Duo Status and Stereotyping 1
RUNNING HEAD: Duo Status and Stereotyping
Unkind to Two of a Kind: Stereotyping Women with Duo Status in a Work Group
Denise Lewin Loyd
Massachusetts Institute of Technology
Judith B. White
Mary C. Kern
Baruch College, City University of New York
Draft. Under revision for resubmission.
Please do not cite without permission.
Author’s note: The authors would like to thank Kim Edgar, Rebecca Gould, Sarah Wang, and Nadia
Zahangir for their research assistance.
Duo Status and Stereotyping 2
Duo Status and Stereotyping 3
Unkind to Two of a Kind: Stereotyping Women with Duo Status in a Work Group
We explore the effects of gender diversity in groups that have few women, calling attention to the
particular case of two women in a group. Drawing on theories of how impressions are formed of
individuals, duos, and groups, we propose that men evaluate women more stereotypically when there are
two women in the group than when there are one or three. Study 1 demonstrated this effect with a sample
of men who participated in an online experiment. Study 2 extended the finding to interacting groups,
showing that men evaluated a woman in their group as contributing less leadership and having fewer
skills when there were two women in the group than when there were one or three. We also found that
men viewed groups with two women as less cohesive than those with one or three women. The results
shed light on duo status, being one of exactly two members of a social category in a group, and on critical
mass, the smallest number of category members in a small group required to reduce stereotyping from the
majority. We discuss the implications of these effects for managing diversity in groups and organizations.
Keywords: duo status, stereotyping, subgroups, gender, entitativity, cohesion
Duo Status and Stereotyping 4
In 2010, the nine-member United States Supreme Court convened for the first time with three
female justices. The first, Justice Sandra Day O'Connor, stated she experienced an excessive amount of
attention as the only woman on the court, and was glad when a second woman was appointed (Rudenstine
2000). However, the second, Justice Ruth Bader Ginsberg, noted that as one of two women, attorneys
arguing before the court confused her with Justice O’Connor at least once a term, even though she and
O'Connor hailed from opposite sides of the political spectrum and did not look alike. After O'Connor
retired in 2006, Justice Ginsberg was glad when she was joined by Justice Sonia Sotomayor in 2009. But
it was not until the confirmation of Justice Elena Kagan in 2010 that Ginsberg opined "the days in which
the first two women were 'curiosities' are finally over" (Sherman 2010). This paper supports Ginsburg’s
opinion, arguing that a distinctive individual in a group is viewed differently by the majority when there
are two of his or her kind (a duo), compared to when there are either one or three. Our language reflects
this tendency to see a duo as a special unit—a “pair,” “two of a kind,” or “peas in a pod.” We know, for
example, that "two is company,” and "one is the loneliest number,” but “three is a crowd.” Psychologists
have theorized about differences in how individuals versus groups are perceived (for a review, see
Hamilton and Sherman 1996), but a duo is different both from an individual and from a group of three
(Simmel and Wolff 1950). Little attention has been paid to the psychological processes associated with
our perception of individuals when they are members of duos and their implications for diverse groups.
Like the term solo status, which refers to being the only member of one's social category in a
group, the term duo status means being one of exactly two members of one's social category in a group.
Recent theory about the absolute number of distinctive category members in a group proposes that in
groups, members of a distinctive duo lack certain advantages that come with solo status, such as visibility,
yet retain certain disadvantages that come from not having critical mass, such as categorization and
stereotyping (Loyd et al. 2008). Little empirical research, however, has examined the effects of group
composition on members of duos relative to solos or to members of larger subgroups. We argue that the
tendency to view an individual who is a member of a duo differently than an individual who is a solo or a
member of a subgroup of three has implications for the dynamics within small groups. As a first step to
Duo Status and Stereotyping 5
understanding this phenomenon, in this paper we address implications of duo status for women in
organizations, specifically whether duo status affects the extent to which women will be stereotyped.
