Content uploaded by Geoffrey J Leonardelli
Author content
All content in this area was uploaded by Geoffrey J Leonardelli on Aug 14, 2020
Content may be subject to copyright.
Optimal Distinctiveness and Membership Trust 1
RUNNING HEAD: OPTIMAL DISTINCTIVENESS AND MEMBERSHIP TRUST
Optimal Distinctiveness Signals Membership Trust
Geoffrey J. Leonardellia
Denise Lewin Loydb
This is the pre-peer reviewed version of the following article.
Leonardelli, G.J., & Loyd, D.L. (2016). Optimal distinctiveness signals membership
trust. Personality and Social Psychology Bulletin.
Author note. a Corresponding author. Rotman School of Management and Department of
Psychology, University of Toronto, ON M5S 3E6, Canada. Email:
Geoffrey.leonardelli@rotman.utoronto.ca; 416-946-0731 (phone); 416-978-4629 (fax). b Email:
loyddl@illinois.edu; 217-300-6750 (phone); Department of Business Administration, University
of Illinois at Urbana-Champaign, 1206 S 6th Street, Champaign, IL 61820, USA. This work was
supported by SSHRC of Canada grant to the first author (410-2010-1221). We thank Jane Yao
and Konstantin Chestopalov for help in data collection, Marilynn Brewer, Bill von Hippel and
Bob Lount for comments on former drafts, and the Self and Identity Lab (SAIL) for feedback.
Early presentations were discussed at the INGRoup conference (2010) and the Society for
Experimental Social Psychology (2013). GJL devised theory with input from DLL. GJL/DLL
devised method and oversaw data collection. GJL analyzed the data and conducted the literature
review. Authors co-wrote the paper.
Manuscript Word Count: 11653 (total) – 2257 (supplemental) = 9396 (main).
Optimal Distinctiveness and Membership Trust 2
Abstract
According to optimal distinctiveness theory (Brewer, 1991), sufficiently small minority groups
are associated with greater membership trust, even among members otherwise unknown, because
the groups are seen as optimally distinctive. This paper elaborates on the prediction’s
motivational and cognitive processes and tests whether sufficiently small minorities (defined by
relative size; e.g., 20%) are associated with greater membership trust relative to mere minorities
(45%), and whether such trust is a function of optimal distinctiveness. Two experiments,
examining observers’ perceptions of minority and majority groups and using minimal groups and
(in Exp. 2) a trust game, revealed greater membership trust in minorities than majorities. In
Experiment 2, participants also preferred joining minorities over more powerful majorities. Both
effects occurred only when minorities were 20% rather than 45%. In both studies, perceptions of
optimal distinctiveness mediated effects. Discussion focuses on the value of relative size and
optimal distinctiveness, and when membership trust manifests.
Words: 150
Keywords: trust; cooperation; relative group size; optimal distinctiveness; social comparison
Optimal Distinctiveness and Membership Trust 3
Human beings are innately social creatures. Such sociality has its functions, allowing
people the opportunity to collaborate and depend on others, to form groups, and expand our
capabilities beyond those of a single individual. But what sustains that cooperation and maintains
an entire community of people collaborating? We believe it has to do at least in part with group
size, particularly individuals’ perceptions of relative group size. We argue that sufficiently small
groups are more likely to be seen as trustworthy, and when seeking to trust others, individuals
will prefer membership in such groups. We draw from optimal distinctiveness theory (Brewer,
2003; Leonardelli, Pickett, & Brewer, 2010) to support these assertions, and test these
predictions for the first time in two controlled experiments.
Optimally distinctive group memberships are defined by balancing two fundamental
human motives (Brewer, 1991, 2003), the need for inclusion and the need for differentiation.
First, individuals seek group memberships that are large enough to meet a need for inclusion,
where they desire security, group immersion, and similarity. Second, in contrast, individuals also
seek groups that are sufficiently small to permit them to feel distinctive. These motives are
fundamental but also in opposition, an assumption with a long history in psychology (e.g.,
Codol, 1984; Lemaine, 1974; Maslach, 1974; Snyder & Fromkin, 1980; Ziller, 1964):
individuals wish to meet both needs, seeking to stand out, but not too much.
Moreover, these motives are presumed to operate at different levels of self (Brewer,
1991; Brewer & Roccas, 2001). Not only do individuals seek to meet both needs, but they can
meet them with reference to different identities, not only their individual identity but also
increasingly inclusive nested identities (i.e., individual, subgroups, groups, superordinate groups,
etc.; Brewer, 1991; see also Hornsey & Jetten, 2004). Individuals can feel that their personal and
group identities can feel distinctive or inclusive. However, meeting a need with one level of self
Optimal Distinctiveness and Membership Trust 4
does not mean that the need is met at every level of self. Feeling individually similar to or
included with fellow group members (within-group inclusion) does not pre-suppose that the
group itself is also seen as similar to other groups (between-group inclusion). As a consequence,
even though the needs oppose each other at the same level of self, one may feel simultaneously
distinctive and included, with the needs met at different levels of the self. Herein lies the
experience of optimal distinctiveness; it can be achieved when individuals belong to groups that
are more distinctive than other groups but simultaneously yield a sense of inclusion within the
group. Achieving optimal distinctiveness – that is, by simultaneously meeting the needs via
between-group distinctiveness and within-group inclusion – is presumed to be predicted by
relative group size and to predict membership trust.
Individuals find groups to be optimally distinctive by considering the group’s size
relative to the surrounding population (Brewer, 1991): a group of 500 is large in a population of
600, but small in a population of 6000. The relative size of groups could thus be conceptualized
as a continuum and may be indexed by percentages, ranging from extremely small groups (such
as 1% or less of the population) to extremely large groups (99% or more) to any number in
between. Groups that are relatively small (hereafter referred to as “minorities”) have the potential
to be distinctive, and sufficiently small minorities are expected to be optimally distinctive.
Relative group size is expected to yield optimal distinctiveness based on how size affects
social categorization (Allport, 1954; Bruner, 1957), the cognitive process by which individuals
form perceptions of groups (Wilder, 1981). Social categorization refers to a process of
understanding individuals by knowing to which other people they are similar, and from which
other people they differ (adapted from McGarty, 1999). People form categories when they
observe meta-contrast (Campbell, 1958), that is when individuals or objects are perceived to be
Optimal Distinctiveness and Membership Trust 5
increasingly similar to some other individuals or objects, and these individuals or objects are
unlike others (for a recent review, see Leonardelli & Toh, 2015; see also Rosch, 1978; Turner,
1987; Tversky & Gati, 1978). Decreasing minority size accentuates category formation,
increasing between-group distinctiveness and within-group inclusion (Brewer, 1991). Evidence
supports this prediction: relative to numerical majority categories, moderately small minority
categories are more distinctive (McGuire & McGuire, 1988) and minority versus majority group
members are perceived to be more similar to each other (Nelson & Miller, 1995). In this regard,
optimally distinctive groups can simultaneously meet both needs, via between-group
distinctiveness and within-group inclusion.
According to Brewer (2003), between-group distinctiveness and within-group inclusion
produced by relative size will strengthen mutual expectations of trust and obligatory cooperation
within the category. The more distinctive size of the minority group reduces the risk of
nonreciprocation of cooperation; after all, holding all else constant, depending on fewer people is
expected to increase mutual accountability among existing group members. Moreover, greater
inclusion within-group extends trust expectations to all members, even those otherwise
unknown. In this regard, it is the shared features that lead individuals to trust others, even if the
group members are otherwise unknown to the individual. Thus, we elaborate on how perceptions
of minority group size, optimal distinctiveness, and membership trust came to be associated with
each other, and test such predictions in two studies.
We take a different approach from other research on group size and cooperation. In
contrast with Dunbar’s (1993) research connecting larger primate brains with absolute group size
and cooperation among familiar others, our work targets relative group size, social comparison
and categorization processes, and implications for trust and cooperation even for unfamiliar
Optimal Distinctiveness and Membership Trust 6
group members. Crucially, too, Dunbar’s work does not depend on and cannot explain the
connection between cooperation and perceptions of relative size and optimal distinctiveness,
upon which our theoretical perspective rests. In two experiments, we test for the first time
whether individuals associate minority size with membership trust, and whether such trust is a
function of optimal distinctiveness. Previous research has established that group members trust
fellow members more than non-members (e.g., Brewer & Kramer, 1986; Foddy et al., 2009;
Lount, 2010). Investigating the role of relative size in membership trust requires tests that
disentangle group size effects from those of shared group membership. Thus, our experiments
determine whether individuals associate more membership trust with sufficiently small minority
groups when they are members of neither group.
