findings from Study 2 indicate that participants low in need to

belong, but high in personal importance, scored closer to zero

than any other group. That is, they neither overestimated nor

underestimated consensus for their opinions.

In the future, it would be illuminating to test whether

personal importance ever leads to underestimates of con-

sensus. For instance, people who are high in need for

uniqueness (Imhoff & Erb, 2009; Kernis, 1984; Snyder &

Fromkin, 1980) may be inclined to think that their opinions

on personally important issues are uncommon. For these

individuals, the motivation to be distinctive may override

any competing needs they may have to fit in with others

(see Brewer, 1991). Additionally, given that opinions have

been shown to produce larger false consensus effects than

other domains, such as perceptions of (high) ability (Campbell,

1986), it remains unclear if the effects of need to belong on

consensus perceptions are specific to opinions. One likely

possibility is that the need to belong triggers greater

false consensus effects in socially undesirable domains, be

they opinions (e.g., racial prejudice) or abilities (e.g., low

intelligence), than in socially desirable domains (e.g., racial

tolerance, high intelligence). In the latter domains, belong-

ingness needs may even lead to false uniqueness effects (Suls

& Wan, 1987; see also Blanton & Christie, 2003). Another

possibility is that the relationship between need to belong and

overestimates of consensus is more pronounced for opinions

(e.g., attitudes toward political issues) than for preferences

(e.g., taste in music), on which people typically have a stronger

desire to be distinctive (see Spears, Ellemers, & Doosje, 2009).

Given that perceived consensus for one’s opinion has been

shown to increase opinion expression (Matthes et al., 2010;

Miller & Morrison, 2009), the current results have somewhat

ironic implications for who will voice their opinions on

controversial issues and why. If people who have a high need

to belong develop false beliefs that others share their opinions

on personally important issues, they may ultimately decide

to express these opinions, even if they are in fact minorities.

Thus, people’s tendencies to conform to the majority

viewpoint when they are high in need to belong (see Williams

et al., 2000) may not be as pronounced on issues that they

consider central to the self. Such a process could help

explain why Michael Scott from The Office, despite his

seemingly incorrigible need to belong, never hesitates to voice

idiosyncratic opinions on important workplace issues. This

potential relationship between belongingness needs, consensus

perceptions, and opinion expression represents an intriguing

direction for future research.

Conclusion

The present studies demonstrate that individuals high in

need to belong, be it chronic or experimentally primed, will

misperceive others’ opinions on personally important issues

as congruent with their own. These findings indicate that

although the need to belong makes people more accurate social

perceivers on certain dimensions (e.g., decoding nonverbal

cues; Pickett et al., 2004), it makes them less accurate on other

dimensions—for example, recognition of their peers’ opinions

on issues that matter to them personally. People thus seem

motivated to construe the social environment in whichever way

is most likely to satisfy their belongingness needs, even if they

must deceive themselves in the process.

Need to belong and perceived consensus 717

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

ACKNOWLEDGEMENTS

The research conducted in Study 1 of this article was supported

by a grant from the Swiss National Science Foundation as

a part of the National Center of Competence in Research

(NCCR) ‘‘Challenges to Democracy in the 21st Century.’’

REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks: Sage Publications.

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychologi- cal Bulletin, 117, 497–529.

Blanton, H., & Christie, C. (2003). Deviance regulation: A theory of action and identity. Review of General Psychology, 7, 115–149.

Boninger, D. S., Krosnick, J. A., & Berent, M. K. (1995). Origins of attitude importance: Self-interest, social identification, and value relevance. Journal of Personality and Social Psychology, 68, 61–80.

Bosveld, W., Koomen, W., & van der Pligt, J. (1994). Selective exposure and the false consensus effect: The availability of similar and dissimilar others. British Journal of Social Psychology, 33, 457–466.

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.

Campbell, J. D. (1986). Similarity and uniqueness: The effects of attribute type, relevance, and individual differences in self-esteem and depression. Journal of Personality and Social Psychology, 50, 281–294.

Carvallo, M., & Gabriel, S. (2006). No man is an island: The need to belong and dismissing avoidant attachment. Personality and Social Psychology Bulletin, 32, 697–709.

Carvallo, M., & Pelham, B. W. (2006). When fiends become friends: The need to belong and perceptions of personal and group discrimination. Journal of Personality and Social Psychology, 90, 94–108.