We focus on stereotyping because it is seen as a major impediment to women’s progress in
attaining leadership positions in organizations (Heilman 2001). Although the intuitive presumption is that
problems like stereotyping are linearly related to the number of women in a group (gradually abating as
the number of women increases), there are almost no empirical studies comparing how a woman is
evaluated when she is a member of a duo, relative to when she is a solo or a member of a subgroup of
three or more (but see Taylor et al. 1978). In this paper we develop theory to explain how the change from
one to two to three women in a group affects stereotyping. We propose that going from one woman in a
group to two prompts a significant perceptual shift in how men view women in their groups, and that
going from two women to three prompts a different perceptual shift. We draw on the theory and research
of impression formation (Fisk and Neuberg 1990) and the concept of entitativity (Campbell, 1958) to
predict these changes and their effects on stereotyping. Based on this, we propose that stereotyping is
more complex when it takes place in the context of small groups than has previously been acknowledged.
We hypothesize that a woman will be more stereotyped when she is a member of a duo, than when she is
a solo or a member of a subgroup of three.
We report the results of two studies that test our hypotheses. Study 1 is an experiment in which
male observers evaluate a female team member whose performance is held constant across different
group compositions and sizes. Study 2 examines men's evaluations of their female peers' contributions in
interacting task-performing groups with one, two, or three female members, as well as their evaluations of
the group's cohesion. We conclude by discussing the implications of the findings for future research on
duo status and for managing diversity in organizations.
Women in the Group: One, Two, Three
Gender Stereotypes in Organizations
Gender stereotypes have the potential to bias performance evaluations. A stereotype is a socially
shared set of beliefs about the characteristics, attributes, and behaviors of members of a social category
Duo Status and Stereotyping 6
(Greenwald and Banaji 1995; Hilton and von Hippel 1996). Even if people do not personally endorse a
stereotype, it is hard to escape learning about it, and consequently everyone “owns” a bit of knowledge
about common stereotypes whether they like it or not (Devine 1989). Gender stereotypes have two
dimensions, a relationship-oriented dimension of warmth, and a competence-oriented dimension of
potency or agency (Rudman et al. 2001). Women are considered to be high on the warmth dimension and
low on the competence dimension, thus the stereotype of women is that they are sociable and agreeable,
but not intellectually competent or able (Fiske et al. 2002; Fiske et al. 1999). This is consistent with the
social role widely ascribed to women, that they are (or should be) communal but not agentic (Eagly 1987;
Eagly and Karau 2002). When individuals are stereotyped, they are viewed less as unique individuals
with their own traits and characteristics, and more as category members possessing traits and
characteristics consistent with their group’s stereotype. Conversely, when individuals are seen as category
members rather than unique individuals, they tend to be stereotyped (Biernat et al. 1998; Marques and
Yzerbyt 1988). This can lead to biased evaluations at work, in which women are seen as warm and
supportive but lacking the masculine leadership traits (e.g., dominance, assertiveness) necessary for
organizational advancement (Brenner et al. 1989; Eagly and Karau 2002; Heilman 2001; Schein 1973).
Thus, the effects of stereotyping have real implications for women in organizations.
In her landmark book, Kanter (1977) identified stereotyping as a consequence of numeric
distinctiveness in organizations. When there are few women, their gender is salient, and this triggers
information about gender stereotypes in the minds of perceivers (Hilton and von Hippel 1996). Several
studies support this idea, finding that the proportion of women in a work setting predicts bias in their
performance evaluations. For example, Lortie-Lussier and Rinfret (2002) find that female managers
receive lower evaluations from their male peers when seven percent of managers in the organization are
women, compared to when 20% or more are women. The proportion of women is also linked to bias in
performance evaluations in units of the Israeli military (Pazy and Oron 2001), blue-collar work groups
(Sackett et al. 1991) and among law students (Spangler et al. 1978). In general, the smaller the proportion
of women, the more biased the evaluations of them.