Experiment 1: Observer’s Perceptions
This study explored individuals’ perceptions of minorities and majorities when they
themselves belonged to neither group. We expected observers to see members of the minority
more than majority group as likely to trust and assist each other as a function of optimal
distinctiveness specifically when the minority was sufficiently small. This prediction was tested
by manipulating relative group size; the size distribution was more extreme for some participants
(20% v. 80%) than for others (45% v. 55%). Compared to a mere minority (operationalized at
45%), we expected that substantially smaller minorities (operationalized at 20%) would be
perceived as more optimally distinctive, and thus perceived to have higher levels of intragroup
trust relative to majority groups.
Method
Although the study was conducted prior to the developing conclusions the field of
psychology is reaching over how to address the determination of sample size, we sought to make
Optimal Distinctiveness and Membership Trust 7
the sample larger than what previous research has used to determine sample size in studies with
laboratory-created groups (cell size ranged from 20-35; e.g., Hogg & Turner, 1987). Thus, we
sought to recruit between 45-50 participants for each theoretically relevant cell of the design, a 2-
cell (minority relative size: 20%, 45%) between-participants design. Ninety-one individuals were
recruited for this study; 36 were initially recruited through the Rotman School of Management’s
Paid Participant pool; they completed the study in exchange for $3 CA. As the recruiting from
the paid pool was progressing slowly, in order to facilitate data collection, another 55 were
recruited from Amazon's Mechanical Turk (MTurk) website and were paid $1.00 US to
participate. Although MTurk participants reported higher trust ratings (M=5.16) than local
participants (M=4.25), F(1,87) = 20.20, p < .05, differences in recruitment source had no other
significant effects, main effects or moderation, on dependent measures. Consequently, this
difference is no longer addressed.
To reduce alternative explanations associated with some minority groups (e.g., group
history or socioeconomic status), participants were asked to report their perceptions regarding
two supposedly real (but laboratory-created) minimal groups (Tajfel et al., 1971), presumably
defined by preferences for paintings by Wassili Kandinsky or Paul Klee. Participants were also
randomly assigned to one of two relative group size conditions, reading that the painting
preference groups represented 45% and 55% of the population or 20% and 80% of the
population. Here is what the participants’ read:
In this study we are examining how accurately people perform group classifications. Previous
research has shown that simply based on facial features, individuals can at times accurately
classify other people by their group membership. This study will investigate how well
individuals can classify people by their painting preference group membership. For about three
decades, research (Brewer, 1979; Brewer & Pickett, 2005) has established that different painting
preference groups are associated with differences in cognitive psychology, social psychology,
and personality.
The Painting Preferences Classification Task identifies the painting preferences of an individual
and classifies them into one of two painting preference groups, either Kandinsky or Klee. The
Optimal Distinctiveness and Membership Trust 8
results of prior research using this task have shown that roughly [20%, 45%, 55%, or 80%] of
the population is classified into the [smaller, larger] Klee group, whereas [80%, 55%, 45%, or
20%] of the population is classified into the [larger, smaller] Kandinsky group.
One category was always a minority and the other a majority; which category was
smaller was counterbalanced across participants as a control factor. This control factor (minority
group: Kandinsky, Klee) did not yield significant or meaningful effects, and consequently is no
longer addressed.
To confirm they were aware of this size information, we asked participants to report the
size of one group relative to another. For participants to continue with the study, they were
required to correctly report the exact percentage of the two groups.
Participants then reported on the groups’ optimal distinctiveness using a 7-point scale
(1=Kandinsky group much more to 7=Klee group much more). The 4-item measure included
questions about between-group distinctiveness (i.e., “Which group do you feel is more
distinctive?”; “Which group ‘stands out’ more?”) and within-group inclusion (i.e., “In which
group are members more associated with each other?”; “In which group do members fit in more
with the others of their group?”). The item responses were averaged to index perceptions of
optimal distinctiveness (α=.71). Responses were recoded to range from -3 to +3, where positive
scores indicated greater minority optimal distinctiveness relative to the majority group. With this
recoding, zero reflects a point where neither group is rated higher than the other.1
Finally, participants completed a measure of intragroup trust, rating their agreement with
three statements about each group (i.e., “[Klee/Kandinsky] group members think other members
1 Although distinctiveness and inclusion should be negatively related at the same level of self (e.g., when both are
measured within-group), they can be positively correlated when distinctiveness is indexed between-group and
inclusion indexed within-in (Leonardelli et al., 2010). Consistent with this research, the two subscales are positively
related here: Experiment 1, r(89) = .20, p = .06; Experiment 2, r(91) = .30, p = .004. These data support the notion
that the four items should be combined into one index of optimal distinctiveness.
Optimal Distinctiveness and Membership Trust 9
of their group are trustworthy”; “[Klee/Kandinsky] group members trust other members of their
group”; “Members of the [Klee/Kandinsky] group are willing to assist each other”) using a 7-
point scale (1=Disagree Strongly to 7=Agree Strongly). These ratings were recoded to reflect
intragroup trust ratings for the minority or majority group. Internal consistency was high for both
sets of items (αmin = .92; αmaj = .94). Item scores were averaged to form separate scores for the
minority (M = 5.18, SD = 1.05) and majority groups (M = 4.80, SD = 1.16); higher numbers
indicated greater intragroup trust. For more information, the online supplemental materials
summarize exploratory factor analyses, from Experiments 1 and 2, which support the notion that
the distinctiveness, inclusion, and trust measures are indexing different constructs.
Results and Discussion
Two participants failed to accurately report the relative size of the two groups, but
because conclusions of analyses do not substantially differ when excluding them, reported
analyses include the manipulation failures. The counterbalance manipulation yielded no
meaningful or consistent effects here, and is thus excluded from analysis.
Trust perceptions. Scores were submitted to an analysis of variance, with relative group
size (20% v. 80%, 45% v. 55%) as a between-subject factor and target group (minority, majority)
as a within-subject factor. Analysis yielded a target group main effect, F(1, 89) = 7.15, p = .009,
r = .27, which was qualified by a relative group size X target group interaction, F(1, 89) = 5.04,
p = .03, r = .23. Observers expected intragroup trust to be higher among group members
belonging to the 20% minority (M = 5.36, SD = 1.00; Mean CI95% = 5.06, 5.67) than the 80%
majority (M = 4.68, SD = 1.23; Mean CI95% = 4.35, 5.02), F(1, 89) = 12.23, p = .001, r = .35. By
contrast, observers expected intragroup trust for members of the 45% minority (M = 4.98, SD =
1.08; Mean CI95% = 4.68, 5.29) to be similar to that of the 55% majority (M = 4.93, SD = 1.08,
Optimal Distinctiveness and Membership Trust 10
Mean CI95% = 4.58, 5.27), F(1, 89) = .09, p = .76, r < .01. Intragroup trust ratings of the 20%
minority were higher and significantly different than the ratings for every other group, as
indicated by the confidence intervals.
Optimal distinctiveness. Scores were submitted to a t-test, which revealed that the 20%
minority was perceived to be more optimally distinctive (M = .58, SD = 1.45, Mean CI95% = .16,
1.00) than the 45% minority (M = -.08, SD = 1.04, Mean CI95% = -.38, .23), t(81.73) = -2.47, p =
.015, rpb = .25. The confidence intervals also indicated that the 20% minority was perceived to be
more optimally distinctive than the 80% majority (as indexed by the confidence interval not
including zero), but that the 45% minority was no more optimally distinctive than the 55%
majority (confidence interval includes zero).
Mediation. We predicted that the greater perceived trust associated with the 20%
minority group was a function of the greater perceptions of optimal distinctiveness attributed to
the 20% than the 45% minority. To test this predicted mechanism, we conducted an analysis
using PROCESS (SPSS version; Hayes, 2016). PROCESS is an analysis procedure that tests for
different models of moderation and (for our purposes) mediation. The procedure follows a causal
steps framework (Baron & Kenny, 1986) using a series of (in our case, linear regression)
analyses in addition to indirect effects. In this case, we sought to test whether optimal
distinctiveness can explain the effect of relative size on minority membership trust (a single
mediator model; in PROCESS, Model 4).
We submitted minority trust ratings as the outcome measure, relative group size as the
causal variable, and optimal distinctiveness scores as the mediator. Majority trust scores were
included as a covariate on the minority trust ratings to establish that it is perceptions of minority
trust above and beyond that which would be predicted by perceptions of group trust in general
Optimal Distinctiveness and Membership Trust 11
(as indexed by majority trust scores) that would be best predicted by relative size and mediated
by optimal distinctiveness. Relative size was dummy coded (0 = 45% minority, 1 = 20%
minority).