Crano, W. D. (1983). Assumed consensus of attitudes: The effect of vested interest. Personality and Social Psychology Bulletin, 9, 597–608.

Douglas, K. M., & Sutton, R. M. (2004). Right about others, wrong about ourselves? Actual and perceived self-other differences in resistance to persuasion. British Journal of Social Psychology, 43, 585–603.

Imhoff, R., & Erb, H. P. (2009). What motivates nonconformity? Uniqueness seeking blocks majority influence. Personality and Social Psychology Bulletin, 35, 309–320.

Fabrigar, L. R., & Krosnick, J. A. (1995). Attitude importance and the false consensus effect. Personality and Social Psychology Bulletin, 21, 468–479.

Gardner, W. L., Pickett, C. L., & Brewer, M. B. (2000). Social exclusion and selective memory: How the need to belong influences memory for social events. Personality and Social Psychology Bulletin, 26, 486–496.

Gardner, W. L., Pickett, C. L., Jefferis, V., & Knowles, M. (2005). On the outside looking in: Loneliness and social monitoring. Personality and Social Psychology Bulletin, 31, 1549–1560.

Hayes, A. F., Matthes, J., & Eveland, W. P. (2010). Stimulating the quasi- statistical organ: Fear of social isolation motivates the quest for knowledge of the opinion climate. Manuscript submitted for publication.

Hess, Y. D., & Pickett, C. L. (2010). Social rejection and self- versus other- awareness. Journal of Experimental Social Psychology, 46, 453–456.

Hoch, S. J. (1987). Perceived consensus and predictive accuracy: The pros and cons of projection. Journal of Personality and Social Psychology, 53, 221–234.

Hornsey, M. J., & Jetten, J. (2004). The individual within the group: Balancing the need to belong with the need to be different. Personality and Social Psychology Review, 8, 248–264.

Hovland, C. I. (1959). Reconciling conflicting results derived from experimental and survey studies of attitude change. American Psychologist, 14, 8–17.

Jetten, J., Spears, R., & Postmes, T. (2004). Intergroup distinctiveness and differentiation: A meta-analytic integration. Journal of Personality and Social Psychology, 86, 862–879.

Katz, D. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 24, 163–204.

Kernis, M. H. (1984). Need for uniqueness, self-schemas, and thought as moderators of the false-consensus effect. Journal of Experimental Social Psychology, 20, 350–362.

Knowles, M. L., & Gardner, W. L. (2008). Benefits of membership: The activation and amplification of group identities in response to social rejection. Personality and Social Psychology Bulletin, 34, 1200–1213.

Krueger, J. (1998). On the perception of social consensus. Advances in Experimental Social Psychology, 30, 163–240.

Krueger, J. I., & Clement, R. W. (1994). The truly false consensus effect: An ineradicable and egocentric bias in social perception. Journal of Personality and Social Psychology, 67, 596–610.

Lakin, J. L., Chartrand, T. L., & Arkin, R. M. (2008). I am too just like you: Nonconscious mimicry as an automatic behavioral response to social exclusion. Psychological Science, 19, 816–822.

Leary, M. R., Kelly, K. M., Cottrell, C. A., & Schreindorfer, L. S. (2007). Individual differences in the need to belong: Mapping the nomological network. Unpublished manuscript, Duke University.

Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995). Self-esteem as an interpersonal monitor: The sociometer hypothesis. Journal of Person- ality and Social Psychology, 68, 518–530.

Loveland, K. E., Smeesters, D., & Mandel, N. (2010). Still preoccupied with 1995: The need to belong and preference for nostalgic products. Journal of Consumer Research, 37, 393–408.

Maner, J. K., DeWall, C. N., & Baumeister, R. F. (2007). Does social exclusion motivate interpersonal reconnection? Resolving the ‘‘porcupine problem.’’ Journal of Personality and Social Psychology, 92, 42–55.

Marks, G., & Miller, N. (1987). Ten years of research on the false-consensus effect: An empirical and theoretical review. Psychological Bulletin, 102, 72–90.

Matthes, J., Morrison, K. R., & Schemer, C. (2010). A spiral of silence for some: Attitude certainty and the expression of political minority opinions. Communication Research, 37, 774–800.

McClelland, G. H. (2000). Nasty data: Unruly, ill-mannered observations can rule your analysis. In H. T. Reis, & C. M. Judd (Eds.), Handbook of research methods in social and personality psychology (pp. 393–411). New York: Cambridge University Press.