Duo Status and Stereotyping 7
The studies cited above support the link between numeric distinctiveness and stereotyping at the
organizational level, across broad proportions of women and men. To our knowledge, the corresponding
proposition has never been tested at the group level comparing one, two, and three women in a group. A
few studies, however, offer indirect evidence that the relationship between numeric distinctiveness and
stereotyping may not hold in small groups. For example, Taylor et al. (1978) hypothesized, but failed to
support, an inverse linear relationship between stereotyping and the absolute number of women in a small
group. Women in their study appear to have been perceived as warmer and more pleasant when there
were two of them in the group compared to one or three (Taylor et al. 1978; Experiment 3). A more recent
study compared groups with solo (one) or duo (two) women and unexpectedly found that solo women
received more favorable peer evaluations, consistent with less stereotyping, than duo women (Fuegen and
Biernat 2002). Finally, the experiences of professional women themselves suggests that being one of two
is different from being the only one or being one of three. Not only are members of a duo often confused
with one another, as Justice Ginsberg noted above, they also feel categorized and stereotyped (Konrad et
al. 2008). From their study of women on the boards of Fortune 1000 companies, Konrad and Kramer
“Adding a second woman to a board helps reduce the sense of isolation, but it doesn’t always
cause change and may create its own difficulties. Two women may be perceived as a separate
group and may find they have to be careful not to appear to be conspiring. What’s more, they may
not be distinguished from each other. One woman we spoke to explained, “I raised a question at a
board meeting that caused the board to take some important action. Later on, the chairman thanked
the other woman on the board for raising the question. No one said anything to correct him.” A
clear shift occurs when boards have three or more women. At that critical mass, our research
shows, women tend to be regarded by other board members not as “female directors” but simply as
directors, and they don’t report being isolated or ignored." (Konrad and Kramer 2006: 84)
These findings suggest that the phenomenology of duo status should not be thought of as a state
simply midway between solo status and critical mass. We argue that some of the perceptions of women
Duo Status and Stereotyping 8
with duo status, including stereotyping, may be explained in terms of the individuality of a solo,
contrasted with the entitativity of a duo relative to a critical mass of three. We first turn to the theory of
impression formation to describe the perceptual distinction between one woman and two before
discussing how entitativity creates a perceptual distinction between two women and three.
Seeing the Individual, Not the Group
When colleagues form an impression of a coworker, it is important whether that coworker is seen
as an individual or as a representative of a social category. If men see a woman in the group as an
individual, they are less likely to categorize and stereotype her. Fiske and Neuberg's (1990) influential
theory of impression formation posits that perceivers process information about a person along a
continuum. At one end, perceivers form an impression based solely on categorization and no
individuating information, which results in stereotyping and biased evaluations. At the other end,
perceivers form an impression based solely on individuating information and no category information,
which results in accurate person perception and unbiased evaluations. Most of the time, however,
perceivers use a combination of processes that fall somewhere in the middle of the continuum. According
to Fiske and Neuberg, perceivers start to form an impression by categorizing (e.g., Jennifer is a woman),
and then shift towards individuation by taking into consideration information about the specific individual
(e.g., Jennifer is in banking; Jennifer specializes in biotechnology start-ups). How far perceivers shift
toward the individuation end of the continuum depends on two mediating processes: how much attention
they pay to Jennifer’s individual characteristics and behaviors, and how motivated they are to form an
accurate, unbiased impression of Jennifer (Fiske and Neuberg, 1990). Both processes have implications
for how men form impressions of women when there are one, two, or three women in the group.
The first mediating process in impression formation is attention. To the extent a perceiver pays
attention to a woman in a group, he will be exposed to information about her individual attributes,
characteristics and behavior. He may then use that information to form an individuated impression of her.