Prior to conducting PROCESS, we submitted minority trust scores to a regression
analysis with relative size as a predictor and majority trust scores as a covariate. This analysis
revealed that the majority trust ratings were a significant covariate (b = .26, SE = .09, p = .006,
sr=.28), with higher trust from majorities yielding higher trust for minorities. However, even
controlling for majority trust scores, relative size significantly predicted minority trust (b = .44,
SE = .21, p = .04, sr=.21), revealing that observers perceived minority group members displaying
greater intragroup trust when their group was 20% rather than 45%. Albeit conducted in a
different way, these results are consistent with the mixed analysis of variance reported earlier.
These results are presented here to offer a point of comparison for the PROCESS model;
presumably, were optimal distinctiveness to mediate the effect of relative size on minority trust,
then the relative size effect on minority trust should be reduced or become nonsignificant.
The outcomes of the PROCESS analysis are reported in Figure 1. The results revealed
that smaller minority size was related to greater optimal distinctiveness, a significant effect (and
an analysis consistent with our previously reported t-test). An analysis of minority trust with
relative size and optimal distinctiveness included as predictors and majority trust as a covariate
revealed, as would be expected, the covariate continued to be a significant predictor of minority
trust (b = .28, SE = .09, p = .002). More importantly, and consistent with the prediction, optimal
distinctiveness positively predicted minority trust such that higher optimal distinctiveness scores
were associated with higher membership trust, but relative group size no longer predicted
minority trust (ns). A bootstrap test of the indirect effect through optimal distinctiveness did not
Optimal Distinctiveness and Membership Trust 12
include zero in the confidence interval, indicating evidence consistent with mediation (b = .12,
CI95% = 0.0012, .369; 10000 resamples). To conclude, the evidence is consistent with the
prediction that individuals associated trustworthiness more with 20% than 45% minorities for
reasons of optimal distinctiveness.
Distinctiveness and inclusion as separate contributors. According to theory (Brewer,
2003), it should be the combined optimal distinctiveness measure that most strongly predicts
minority trust rather than inclusion or distinctiveness scores alone. We explored this assumption
by calculating the correlations for minority trust with optimal distinctiveness scores, and
separately, with inclusion and distinctiveness scores. Positive correlations would indicate that the
more distinctive, included, or optimally distinctive the minority is perceived to be, the more
observers perceive minority intragroup trust. We also calculated the correlations of these
measures with majority trust, included here as a counterpoint. Here, negative correlations would
indicate that the more the majority is perceived to be distinctive, included, or optimally
distinctive, the more observers perceive majority trust (we did not expect optimal distinctiveness
to be related to majority intragroup trust). Table 1 summarizes these correlations in the first two
rows of the table.
Three features of the data are worth discussing. First, the combined optimal
distinctiveness measure is significantly related to greater minority trust, and although
nonsignificant, distinctiveness and inclusion individually are positively related to trust. Second,
the optimal distinctiveness correlation is bigger than either need satisfaction measure alone,
accounting for 1% more variance, consistent with the idea that the combined total is more
predictive of intragroup trust than either alone. Third, the needs do not correlate with majority
trust in the same way. If anything, the data seem to suggest that while greater majority inclusion
Optimal Distinctiveness and Membership Trust 13
relates to greater majority trust, distinctiveness does not, and neither did the combined optimal
distinctiveness index. The data are certainly consistent with our predictions. While correlations
between minority trust and the distinctiveness and inclusion were not significant, our theoretical
approach does not require significant effects of each; rather, it focuses on the overall concept of
optimal distinctiveness, which was significantly related to minority trust. The Supplemental
Online Materials include some additional mediation analyses exploring between-group
distinctiveness and within-group inclusion as independent mediators to membership trust, using
data from Experiments 1 and 2.
Experiment 2: Membership Preferences
Although in Experiment 1 individuals associated trustworthiness more with 20% rather
than 45% minorities for reasons of optimal distinctiveness, it was not clear whether this
association between trustworthiness and minority size was desirable and would manifest
behaviorally. Experiment 2 sought to test whether individuals would prefer and join a minority
or majority group before doing a task where they would depend on the trustworthiness of a
member of their selected group. Moreover, as in the first study, we experimentally manipulated
relative group size, where the distribution between minority and majority groups was more
extreme for some participants (20% v. 80%) than for others (45% v. 55%) and measured optimal
distinctiveness. As reported in Experiment 1, we expected that the 20% minority would be
perceived to be more optimally distinctive and its members more trustworthy compared to the
45% minority. Furthermore, we expected that individuals would prefer membership in and join
minorities more than majorities when the minorities were 20% versus 45% of the population as a
result of perceptions of optimal distinctiveness and intragroup trust. This model is presented in
Figure 2.
Optimal Distinctiveness and Membership Trust 14
Method
As in Experiment 1, we targeted a total sample size of 45-50 participants per theoretically
meaningful cell of our 2-cell (minority relative size: 45%, 20%) between-participant design. A
total of 99 participants were recruited from Amazon's Mechanical Turk website and paid $0.75
to participate in this study. Once recruited, participants were presented with a decision to join
one of two groups (Group A and B) – a numerical minority or majority – before participating in a
trust game with an anonymous member of their selected group.
In the trust game (Berg et al., 1995), a participant is given a sum of money and decides
how much to send to an anonymous recipient. The money sent is tripled and the recipient decides
how much (if anything) to return to the participant. More money sent by the participant
represents a greater degree of trust in the recipient. In our study, participants decided how to
distribute $10 between themselves and another, described only as a member of their selected
group. Participants first read about the trust game and their group decision (all participants
received these instructions):
For the task, first you will choose which group to join (Group A or Group B). Then, you will
decide on a quantity of dollars ranging from zero to $10 that you would like to send to another
person, who will be a randomly chosen member from the group you join. The amount that you
send will be multiplied by 3 and given to that group member (i.e., if you choose to send $10, the
group member will receive 3 x $10 = $30). The group member will then respond with an amount
that he/she wants to return to you.
Prior to making their group choice, participants read some additional details about the
trust game. Group size was manipulated as follows: “For the purposes of this study, these groups
have been labeled Group A, which represents 20% [45%] of the population, and Group B, which
represents 80% [55%] of the population.” Group size was counterbalanced with category label;
for about half the participants Group A was the smaller group. As counterbalance yielded no
Optimal Distinctiveness and Membership Trust 15
consistent effects, it is no longer discussed. Participants then had to report the size of each group
before continuing.
Participants then completed measures of power, trust, and membership preference with
single-item questions (i.e.,“Which group do you feel is more powerful?”;“Who would you be
more likely to trust: A Member of Group A or B?”;“Which group do you prefer more?”) as well
as the optimal distinctiveness measure from Experiment 1. Participants completed the measures
using the following response scale (1=Group A much more, 4=Neither Group more than the
other, to 7=Group B much more). The item responses were averaged to index perceptions of
optimal distinctiveness (α=.78). Responses on these measures were recoded to range from -3 to
+3, with positive scores indicating greater optimal distinctiveness, power, membership trust, and
membership preference of the minority than majority group. With this recoding, zero reflects a
point where neither group is rated higher than the other.
Finally, participants chose their group membership, and made decisions of how much of
$10 to send to the other party. Before completing any dependent measures, participants were told
that 10% of the trust games would be conducted using real money, and after making their trust
game decision, they were informed whether their game was selected to be played with real
money. Participants were then paid for their participation. For participants who were chosen to
play trust games with real money, we tripled the amount of the outcomes sent to the other party,
gave this new total to a new set of individuals recruited from MTurk, who then decided how
much to keep and how much to return. No experimental manipulations or other measures were
introduced with this second set of participants; they were merely paid according to their decision.
Within two weeks, we then contacted participants in the original study and paid them
accordingly (what they originally decided to keep plus whatever was returned to them).
Optimal Distinctiveness and Membership Trust 16
Results and Discussion
Manipulation checks revealed that all participants correctly reported percentages of the
minority and majority groups depending on their group size condition. Participants’ open-ended
comments about the study revealed that four were confused about the trust game instructions.
Also, one person did not complete the measures, and another was consistently an outlier within
condition as identified by box plot analysis, that is, more extreme than the Tukey fence on
inclusion (-3 < -2.5), trust (3 > 1), and membership preference (3 > 1). These six individuals
were dropped from analysis, resulting N = 93 (ages 18 to 68 years, Mdn=34; 52 females, 41
males). Table 2 contains descriptive statistics and bivariate correlations among the measures.
Group ratings and choice. Individuals perceived the majority to be more powerful than
the minority (M = -1.15, SD = 1.33, Mean CI95% = -1.43, -.87; confidence interval does not
include zero), but relative group size (20% v. 80%, 45% v. 55%) did not affect this perception, t
< 1, p > .54. Logistic regression analysis revealed a relative group size effect on group choice (b
= 1.02, p = .02, requivalent = .25). More participants joined the minority when its relative size was
20% (30/46 = 65%; CI95% = 51%, 77%) than 45% (19/47 = 40%; CI95% =28%, 55%). Likewise,
individuals preferred minority over majority group membership more when the minority was
20% (M = .52, SD = 1.49, Mean CI95% = .11, .94) than 45% (M = -.26, SD = 1.34, Mean CI95% = -
.67, .16), t(91) = -2.65, p = .01, rcontrast = .27.