Mikulincer, M., Orbach, I., & Iavnieli, D. (1998). Adult attachment style and affect regulation: Strategic variation in self-other appraisals. Journal of Personality and Social Psychology, 75, 436–448.

Miller, N., & Marks, G. (1982). Assumed similarity between self and other: Effect of expectation of future interaction with that other. Social Psychology Quarterly, 45, 100–105.

Miller, D. T., & Morrison, K. R. (2009). Expressing deviant opinions: Believing you are in the majority helps. Journal of Experimental Social Psychology, 45, 740–747.

Miyake, K., & Zuckerman, M. (1993). Beyond personality impressions: Effects of physical and vocal attractiveness on false consensus, social comparison, affiliation, and assumed and perceived similarity. Journal of Personality, 61, 411–437.

Morrison, K. R., & Wheeler, S. C. (2010). Nonconformity defines the self: The role of minority opinion status in self-concept clarity. Personality and Social Psychology Bulletin, 36, 297–308.

Mullen, B., & Hu, L. (1988). Social projection as a function of cognitive mechanisms: Two meta-analytic integrations. British Journal of Social Psychology, 27, 333–356.

Nisbett, R. E., & Kunda, Z. (1985). Perception of social distributions. Journal of Personality and Social Psychology, 48, 297–311.

Peterson, R. A. (2001). On the use of college students in social science research: Insights for a second-order meta-analysis. Journal of Consumer Research, 28, 450–461.

Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on response to argument quantity and quality: Central and peripheral routes to persuasion. Journal of Personality and Social Psychology, 46, 69–81.

Pickett, C. L., & Brewer, M. B. (2001). Assimilation and differentiation needs as motivational determinants of perceived in-group and out-group homogeneity. Journal of Experimental Social Psychology, 37, 341–348.

Pickett, C. L., Gardner, W. L., & Knowles, M. (2004). Getting a cue: The need to belong and enhanced sensitivity to social cues. Personality and Social Psychology Bulletin, 30, 1095–1107.

Pomerantz, E. M., Chaiken, S., & Tordesillas, R. S. (1995). Attitude strength and resistance processes. Journal of Personality and Social Psychology, 69, 408–419.

Robbins, J. M., & Krueger, J. I. (2005). Social projection to ingroups and outgroups: A review and meta-analysis. Personality and Social Psychology Review, 9, 32–47.

Ross, L., Greene, D., & House, P. (1977). The false consensus phenomenon: An attributional bias in self-perception and social perception processes. Journal of Experimental Social Psychology, 13, 279–301.

Sherman, S. J., Presson, C. C., Chassin, L., Corty, E., & Olshavsky, R. (1983). The false consensus effect in estimates of smoking prevalence. Personality and Social Psychology Bulletin, 9, 197–207.

Snyder, C. R., & Fromkin, H. L. (1980). Uniqueness: The human pursuit of difference. New York: Plenum Press.

Sommer, K. L., & Baumeister, R. F. (2002). Self-evaluation, persistence, and performance following implicit rejection: The role of trait self-esteem. Personality and Social Psychology Bulletin, 28, 926–938.

Spears, R., Ellemers, N., & Doosje, B. (2009). Strength in numbers or less is more? A matter of opinion and a question of taste. Personality and Social Psychology Bulletin, 35, 1099–1111.

rg Matthes718 Kimberly Rios Morrison and Jö

Eur. J. Soc. Psychol. 4 , 707 719 (2011)–1

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.

Thompson, E. R. (2007). Development and validation of an internationally reliable short form of the Positive and Negative Affect Schedule (PANAS). Journal of Cross-Cultural Psychology, 38, 227–242.

Thomsen, C. J., Borgida, E., & Lavine, H. (1995). The causes and con- sequences of personal involvement. In R. E. Petty, & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 191–214). Mahwah, NJ: Lawrence Erlbaum Associates.

Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762.

Wojcieszak, M., & Price, V. (2009). What underlies the false consensus effect? How personal opinion and disagreement affect perception of public opinion. International Journal of Public Opinion Research, 21, 25–46.

APPENDIX

Need to belong items

1. One of the worst things that can happen to me is to be

excluded by people I know.

2. It is scary to think about not being invited to social

gatherings by people I know.

3. It would bother me if no one wanted to be around me.

4. It is important to me to fit into the group I am with.

5. I dislike feeling left out of social gatherings, parties, or

other social functions.

Need to belong and perceived consensus 7 91

, 707 719 (2011)–1