Research suggests that a solo woman will receive more attention than a member of a duo or larger
subgroup. Some evidence for this assertion comes from studies based on Taylor et al.'s (1978) "Who said
Duo Status and Stereotyping 9
what?" experimental paradigm. Participants watch a tape or see a slide show of a group discussion, then
are tested for their memory of who said what. The rationale is that participants can remember what a
group member said only if they paid attention to that member. Participants pay more attention to a
minority member when s/he is the only minority in the group, than when s/he is a member of a subgroup
of two or three (Taylor et al. 1978). A recent study also finds more accurate memory for a solo woman's
contributions than for the contributions of a woman when there are two or more in the group (Klauer et al.
2002). Increased individuation as a result of attention may seem to run counter to the fact that the
distinctiveness of a solo makes her gender salient, which immediately triggers categorization and
stereotyping. According to the theory of impression formation, however, if a solo behaves in any way
inconsistent with the stereotype, the attention she is paid will mediate the processing of that individuating
information, reducing stereotyping (Fiske and Neuberg, 1990). We argue, therefore, that men's impression
of a solo will be more individuated and further toward the unbiased end of the continuum, than their
impression of a member of a duo.
The second mediating process in impression formation is motivation. The more a colleague is
motivated to form an accurate, unbiased impression of a coworker, the more likely he is to see her as an
individual and not a category member. Men may be motivated to form an unbiased impression of female
coworkers because research suggests that they do not want to appear sexist in the workplace (Koenig and
Richeson 2010). However, because stereotyping is automatic, they must be vigilant to avoid it. If it is
unlikely that they will be seen as sexist, self-correction is unlikely. For example, when undergraduate men
are first given an opportunity to demonstrate that they are not sexist, or primed to think of themselves as
objective, they are less guarded in their subsequent judgments of women (Monin and Miller 2001;
Uhlmann and Cohen 2007). We argue that a solo woman's distinctiveness acts as a contextual cue to
make men vigilant about not stereotyping her. Therefore, the motivation not to be seen as sexist has
implications for members of female duos relative to solos. Moreover, a man’s evaluation of a solo woman
in a group may be construed as a signal of whether or not he is sexist, because the lack of other women in
the group makes it harder to differentiate his evaluation of this woman from his evaluation of all women.
Duo Status and Stereotyping 10
When there is more than one woman present in the group, however, a man's evaluation of any individual
woman may be attributed to something about her as an individual rather than to his attitudes about women
in general. Thus, if he is concerned about not appearing sexist, a man should stereotype a solo woman less
than he stereotypes a member of a female duo or larger subgroup. To accomplish this, a man may either
carefully judge the woman according to her individual characteristics and behavior or, if that’s not
possible, he may self-correct for any possible negative effects of stereotyping by being a little extra
positive when evaluating her (Greenwald and Banaji 1995).
In sum, a solo stands out more and receives more attention than a member of a duo. The salience
of her gender acts as a cue to remind men to be vigilant about not stereotyping her. Because they are
motivated not to be sexist, and naturally pay more attention to her, we hypothesize that men's impressions
of a solo woman will shift further down the impression formation continuum toward an unbiased,
individuated impression, than will their impressions of a member of a female duo. Ironically, the
individuation of the solo results in comparatively more biased evaluations for the member of a duo.
Therefore, we hypothesize:
Hypothesis 1. Men will evaluate a woman more stereotypically when there are two women in her
group, compared to when there is one.
If Hypothesis 1 appears to favor solo status for women over duo status, that is not our intention.
There are other distinctions that could be made between groups that have one woman and groups that
have two, and in addition there are many interesting aspects to the dynamic between two women in a
group. Our emphasis here, however, is on how men perceive women in a group when there are exactly
two women. To bring duo status into sharper focus we must also consider how it compares to a situation
where there are three women in a group, compared to two. Whereas one woman may be perceived as an
individual, more than one woman may be perceived as a collection of individuals, or as a subgroup within
the group. We examine the shift from two to three women in a group through the theoretical lens of
entitativity (Campbell 1958).