Similarly, participants rated the 20% minority as more optimally distinctive (M = .86, SD
= 1.09, Mean CI95% = .56, 1.16) and its members as more trustworthy (M = .54, SD = 1.49, Mean
CI95% = .14, .95) than the 45% minority (optimal distinctiveness: M = -.23, SD = .95, Mean CI95%
= -.53, .06; trustworthiness: M = -.23, SD = 1.25, Mean CI95% = -.63, .16), ts < -2.70, ps < .009,
rscontrast > .28. Also, the 20% minority was perceived to be more optimally distinctive and its
Optimal Distinctiveness and Membership Trust 17
members more trustworthy than the 80% majority, as indicated by the confidence intervals not
including zero. There was no difference between the 45% minority and the 55% majority on
perceptions of optimal distinctiveness or trust (i.e., confidence intervals include zero).
Sequential mediation tests. We sought to replicate the mediation effects from
Experiment 1, where optimal distinctiveness mediated the effects of relative group size on
membership trust, and to extend this model one step further, by demonstrating that individuals
will prefer and choose membership in a sufficiently small minority for reasons of membership
trust. Such a prediction requires a sequential mediation model, with relative size predicting
optimal distinctiveness, which in turn predicts membership trust, which in turns predicts
membership choice (or preference; these models are depicted in Figure 2, upper and lower panels
respectively). We tested this model using PROCESS (Hayes, 2016, Model 6), first using
membership choice as the outcome measure, and conducting it again with membership
preference. We expected analyses with both outcomes would yield consistent effects.
Figure 2 summarizes the coefficients for each of the predicted paths for membership
choice (upper panel) and preference (lower panel). Consistent with predictions, the sequential
mediation path – namely that a smaller relative size would yield the inference that the minority
group was optimally distinctive, and thus its members more trustworthy, and in turn preferred or
chosen more – was supported (bolded line); all paths were in the predicted direction and
significant. The direct effect from relative size to membership choice or preference was no
longer significant and tests indicate that the indirect sequential mediation effect was significant
on choice (b = .57, CI95% = 0.19, 1.25; 10000 resamples) and membership preference (b = .55,
CI95% = 0.28, .97; 10000 resamples). None of the indirect effect tests through only one mediator
was significant on choice or preference, although evidence suggests the effect of relative size on
Optimal Distinctiveness and Membership Trust 18
membership preference may also be mediated by optimal distinctiveness alone. While the
indirect effect test for this path was not significant, future research would benefit from exploring
this potential separate effect with additional testing. Overall, as minority size decreased, the
minority was perceived to be optimally distinctive, which provoked greater membership trust,
which in turn was associated with minority membership preference and choice (Figure 2).
One possible alternative explanation might be that observers saw intragroup trust as a
necessary feature of a less attractive and weaker minority group; perhaps members of such
groups have nothing else to depend on, so they depend on each other. This explanation is
unlikely and the data in this study help to reject it. If this alternative were supported, we should
have found a negative correlation between trust and power ratings; such a correlation would
indicate that the more power the majority has, the more observers believe that members of the
weaker minority group trust each other. However, we find no such correlation (r = .07; Table 2).
Analyzing this correlation within experimental conditions also reveals no correlation between
power and trust ratings in the 45% minority condition (r = .15) or in the 20% minority condition
(r = .006). Among the data reported in this paper, there is no evidence to support this alternative
explanation. Trustworthiness thus appears to be a valued attribute of group membership (Cottrell
et al., 2007; Leach et al., 2007).
Trust game decision. Analysis of the trust game decision revealed that those who joined
the minority gave marginally more (M = 8.52, SD = 2.90) than those who joined the majority (M
= 7.36, SD = 3.21), t(91) = -1.94, p = .06, r = .20. Analyzing only participants in the 20%
minority condition produced the same pattern; participants gave significantly more money when
they joined the minority (M = 8.17, SD = 3.25) than when they joined the majority (M = 6.00, SD
= 2.83), t(43) = -2.24, p = .03, r = .20. By contrast, for those in the 45% minority condition, this
Optimal Distinctiveness and Membership Trust 19
difference did not reach significance, t = -1.21, p = .22. On the surface, these data appear to
support our basic notion that group members will exhibit higher levels of trust and cooperation
when they are members of a sufficiently small minority rather than majority group members.
Interestingly, in this paradigm, where individuals selected their group membership before
engaging in the decision to trust the other party, we may have expected to see no difference in
trust game decisions. After all, those who chose to be in the majority would have likely
optimized their investment decision given their preferred group membership. It would be helpful
to test whether the trust decision effect we find here in this paradigm replicates.
Demographic characteristics. Of note, age and income were not related to any of the
dependent measures (ps > .15). Gender yielded a marginally significant effect on membership
trust, t(91) = -1.96, p = .052, where women were more likely to trust minority group members
(M = .40, SD = 1.32) than men were (M = -.17, SD = 1.50). Gender also yielded a marginally
significant effect on group power, t(91) = -1.71, p = .08, where men considered majorities more
powerful (M = -1.41, SD = 1.14) than women did (M = -.94, SD = 1.45). However, gender did
not moderate any of the effects yielded by the minority size manipulation (ps > .33).
Distinctiveness and inclusion as separate contributors to trust. The combined optimal
distinctiveness measure should more strongly predict membership trust than inclusion or
distinctiveness scores alone. We explored this assumption by calculating the correlations
between membership trust and optimal distinctiveness, inclusion and distinctiveness scores.
Positive correlations would indicate that the more distinctive or included the minority is
perceived to be, the more participants believed members of the minority group are trustworthy.
Table 1 summarizes these correlations in the bottom row.
Optimal Distinctiveness and Membership Trust 20
Consistent with Experiment 1, distinctiveness and inclusion were individually and
positively related to trust (and in this study, the correlations were significant) as was the optimal
distinctiveness measure, although the optimal distinctiveness correlation was bigger than either
distinctiveness or inclusion alone, accounting for 11-13% more variance, consistent with the idea
that the combined total is more predictive of intragroup trust than either alone. The data are
consistent with our predictions, and the effects overall are larger here than they were in
Experiment 1, consistent with the notion that the paradigm used in Experiment 2 was more
personally relevant, which in turn made the inferences about trust larger.
General Discussion
This paper is the first to empirically demonstrate the connections between relative size
and intragroup trust as mediated by optimal distinctiveness. Moreover, this is the first paper to
show, to our knowledge, a membership preference for a 20% minority group in spite of seeing it
as less powerful than the numerical majority (Experiment 2). Such data – collected with highly
controlled procedures and experimental manipulations – allow for causal inference and rule out
additional explanations associated with self-justification (through employing participants as
observers rather than group members), brain size (through manipulating relative rather than
absolute size) and status (through using minimal groups). These effects were replicated with
different samples, procedures, and measures of trust, adding confidence to the findings. Overall,
the evidence supports the conclusion that sufficiently small minority group size is associated
with optimal distinctiveness, membership trust, and membership preference. Although much of
this program was under way prior to psychology’s replicability crisis, we find it encouraging
that, given our sample sizes (N = 89, Study 1; N = 93, Study 2) and effect size estimates (.23 < rs
< .35), our tests had acceptable observed power ranging from .72 to above .90.
Optimal Distinctiveness and Membership Trust 21
The Function of Relative Size and Optimal Distinctiveness
The evidence reported within this paper supports, more generally, the notion that when it
comes to decisions related to trust, cooperation and group effectiveness, a sufficiently small
group size is an important determinant. Other literature has identified benefits to optimally sized
groups, whether it be more effective coordination (Hackman, 1987) or evolutionary explanations
for cooperation rooted in genetics (West et al., 2011) or ecology (Boyd & Richerson, 1988).
What our conclusions indicate, however, is that individuals are aware of such associations and
they use them to guide their decision-making about groups. Individuals’ attention to a group’s
relative size and degree of optimal distinctiveness suggests that, when motivated to do so,
individuals seek to balance tradeoffs between group capability and sustainability. Individuals
care about the cooperative success of their communities, and one way they can safeguard it is by
ensuring their communities and groups are perceived by them to be optimally distinctive.
Some evidence supports the idea that individuals will actively regulate the size of their
groups. Take the Hutterite community (as described in Dunbar, 1993; Hardin, 1988); these
communities typically swell to a particular group size before steps are taken to divide into two
smaller communities. Actively managing group size, we argue, is a less aggravating way of
adjusting group size when resources run thin relative to means such as starvation, pestilence, or
war (Malthus, 1826). Fissioning (as it is called when groups subdivide; Hart & van Vugt, 2006)
can be considered a less aggravating circumstance wherein individuals or their group members
could use relative size and perceived optimal distinctiveness to regulate group memberships.