Duo Status and Stereotyping 11
The Entitativity of Minority Subgroups
Entitativity is the degree to which a collection of individuals has the property of an entity, of
being a coherent whole (Campbell 1958). Members of an entitative group tend to be perceived as similar,
sharing traits and attributes (Crawford et al. 2002). The degree of entitativity also represents the strength
of the associative link among individual group members (Brewer and Harasty 1996). Thus, if women in a
group are perceived as entitative, it will be hard for men to think of one individual woman without
thinking of the other(s). Consistent with this, Hamilton and Sherman (1996) suggest that members of
highly entitative groups are seen as interchangeable, echoing Justice Ginsberg's experience. As a result,
members of a highly entitative group are perceived more toward the categorization end of the impression
management continuum, and less toward the individuation end, than members of a less entitative group.
In other words, entitativity facilitates stereotyping (Crawford et al. 2002).
We claim here that a duo is perceived as more entitative than a trio (a subgroup of three). Our
support comes primarily from Simmel (Simmel and Wolff 1950), who wrote at length on the sociology of
small numbers. Simmel contrasted the duo with small groups of three. The duo, he wrote, was inseparable
from its members. If any member exits, the social unit no longer exists. “For the outsider, the group
consisting of two may function as an autonomous, super-individual unit” (Simmel and Wolff 1950: 123).
In contrast, Simmel considered a trio to exist as a social unit independent of any individual member—if
any member exits, the social unit still exists. Simmel proposed further that a collection of three gives any
two members the opportunity to differ without threatening the group as a whole. This implies that
members of a trio may be perceived as less similar to one another than members of a duo. Konrad et al.
(2008) found support for this idea, citing a male CEO who said that once there are three women on a
board, the male members see that the women differ on multiple dimensions. "The three women don’t
always agree with each other, and that is healthy for the men to see. They are independent," (Konrad et al.
2008: 154). Simmel noted that in this regard, there is a vast difference between a duo and a trio, but little
marginal difference between a trio and a group of four or more. Thus, we assert, a minority duo is highly
entitative and its members strongly associated with one another and a minority trio is less entitative and
Duo Status and Stereotyping 12
its members less strongly associated with one another.
According to Fiske and Neuberg's (1990) impression management continuum, a member of a
female duo or trio will be immediately categorized by gender. But, because of the high entitativity of a
duo, each member is likely to be perceived as strongly associated with the other woman in the duo, even
to the point of being interchangeable with her. In contrast, a member of a trio will be perceived as less
strongly linked to either of the other two women, and assumed to be less similar to them. Because
entitativity is associated with stereotyping, we draw the inference that men will stereotype a member of a
female duo more than a member of a subgroup of three. We hypothesize:
Hypothesis 2: A woman will be evaluated more stereotypically when there are two women in her
group than when there are three.
Study 1 was conducted to test Hypotheses 1 and 2, that men evaluate a woman more
stereotypically when there are two women in the group than when there are one or three. To test this, we
designed an experiment in which participants read about a target woman's contributions to a group
discussion, and we manipulated whether the woman was a female solo, a member of a female duo, or a
member of a female trio. To disentangle the absolute number of women from the proportion of women in
the group, we crossed the number of women (1, 2, 3) with group size (5, 7, 10, 14) to test whether
stereotyping of a female duo was due to the absolute number of women in a group, in which case we
would observe only the predicted main effect, or to the proportion of women in a group, in which case we
would observe an interaction between the number of women in the group and the group size.
Participants and design. We recruited participants from Amazon's Mechanical Turk website
(MTurk). The MTurk worker pool is closely representative of the general population of the United States,
albeit with slightly more education and a slightly lower income (Paolacci et al. 2010). When they are
offered a reasonable wage, MTurk workers provide data that is as reliable and valid as other internet or
university subject pools (Buhrmester et al. in press). We offered $1.01 in exchange for completing a 7-10
Duo Status and Stereotyping 13
minute online survey, roughly equivalent to the current minimum wage in the United States. We limited
participation to male workers at least 18 years of age whose accounts were registered in the United States.
Participants were randomly assigned to one of 12 conditions resulting from crossing the number of
women (1, 2, 3) by group size (5, 7, 10, 14).
Procedure and measures. We created a context for participants to evaluate a female group
member based on a group task called ACME Investments (McLeod et al. 1997). Depending on condition,
participants saw one of 12 different graphics composed of male and female avatars representing ACME
Management. Figure 1 shows the graphic that appeared in the 3 women, 10 member condition. The target
female for evaluation was always in the same position, on the left of the graphic. In the duo female
conditions, the female on the lower right was replaced with a male. In the solo female condition, both the
lower right and lower left females were replaced with males. In all conditions, avatars were the same size
and the graphic as a whole took up the same amount of space on the screen. The graphic remained on the
screen throughout the survey until the demographic questions were presented at the end.
--------Insert Figure 1 about here--------
Participants first answered four questions about the team represented in the graphic: the number
of team members, how they were dressed, the number of men and women, and the shape of their
arrangement. Only participants who gave correct answers were allowed to proceed. If they gave incorrect
answers, they were asked to try again. Next, the target female was identified as Jennifer,1
1 "Jennifer" was the most popular girl's name in the United States in the 1970s and 1980s, according to the
member the participant was to evaluate. Participants next read a summary in six segments of the ACME
Management Team's discussion about which of several target companies it should acquire. For each
segment, participants answered a question about the discussion summary. Again, only participants who
gave correct answers were allowed to proceed. If they gave incorrect answers, they were asked to try
Duo Status and Stereotyping 14
again. Jennifer's contributions were attributed to her (e.g., "Jennifer suggested,"); other team members'
contributions were attributed anonymously (e.g., "someone suggested"). Altogether, Jennifer made
contributions in four out of six of the segments. In the third part of the survey, participants were asked to
give their impressions of Jennifer, and told there were no correct answers. They evaluated Jennifer on
eight bipolar scales measuring gender stereotypes from +3 to -3 (Rudman et al. 2001). These were strong-
weak, dominant-submissive, harsh-lenient, hard-soft (the left-most word in each pair representing gender
potency), and warm-cold, supportive-detached, trusting-skeptical, caring-distant (the left-most word in
each pair representing gender warmth). Following Rudman et al., we subtracted the sum of the gender
potency items from sum of the gender warmth items to create an index of female stereotypicality.
In the final segment, participants provided information about themselves, including age,
employment and student status, ethnicity, and level of education. Level of education was measured on an
ordinal scale from 1 (did not finish high school) to 8 (doctoral degree).
A total of 231 men completed the survey. They ranged in age from 18 to 67 (M = 30.75, SD =
10.67). Most (94%) were native English speakers. They were 78% White, 8% Black, 7% Asian, 4%
Hispanic, and 2% Native American (2% preferred not to say). Thirty-eight percent had some college or an
associate's degree, 33% had a bachelor's degree, and 15% had a master's or professional degree. Most
were employed (48% full-time; 18% part-time). They had between 0 and 42 years' work experience (M =
10.11, SD = 10.06). Twenty-seven percent were students. Stereotypicality scores ranged from -16 to 7, M
= -1.51, SD = 3.74. On average, Jennifer was perceived as having greater gender potency than gender
warmth, which is not surprising since her contributions gave direction to the group's task rather than
support to the group's members. One outlier (female stereotypicality greater than 3 standard deviations
from the mean) was eliminated from hypothesis tests. Means, standard deviations, and correlations among
study variables appear in Table 1. Because age and level of education were correlated with gender
potency and female stereotype, respectively, these were retained as control variables.