Given increased individual migration patterns later in life, say through the departure from
home as individuals enter university or find work, individuals may find such associations helpful
as they seek membership in communities that are self-sustaining. In this instance, we can see
Optimal Distinctiveness and Membership Trust 22
parallels between a personal life experience and the paradigm used in Experiment 2. As an
individual starts life away from parents or guardians and wishes to build a community for
himself or herself, to which groups does he or she turn? Were the individual seeking a basis for
trust and cooperation, holding all else constant, a group perceived as sufficiently small – one that
is perceived to be optimally distinctive – would be one way to do so. We see such parallels
between the paradigms used here and some instances of real-life events.
That noted, we expect that existing group memberships – particularly those to which
individuals were born or assigned – played an important role in sharing individuals’ perceptions
on the relatedness of relative group size, optimal distinctiveness and membership trust.
Evolutionary explanations for cooperation, whether rooted in genetics (W.D. Hamilton, 1964),
ecology (Boyd & Richerson, 1988), or some combination of the two, propose that cooperation
will be sustained with sufficiently small groups. Such theories give rise to the possibility that the
structural realities of group life would have reinforced perceived associations people formed
between relative size, optimal distinctiveness, and trust throughout an individual’s extended
period of dependence by repeated assembly (Caporael, 1997) into groups that are part of larger
populations (that would have allowed comparisons of relative size). With such a linkage
established between optimal distinctiveness and trust, individuals are expected to cooperate with
and trust others who they see as members of their optimally distinctive group, even if the group
is vast (but still a minority) and far from their place of origin.
An additional feature goes beyond the scope of this article but would benefit from
additional investigation. Limited dispersal, defined as reduced movement and mixing of human
groups from natal and breeding sites (for a recent review, see Ronce, 2007), is considered an
important boundary condition for distal genetic and ecological theories linking group size and
Optimal Distinctiveness and Membership Trust 23
cooperation. Smaller group size is related to greater intragroup cooperation under limited
dispersal. It would be interesting to determine whether the perceived connections among relative
size, optimal distinctiveness, and intragroup trust depend on the degree of limited dispersal
experienced by individuals, perhaps the degree of dispersal during their developmental history.
The Conditions for Membership Trust
A powerful application that this theory and evidence offers is that leaders of groups (even
large ones) may facilitate more cooperative interdependence among their members if they frame
the group as being sufficiently small percentage of a reference population. It speaks to the
potential value that shaping relative size perceptions of the teams, units, and the institution or
organization as a whole may have on members of the industry or government. Importantly, as
this work demonstrates, it is not enough to be a numerical minority, but rather, a sufficiently
small minority, one that would conjure perceptions of optimal distinctiveness. We showed that
participants found 20% groups as more optimally distinctive than 45% groups, but it would be
helpful to explore whether there is a more precise estimate (or range of estimates) that would be
considered optimally distinctive.
Part of what may inform the conditions of membership trust, at least as such perceptions
result from the interplay of distinctiveness and inclusion, is to consider how features of these
motives co-exist and mutually sustain and constrain each other (Brewer, 1991), particularly the
interplay of the motives at the same levels of analysis (in opposition) and across levels of self
(complementary). Taking these co-existing features of the motives one step further, meeting a
need at one level of self (at intragroup or group) is not expected to completely satiate the need at
another level of self. For example, individuals may belong to and identify with an optimally
distinctive group membership – with between-group distinctiveness and within-group inclusion –
Optimal Distinctiveness and Membership Trust 24
and also choose to identify themselves with optimally distinctive subgroups within that group
(what has been called “dual identity”; for supporting evidence, see Leonardelli, Pickett, Joseph,
& Hess, 2011). In this regard, individuals may achieve optimal distinctiveness at multiple levels
of identification, and relatedly, membership trust at multiple levels of identification.
That such needs mutually constrain each other also may help address whether
membership trust is likely to manifest with really exclusive minority groups (e.g., 1%). Greater
group distinctiveness is presumed to reduce the risk of noncooperation from others and within-
group inclusion is expected to extend expectations of trust to all group members. By this
rationale, extremely small groups are also likely to be highly distinctive between-group and
highly included within. Groups that are too small are believed to be insufficient to expand
capabilities. Moreover, such a prediction neglects the assumption that the needs are in opposition
at the same level of analysis; holding all else constant, complementarity (between-group
distinctiveness and within-group inclusion) is expected to be an insufficient substitute for,
respectively, individual distinctiveness and between-group inclusion.
Consistent with this prediction, evidence has indicated that individuals are less likely to
identify with groups that are too small (Abrams, 2009, Study 1; Bearman & Bruckner, 2001;
Lau, 1989) especially when individuals are motivated to be included in their group (Leonardelli
et al., 2011). Future research would benefit from determining whether optimal distinctiveness
effects resulting from trust will occur with groups that are extremely small, and whether group
identification may indirectly affect trust effects by reducing individuals’ tendency to identify
with extremely small groups.
Conclusion
Optimal Distinctiveness and Membership Trust 25
Our findings help to substantiate why individuals in optimally distinctive groups –
whether defined by race, music preferences, location or minimal classification – exhibit greater
identification with or support on behalf of their group relative to those in numerical majorities
(for a review, Leonardelli et al., 2010; e.g., Abrams, 2009; Badea et al., 2010; Bernardo &
Palma-Oliveira, 2012; Blanz et al., 1995; Brewer et al., 1993; Ellemers & Van Rijswijk, 1997;
Ellemers et al., 1999; Feather, 1995; Hornsey & Hogg, 1999; Lau, 1989; Leonardelli & Brewer,
2001; Leonardelli et al., 2011; Picket, Bonner, & Coleman, 2003; Lucken & Simon, 2005; Simon
& Brown, 1987; Simon & Hamilton, 1994, Experiment 1; Vignoles & Moncaster, 2007).
Specifically, these individuals are supporting group memberships that are perceived to be
optimally suited for sustained group cooperation, such that they feel obliged to cooperate on
behalf of group members beyond those known through personal connections (Leonardelli et al.,
2010). Future research would benefit from exploring at what point (if any) the burden of
cooperation outweighs the benefits of shared trust expectations.
Optimal Distinctiveness and Membership Trust 26
References
Abrams, D. (2009). Social identity on a national scale: Optimal distinctiveness and young
people's self-expression through musical preference. Group Processes & Intergroup
Relations, 12, 303-317. doi:10.1177/1368430209102841
Allport, G. W. (1954). The historical background of modern social psychology. In G. Lindzey
(Ed.), Handbook of Social Psychology (Vol. 1, pp. 3–56). Cambridge, MA: Addison-
Wesley Publishing Company.
Badea, C., Jetten, J., Czukor, G., & Askevis-Leherpeux, F. (2010). The bases of identification:
When optimal distinctiveness needs face social identity threat. British Journal of Social
Psychology, 49, 21-41. doi:10.1348/000712608x397665
Baron, R. M., & Kenny, D. A. (1986). The moderator mediator variable distinction in social
psychological-research - conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182. doi:10.1037/0022-3514.51.6.1173
Bearman, P. S., & Brückner, H. (2001). Promising the Future: Virginity Pledges and First
Intercourse. American Journal of Sociology, 106(4), 859-912. doi:10.1086/320295
Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social-history. Games and
Economic Behavior, 10, 122-142. doi:10.1006/game.1995.1027
Bernardo, F., & Palma-Oliveira, J. M. (2012). Place identity: a central concept in understanding
intergroup relationships in the urban context. The Role of Place Identity in the Perception
Understanding, and Design of Built Environments (pp. 45-62). Bentham.
Blanz, M., Mummendey, A., & Otten, S. (1995). Positive–negative asymmetry in social
discrimination: The impact of stimulus valence and size and status differentials on
Optimal Distinctiveness and Membership Trust 27
intergroup evaluations. British Journal of Social Psychology, 34(4), 409-419. doi:
10.1111/j.2044-8309.1995.tb01074.x
Boyd, R., & Richerson, P. J. (1988). The evolution of reciprocity in sizable groups. Journal of
Theoretical Biology, 132, 337-356. doi:10.1016/s0022-5193(88)80219-4
Brewer, M. B., & Kramer, R. M. (1986). Choice behavior in social dilemmas – effects of social
identity, group-size, and decision framing. Journal of Personality and Social Psychology,
50, 543-549. doi:10.1037/0022-3514.50.3.543
Brewer, M. B., Manzi, J. M., & Shaw, J. S. (1993). In-group identification as a function of
depersonalization, distinctiveness, and status. Psychological Science, 4, 88-92.
doi:10.1111/j.1467-9280.1993.tb00466.x
Brewer, M. B. (1991). The social self: On being the same and different at the same
time. Personality and social psychology bulletin, 17(5), 475-482.
doi:10.1177/0146167291175001
Brewer, M. B. (2003). Optimal distinctiveness, social identity, and the self. In M. R. Leary & J.