A number of women (1, 2, 3) by group size (5, 7, 10, 14) ANCOVA, with participant's age and
Duo Status and Stereotyping 15
education level entered as covariates, found a significant effect of number of women on evaluations of the
target woman, Jennifer, F(2, 216) = 4.43, p = .01. Jennifer was viewed as more stereotypically female
when she was one of two women (M = -0.59, SE = .41) than when she was a solo (M = -2.31, SE = .42, p
< .01, one-tailed) or one of three women (M = -1.70, SE = .43, p < .05, one-tailed), supporting
Hypotheses 1 and 2. Follow-up analyses revealed that Jennifer was viewed as significantly less potent (p
< .01, one-tailed) and more warm (p = .05, one-tailed) in the duo condition than in the solo condition, and
as more warm (p < .05, one-tailed) in the duo condition than in the trio condition. Neither the group size,
F(3, 216) = 1.08, p = .36, nor the Number of Women X Group Size interaction, F(6, 216) = 0.98, p = .44,
We found that men stereotyped Jennifer more when she was a member of a female duo than when
she was a solo or one of three women. Both Hypotheses 1 and 2 were supported. In Study 1 the
contributions of a target woman, Jennifer, remained constant across all conditions; only the group
composition changed. Because there was no interaction between the number of women and the group
size, we infer that the absolute number of women, rather than a strict proportion, predicts stereotyping in a
small group. The logical inference is that when there are exactly two women in a group as opposed to one
or three, men view them more as members of a social category, as women, and less as individual
members of the group. This has implications for the men's view of the group's cohesion as well as their
evaluation of its female members. However, Study 1 measured the evaluations of male observers, not of
men who were themselves members of the group. In Study 2 we extended our findings to interacting
groups which allows us to further examine the implications of duo status for women in small groups in
Implications of Two Women for Group Cohesion
The individuation versus entitativity of a small number of women in a group has implications for
group dynamics as well as for stereotyping individual group members. Cohesion, or social integration, is
the extent to which group members feel connected to one another and will continue to work together
Duo Status and Stereotyping 16
(O'Reilly et al. 1989). Previous research has found an overall negative relationship between gender
diversity and cohesiveness (Harrison et al. 1998; Randel 2002), but has not examined the effect of one
versus two versus three women in a group on male members' attitudes of group cohesiveness. Based on
our reasoning that men will individuate a solo female more than a member of a duo, we would predict that
male members will have stronger ties with a female solo in the group and feel more cohesion in groups
with one versus two women. Our prediction is also consistent with research showing that individuals are
treated more favorably than groups. In particular, more cooperation is extended to individuals than to
groups, even to groups of two (Insko et al. 1992). Therefore, we hypothesize:
Hypothesis 3. Men in groups with two women will experience the group as less cohesive than
will men in groups with one woman.
The entitativity of a subgroup within the group should also have an impact on the strength and
nature of interpersonal ties between members of the subgroup and the rest of the group. If a subgroup of
two, a duo, is perceived as more entitative than a subgroup of three, then other group members may not
reach out to members of the duo. People perceive a pair to be balanced, stable, symmetrical and good
(Menon and Phillips in press). As described above, a duo appears more like a unified entity, its members
natural allies. We predict that male members will establish stronger interpersonal ties with and feel more
connected to members of a less entitative subgroup than a more entitative subgroup, so they will have
stronger ties with women in a group when there are three than when there are two. O‘Leary and
Mortensen (2010) found analogous evidence of this dynamic in geographically diverse teams. They found
that six-member teams with one off-site isolate (a solo) had greater team identification, greater transactive
memory, lower conflict, and fewer coordination problems than teams with two or three members at a
second site. The worst configuration to have, in terms of team processes, was two members at one site
and four at the other. We speculate that cohesiveness was also lower in teams that were split 4/2 than in
any other configuration, and expect to observe the same result in gender diverse teams. Therefore, we
Hypothesis 4: Men in groups with two women will experience the group as less cohesive than
Duo Status and Stereotyping 17
will men in groups with three women.