P. Tangney (Eds.), Handbook of self and identity (pp. 480-491). New York, NY, US:
Guilford Press.
Brewer, M. B., & Roccas, S. (2001). Individual values, social identity, and optimal
distinctiveness. Psychology Press.
Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123-152. doi:
10.1037/h0043805
Campbell, D. T. (1958). Common fate, similarity, and other indices of the status of aggregates of
persons as social entities. Behavioral Science, 3(1), 14-25. DOI: 10.1002/bs.3830030103
Optimal Distinctiveness and Membership Trust 28
Caporael, L. R. (1997). The evolution of truly social cognition: The core configurations model.
Personality and Social Psychology Review, 1, 276-298.
doi:10.1207/s15327957pspr0104_1
Codol, J-P. (1984). Social differentiation and nondifferentiation. In H. Tajfel (Ed.), The social
dimension (Vol. 1, pp.314 – 337). Cambridge: Cambridge University Press.
Cottrell, C. A., Neuberg, S. L., & Li, N. P. (2007). What do people desire in others? A
sociofunctional perspective on the importance of different valued characteristics. Journal
of personality and social psychology, 92(2), 208. http://dx.doi.org/10.1037/0022-
3514.92.2.208
Dunbar, R. I. (1993). Coevolution of neocortical size, group size and language in
humans. Behavioral and brain sciences, 16(4), 681-693.
Ellemers, N., & Van Rijswijk, W. (1997). Identity needs versus social opportunities: The use of
group-level and individual-level identity management strategies. Social Psychology
Quarterly, 60(1), 52–65. doi: 10.2307/2787011
Ellemers, N., Kortekaas, P., & Ouwerkerk, J. W. (1999). Self-categorisation, commitment to the
group and group self-esteem as related but distinct aspects of social identity. European
Journal of Social Psychology, 29(2–3), 371–389. doi: 10.1002/(SICI)1099-
0992(199903/05)29:2/3<371::AID-EJSP932>3.0.CO;2-U
Feather, N. T. (1995). National identification and ingroup bias in majority and minority groups:
A field study. Australian Journal of Psychology, 47(3), 129–136.
doi:10.1080/00049539508257513
Optimal Distinctiveness and Membership Trust 29
Foddy, M., Platow, M. J., & Yamagishi, T. (2009). Group-based trust in strangers: The role of
stereotypes and expectations. Psychological Science, 20, 419-422. doi:10.1111/j.1467-
9280.2009.02312.
Hackman, J. R. (1987). The design of work teams. In J. W. Lorsch (Ed.), Handbook of
organizational behavior (pp. 315-342). Englewood Cliffs, NJ: Prentice-Hall.
Hamilton, W. D. (1964). Genetical evolution of social behavior. Journal of Theoretical Biology,
7(1), 1-52. doi:10.1016/0022-5193(64)90038-4
Hardin, G. (1968). Tragedy of the commons. Science, 162, 1243-1248.
DOI: 10.1126/science.162.3859.1243
Hardin, G. (1988) Common failing. New Scientist, 120(1635), 76.
Hart, C. M., & Van Vugt, M. (2006). From fault line to group fission: Understanding
membership changes in small groups. Personality and Social Psychology Bulletin, 32(3),
392-404. doi:10.1177/0146167205282149
Hayes, A. F. (2014). PROCESS (Version 2.13) [Software]. Available from
http://www.processmacro.org/download.html.
Hogg, M. A., & Turner, J. C. (1987). Intergroup behaviour, self‐stereotyping and the salience of
social categories. British Journal of Social Psychology, 26(4), 325-340. doi:
10.1111/j.2044-8309.1987.tb00795.x
Hornsey, M. J., & Hogg, M. A. (1999). Subgroup differentiation as a response to an
overly‐inclusive group: a test of optimal distinctiveness theory. European Journal of
Social Psychology, 29(4), 543-550.
Optimal Distinctiveness and Membership Trust 30
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(3),
248-264. doi: 10.1207/s15327957pspr0803_2
Lau, R. R. (1989). Individual and contextual influences on group identification. Social
Psychology Quarterly, 52, 220-231. doi:10.2307/2786717
Leach, C. W., Ellemers, N., & Barreto, M. (2007). Group virtue: The importance of morality (vs.
competence and sociability) in the positive evaluation of in-groups. Journal of
Personality and Social Psychology, 93, 234-249. doi:10.1037/0022-3514.93.2.234
Lemaine, G. (1974). Social differentiation and social originality. European Journal of Social
Psychology, 4(1), 17-52. doi: 10.1002/ejsp.2420040103
Leonardelli, G. J., & Brewer, M. B. (2001). Minority and majority discrimination: When and
why. Journal of Experimental Social Psychology, 37, 468-485.
doi:10.1006/jesp.2001.1475
Leonardelli, G.J., Pickett, C.L., & Brewer, M.B. (2010). Optimal distinctiveness theory: A
framework for social identity, social cognition and intergroup relations. In M. Zanna & J.
Olson (Eds.), Advances in experimental social psychology (Vol. 43, pp. 65-115). New
York: Elsevier.
Leonardelli, G. J., Pickett, C.L., Joseph, J.E., & Hess, Y.D. (2011). Optimal distinctiveness
theory in nested categorization contexts: Moving from dueling identities to a dual
identity. In R.M. Kramer, G.J. Leonardelli, & R.W. Livingston (Eds.), Social Cognition,
Social Identity, and Intergroup Relations: A Festschrift in Honor of Marilynn Brewer
(pp. 103-125). Psychology Press Festschrift series. New York, NY: Taylor & Francis.
Optimal Distinctiveness and Membership Trust 31
Leonardelli, G. J., & Brewer, M. B. (2001). Minority and majority discrimination: When and
why. Journal of Experimental Social Psychology, 37(6), 468–485. doi:
10.1006/jesp.2001.1475
Leonardelli, G. J., & Toh, S. M. (2015). Social categorization in intergroup contexts: Three kinds
of self‐categorization. Social and Personality Psychology Compass, 9(2), 69-87. doi:
10.1111/spc3.12150
Lount, R. B., Jr. (2010). The impact of positive mood on trust in interpersonal and intergroup
interactions. Journal of Personality and Social Psychology, 98, 420-433.
doi:10.1037/a0017344
Lücken, M., & Simon, B. (2005). Cognitive and affective experiences of minority and majority
members: The role of group size, status, and power. Journal of Experimental Social
Psychology, 41(4), 396-413. doi: 10.1016/j.jesp.2004.08.006
Malthus, T. R. (1826). An Essay on the Principle of Population (6th Ed.). London, UK: John
Murray.
Maslach, C. (1974). Social and personal bases of individuation. Journal of Personality and
Social Psychology, 29(3), 411. doi: 10.1037/h0036031
McGarty, C. (1999). Categorization in social psychology. London: Sage.
McGuire, W. J., & McGuire, C. V. (1988). Content and process in the experience of the self. In
L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 97-144).
San Diego, CA: Academic Press.
Nelson, L. J., & Miller, D. T. (1995). The distinctiveness effect in social categorization – you are
what makes you unusual. Psychological Science, 6, 246-249. doi:10.1111/j.1467-
9280.1995.tb00600.x
Optimal Distinctiveness and Membership Trust 32
Pickett, C. L., Bonner, B. L., & Coleman, J. M. (2002). Motivated self-stereotyping: heightened
assimilation and differentiation needs result in increased levels of positive and negative
self-stereotyping. Journal of personality and social psychology, 82(4), 543-562.
http://dx.doi.org/10.1037/0022-3514.82.4.543
Ronce, O. (2007). How does it feel to be like a rolling stone? Ten questions about dispersal
evolution. Annual Review of Ecology, Evolution, and Systematics, 231-253. doi:
10.1146/annurev.ecolsys.38.091206.095611
Rosch, E. (1978). Principles of categorization. In E. Rosch & B.B. Lloyd (Eds), Cognition and
categorization (pp.27-48). Hillsdale, NJ: Lawrence Erlbaum Associates.
Simon, B., & Brown, R. (1987). Perceived intragroup homogeneity in minority-majority
contexts. Journal of Personality and Social Psychology, 53(4), 703–711. doi:
10.1037/0022-3514.53.4.703
Simon, B., & Hamilton, D. L. (1994). Self-stereotyping and social context: The effects of
relative in-group size and in-group status. Journal of Personality and Social Psychology,
66(4), 699–711. doi: 10.1037/0022-3514.66.4.699
Snyder, C. R., & Fromkin, H. L. (1980). Uniqueness: The human pursuit of difference. New
York: Plenum.
Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and
intergroup behaviour. European Journal of Social Psychology, 1, 149-178. doi:
10.1002/ejsp.2420010202
Turner, J.C. (1987). A self-categorization theory. In Turner, J. C., Hogg, M. A., Oakes, P. J.,
Reicher, S. D., & Wetherell, M. S. (Eds). Rediscovering the social group: A self-
categorization theory. Basil Blackwell.
Optimal Distinctiveness and Membership Trust 33
Tversky, A., & Gati, I. (1978). Studies of similarity. In E. Rosch & B.B. Lloyd (Eds), Cognition
and categorization (pp.79-98). Hillsdale, NJ: Lawrence Erlbaum Associates.
Vignoles, V. L., & Moncaster, N. J. (2007). Identity motives and in‐group favouritism: A new
approach to individual differences in intergroup discrimination. British Journal of Social
Psychology, 46(1), 91-113. DOI:10.1348/014466605X85951
West, S. A., El Mouden, C., & Gardner, A. (2011). Sixteen common misconceptions about the
evolution of cooperation in humans. Evolution and Human Behavior, 32, 231-262.
doi:10.1016/j.evolhumbehav.2010.08.001
Wilder, D. A. (1981). Perceiving persons as a group: Categorization and intergroup relations. In
D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp.
213-257). Hillsdale, NJ: Erlbaum.
Ziller, R. C. (1964). Individuation and socialization: A theory of assimilation in large
organizations. Human Relations 17(4), 341-360. doi:10.1177/001872676401700403=
Optimal Distinctiveness and Membership Trust 34
Table 1
Bivariate Correlations between Trust and Distinctiveness, Inclusion, and Optimal Distinctiveness
Distinctiveness Inclusion Optimal
distinctiveness
Minority Trust (Exp. 1, N=93) .15
(p = .14)
.20m
(p = .051)
.23*
(p = .027)
Majority Trust (Exp. 1, N=93) .08
(p = .44)
-.28*
(p = .007)
-.13
(p = .23)
Trust rating (Exp. 2, N=91; higher
scores, more trust to minority)
.47*
(p < .001)
.44*
(p < .001)
.58*
(p < .001)
Note. For minority trust ratings (Exp. 1) and trust ratings (Exp. 2), positive correlations indicate
that the more the minority is perceived to be distinctive, included, or optimally distinctive, the
more members of this group were expected to trust each other (Exp. 1) or appeared trustworthy
(Exp. 2). For majority trust ratings, one would expect negative correlations, which would
indicate that the more the majority is perceived to be distinctive, included, or optimally
distinctive, the more its members are perceived to be trustworthy. Overall, the data indicate that
the combined measure of optimal distinctiveness has a stronger effect (as noted by a larger r-
value) than distinctiveness or inclusion alone.
Optimal Distinctiveness and Membership Trust 35
Table 2
Descriptive Statistics and Bivariate Correlations: Experiment 2
M SD Range 1 2 3 4
1. Optimal
distinctiveness
.31 1.16 -2.50 to +3 1.0
2. Membership trust .15 1.42 -3 to +3 .58* 1.0
3. Group power -1.15 1.33 -3 to +3 .15 .07 1.0
4. Membership
preference
.13 1.46 -3 to +3 .57* .77* .17 1.0
Note. N=93. *p<.05.
Optimal Distinctiveness and Membership Trust 36
Figure 1. Relative Size affects Minority Trust via Optimal Distinctiveness: Experiment 1
Note. Unstandardized regression coefficients reported. Bolded lines indicate a path through
mediator: optimal distinctiveness and minority trust. Relative size no longer significantly
predicted minority trust (dashed lines) when simultaneously predicted by optimal distinctiveness.
Analyses involving minority trust included majority trust as a covariate * p < .05; ns p > .10.
Optimal Distinctiveness and Membership Trust 37
Figure 2. Relative Size affects Membership Choice and Preference via Optimal Distinctiveness
and Membership Trust: Experiment 2
Note. Analyses conducted with membership choice and preference are in the upper and lower
panels respectively. Unstandardized regression coefficients reported. Bolded lines indicate the
predicted path through optimal distinctiveness and member trust. Other paths were no longer
significant or reduced in strength (dashed lines). * p<.05; ns p>.34.
Optimal Distinctiveness and Membership Trust 38
Supplemental Online Materials
Some reviewers wondered whether inclusion and trust could be treated as one and the
same. Conceptually, they are quite different. Inclusion refers to being immersed in, similar to,
and accepted by other group members, whereas trust refers to the intention to depend on another
to cooperate on your behalf (Rousseau et al., 1998). To explore this further, we conducted
exploratory factor analyses (maximum likelihood estimation and oblique rotation; specifically, in
SPSS, we used Direct oblimin rotation with parameter set to 0) in both studies to determine
whether, as we measured them, they could also be considered empirically distinct.
Also of interest, we explored whether distinctiveness or inclusion could individually
explain the connection between the relative group size manipulation and membership trust.
Within each study, we conducted a series of analyses using PROCESS (Hayes, 2016) to explore
these possibilities. It is important to note, however, that these alternative models are different in
an important way to the one reported in the paper. In the paper, ratings of both motives are
combined together into one, and thus, optimal distinctiveness scores comprised the variance in
each motive as well as the shared variance (the positive relation between the two). By contrast,
the models we entertain below either exclude the shared variance from the model (in what is
called the “separate” motives mediation model) or the covariance is assigned a causal role (in the
“sequential” motives mediation model). Ultimately, the role that this covariance plays in theory
will be the basis for comparing these models.
Study 1: Additional Analyses
Exploratory Factor Analysis
We submitted to the analysis the two items measuring inclusion, the two measuring
distinctiveness, and the three measuring minority intragroup trust (attempts to include majority
Optimal Distinctiveness and Membership Trust 39
trust items prevented the generation of a pattern matrix, as no local minimum could be found,
even when iterations were set to the maximum). The scores submitted to analysis were those
transformed so that higher numbers indicated greater trust, inclusion, or distinctiveness of the
minority. This analysis produced a pattern matrix which yielded a three-factor simple structure,
where each predicted set of items loaded strongly on separate factors, but weakly on the other
factors (see Table S1). Also, as a result of using oblique rotation, the analysis estimated
correlations among the three factors, all of which were small, but positive (Table S2).
Table S1: Factor Loadings for Distinctiveness, Inclusion, and Minority Trust Ratings (Study 1)
Pattern Matrixa
Factor
1
Distinctiveness
2
Minority
Trust
3
Inclusion
Distinctiveness - Which group ‘stands out’ more? 1.021 -.038 -.136
Distinctiveness - Which group do you feel is more
distinctive?
.697 .041 .139
Minority Trust - [Klee/Kandinsky] group members
trust other members of their group
.010 .982 -.093
Minority Trust - [Klee/Kandinsky] group members
think other members of their group are trustworthy
-.032 .931 -.018
Minority Trust - Members of the [Klee/Kandinsky]
group are willing to assist each other
.023 .760 .096
Inclusion - In which group are members more
associated with each other?
-.005 -.031 .949
Inclusion - In which group do members fit in more with
the others of their group?
.013 .031 .933
Extraction Method: Maximum Likelihood.
Rotation Method: Oblimin with Kaiser Normalization.
a. Rotation converged in 3 iterations.
b. Loadings less than .10 are dropped from presentation
Note. Responses were collected in reference to Klee and Kandinsky groups, but scores submitted to this analysis
were those transformed so that higher numbers indicated greater trust, inclusion, or distinctiveness to the minority.
Bolded loadings indicate the factor upon which the item most strongly loaded. All items loaded strongly on a single
factor and weakly on the other two, consistent with a simple factor structure.
Optimal Distinctiveness and Membership Trust 40
Table S2: Factor Correlations for Distinctiveness, Inclusion, and Minority Trust (Study 1)
Factor Correlation Matrix
Factor Distinctiveness Minority Trust Inclusion
Distinctiveness 1.000 .124 .199
Minority Trust .124 1.000 .189
Inclusion .199 .189 1.000
Extraction Method: Maximum Likelihood.
Rotation Method: Oblimin with Kaiser Normalization.
Distinctiveness and Inclusion as Predictors of Trust
As we considered how the two needs could be related to trust, we realized there were two
possible ways that the individual motives of between-group distinctiveness and within-group
inclusion could mediate membership trust. On the one hand, it is possible that one motive, the
other, or both could be separate and independent predictors of membership trust. On the other
hand, we also thought it possible that a sequential motive mediation model might be possible,
where relative size might be more likely to yield satisfaction of one motive, which in turn could
meet the other, which in turn could be more closely related to membership trust. We explored
both such models in both studies.
Separate motives mediation test. Given the data reported in Table 1, it is possible that
in Study 1, inclusion would be a better mediator than distinctiveness between size and
membership trust. To more thoroughly explore the separate contributions of distinctiveness and
inclusion, we conducted the same PROCESS analysis (Hayes, 2016, Model 4) as reported in the
paper, but with distinctiveness and inclusion as separate and simultaneous mediators. The results
of individual paths are summarized in Figure S1.
Optimal Distinctiveness and Membership Trust 41
Figure S1. Separate Motives Mediation Model: Experiment 1
Note. Unstandardized regression coefficients reported. Neither indirect path, through distinctiveness or inclusion
alone, could explain size on trust. Analyses involving minority trust included majority trust as a covariate * p < .05;
ns p > .10.
Of interest, relative size significantly predicted distinctiveness (i.e., smaller minorities are
more distinctive), but did not directly predict inclusion. Also of interest, whereas inclusion
positively predicted membership trust (the more included minorities were perceived to be, the
higher membership trust), distinctiveness did not significantly predict membership trust. Neither
indirect effect was significant, whether through distinctiveness (b = .0176, CI95% = -0.0861,
.2011; 10000 resamples) or inclusion (b = .0879, CI95% = -0.0099, .2751; 10000 resamples).
Sequential motives mediation test. The data presented from the simultaneous analysis
suggest that, perhaps, a sequential analysis could explain how distinctiveness and inclusion
individually contribute to membership trust, with size predicting distinctiveness, which predicts
within-group inclusion, which in turns predicts trust. Such a test could be conducted with
PROCESS using Model 6 (Hayes, 2016), and was conducted with majority trust scores as a
covariate. The results of individual paths are summarized in Figure S2.
Optimal Distinctiveness and Membership Trust 42
Figure S2. Sequential Motives Mediation Model: Experiment 1
Note. Numbers reflect unstandardized regression coefficients. Bolded lines indicate an indirect path through two
sequential mediators (with one linkage marginally significant). Other paths were no longer significant (dashed
lines) once included in this analysis. * p < .05; m p = .0519; ns p > .10.
The evidence indicates an indirect sequential path through distinctiveness and then
inclusion could explain the connection between relative size and membership trust; each path is
close to or reaches significance (α < .05). Additionally, the indirect effect for only the sequential
path was significant, as the confidence interval for the estimate did not include zero (b = .0248,
CI95% = 0.0001, .1044; 10000 resamples).
Study 2: Additional Analyses
Exploratory Factor Analysis
In Experiment 2, we could only include five items in the exploratory factor analysis (two
distinctiveness, two inclusion, and one trust item), which meant that the analysis could allow
generation only up to two factors. In spite of this limitation, the factor loadings are consistent
with the idea that trust should be treated as a separate factor. This pattern matrix (Table S3)
describes a simple structure for distinctiveness and inclusion, where the items for these measures
strongly load on only one factor, and weakly load on the other. By contrast, trust ratings do not
show a simple structure, instead loading modestly on both factors to similar degrees.
Optimal Distinctiveness and Membership Trust 43
Table S3: Factor Loadings for Distinctiveness, Trust Ratings, and Inclusion (Study 2)
Pattern Matrixa
Factor
Distinctiveness Inclusion
Distinctiveness - Which group do you feel is more distinctive? 1.016 -.076
Distinctiveness - Which group ‘stands out’ more? .924 .013
Trust - Who would you be more likely to trust: A Member of Group A
or B? .357 .339
Inclusion - In which group do members fit in more with the others of
their group? .129 1.030
Inclusion - In which group are members more associated with each
other? .083 .746
Extraction Method: Maximum Likelihood.
Rotation Method: Oblimin with Kaiser Normalization.
a. Rotation converged in 5 iterations.
b. Loadings less than .10 are dropped from presentation
Note. Responses were collected in reference to Group A and B, but scores submitted to this analysis were those
transformed so that higher numbers indicated greater trust, inclusion, or distinctiveness to the minority. Bolded loadings
indicate the factor upon which the item most strongly loaded. All items loaded strongly on a single factor and weakly on
the other two, consistent with a simple factor structure. Distinctiveness and inclusion loaded onto separate factors,
whereas trust loaded onto both.
We also acknowledge that sample size is somewhat modest, and this can be an issue were
communalities low (.4 and below; MacCallum et al., 1999). Under conditions where
communalities are quite large (.6 and above), sample size matters less (MacCallum et al., 1999).
Admittedly, the data from these studies are not ideal for conducting exploratory factor analyses
given the limited number of items, but, of interest, below we show that using the data from them
yields simple factor structures that are consistent with the notion that they are different
constructs, with communalities quite high where expected to be (Study 1, Mcommunality = .82;
Minimum = .55; In Study 2, Mcommunality = .85; Minimum = .61, Study 2 excluded trust
communality, .34, consistent with the idea that this construct is not easily reducible to the
inclusion factor). Overall, then, these data support the notion that trust is moderately related to
Optimal Distinctiveness and Membership Trust 44
inclusion and should be treated as a factor distinct from inclusion and distinctiveness. The factor
correlation produced by this analysis yielded r = .39.
Distinctiveness and Inclusion as Separate Predictors of Trust
As in Experiment 1, we explored whether between-group distinctiveness and within-
group inclusion could be 1) separate or sequential mediators of membership trust.
Separate motives mediation test. To explore the separate contributions of
distinctiveness and inclusion to membership trust, we submitted trust scores from Experiment 2
to a PROCESS analysis (Hayes, 2016, Model 4), with distinctiveness and inclusion as separate
and simultaneous mediators and relative size as the predictor. The results of individual paths are
presented in Figure S3. Unlike Experiment 1, relative size directly predicted distinctiveness and
inclusion (i.e., smaller groups were more distinctive and included), and each was associated with
membership trust. Both indirect effects were significant, through distinctiveness (b = .395, CI95%
= .108, .872; 10000 resamples) or inclusion (b = .348, CI95% = .0447, .813; 10000 resamples).
Figure S3. Separate Motives Mediation Model: Experiment 2
Note. Unstandardized regression coefficients reported. Both indirect paths, through distinctiveness or inclusion
alone, mediated size on trust. * p < .05; ns p > .10.
Optimal Distinctiveness and Membership Trust 45
Sequential motives mediation test. This analysis was conducted with PROCESS using
Model 6 (Hayes, 2016). The results of individual paths are summarized in Figure S4.
Figure S4. Sequential Motives Mediation Model: Experiment 2
Note. Numbers reflect unstandardized regression coefficients. The sequential path did not significantly predict
membership trust. * p < .05; m p = .0544; ns p > .10.
The evidence indicates an indirect sequential path through distinctiveness and then
inclusion, while consistent with the data in Experiment1, was somewhat weaker than the separate
and simultaneous mediator paths; the indirect effect for the sequential path was not significant,
as the confidence interval included zero (b = .0694, CI95% = -0.0026, .3101; 10000 resamples). A
model with inclusion and distinctiveness reversed in the causal order did not change the
outcome; the indirect effect did not reach significance (b = .0804, CI95% = -0.0022, .2912; 10000
resamples).
Discussion
Across both studies, factor analyses indicated that between-group distinctiveness, within-
group inclusion, and membership trust are distinct, but positively related, constructs. Moreover,
the additional mediation analyses allowed us to explore whether one or both needs contributed to
explaining the relative size effect on membership trust. Here, the conclusions are not fully
consistent. Experiment 1 indicated that if we think about the motives separately, the best model
Optimal Distinctiveness and Membership Trust 46
is perhaps a sequential mediation model, one where relative size predicts between-group
distinctiveness, which then predicts within-group inclusion to predict trust. By contrast, in
Experiment 2, the evidence supported the notion that distinctiveness and inclusion are separate
contributors to membership trust (the evidence for the sequential model was also suggestive, but
did not reach conventional thresholds of significance).
It is interesting to consider these models; after all, perhaps the inconsistency could be
addressed by method differences, with observers more likely to adopt the sequential model
whereas those considering membership more likely to adopt a simultaneous model. Or perhaps,
as we note in the paper, it is their combined presence that best explains the data. After all, it
should be noted that the models reported in the paper differ from the simultaneous analysis in an
important way: the combined presence model incorporates the shared variance into the mediator,
the separate mediation model does not, and the sequential motivational model assigns it a causal
role. Future research will be needed to disentangle and identify which of these best explain the
effects of relative size on membership trust.
Supplemental Materials References
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor
analysis. Psychological methods, 4(1), 84.
Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after
all: A cross-discipline view of trust. Academy of management review, 23(3), 393-404.