In organizations, evaluations are typically made on performance dimensions, rather than personal
traits like dominance and supportiveness that were used in Study 1. Performance evaluations can be
affected by stereotyping when there are relatively few women in a given setting (Pazy and Oron 2001;
Sackett et al. 1991; Spangler et al. 1978), so the results of Study 1 imply that women will receive lower
performance evaluations when they are one of two women in a group than when they are the only woman
or one of three women. Study 2 used a group task as the performance context, so we assessed peer
evaluations of leadership, skills and effort (Hackman and Wageman 2005) as our dependent measures.
These dimensions, particularly leadership, are not gender-neutral standards. Research has shown that
leadership is perceived as agentic and masculine, and that leaders are perceived as male (Eagly and Karau
2002; Heilman et al. 1995). Skills on many tasks are also considered masculine. Therefore, consistent
with Hypotheses 1 and 2, if women are stereotyped more as members of female duos, we predicted they
would be evaluated as contributing less leadership and fewer skills, but not different levels of effort, than
women who were solos or members of a subgroup of three. Consistent with this, we also expected that
women would feel more stereotyped when there were two women in the group, compared to one or three.
In this study we also tested Hypotheses 3 and 4, that men would perceive their groups to be less cohesive
when there were two women in the group, compared to one or three. We tested these predictions in
interacting groups of seven or eight members engaged in a classroom task to assemble a model person
constructed of interlocking construction pieces.
Pretest participants. Although stereotyping results in lower performance evaluations for women
when they are in the minority, this effect has been shown to occur only when the task itself is masculine
(Chatman et al. 2008; Karakowsky and Siegel 1999). We therefore conducted a pretest to determine that
our task was, in fact, perceived as masculine. Using email, we recruited 169 students at a private U.S.
university (40% male; 60% White, 18% Asian, 4% Black, 4% Hispanic, 13% did not specify their
Duo Status and Stereotyping 18
ethnicity), ages 18-25, M = 20.97, SD = 0.67. They agreed to complete an online survey in exchange for
Pretest method and results. Pretest participants read a scenario that described the group task in
Study 2: "Imagine you are doing an in-class activity in an organizational psychology class. In the front of
the room is a model person constructed of interlocking construction pieces. The instructor assigns
everyone to a group with seven members. Each group gets an identical kit with the required pieces and
must work together to assemble an exact replica of the model as quickly as possible." They then
responded to four statements adapted from Karakowsky and Siegal (1999) on scales of 1 (strongly
disagree) to 7 (strongly agree). Two items asked whether men or women would be considered more
expert at the task, and two items asked whether men or women would be perceived as more
knowledgeable about the task. Pretest participants agreed strongly that men (M = 4.24, SE = .13) would
be perceived as more expert at the task than women (M = 2.71, SE = .10), F(1, 164) = 74.62, p < .01, and
that men (M = 4.17, SE = .13) would be perceived as more knowledgeable about the task than women (M
= 2.68, SE = .10), F(1, 164) = 82.07, p < .01. There were no significant interactions with pretest
participant gender on either set of items. We concluded that both men and women perceived the task to be
Participants and design. Study participants were 165 students (37 women) in two sessions of a
non-credit summer education program at a U.S. business school (14% International; 10% Asian, 6%
Hispanic, 5% Black, 11% did not specify their ethnicity), ages 18-25, M = 21.54, SD = 1.13. Participants
were offered a chocolate bar in exchange for completing a questionnaire at the end of a classroom
exercise. We composed groups with seven or eight members that were majority male and had either one
(solo condition), two (duo condition), or three (trio condition) female members. Within this design,
individuals were randomly assigned to groups.2
Task. We used a task commonly used in MBA classrooms to illustrate group dynamics, the
2 Students who worked in groups with other gender combinations were not included in the study.
Duo Status and Stereotyping 19
LegoTM Person assembly (Reddy and Kroeger 1972). The instructor placed a model LegoTM Person in the
front of the room. Groups were provided with the LegoTM pieces required for the task, which was to
assemble an exact replica of the model as quickly as possible after being allowed up to 40 minutes of free
time to plan. An observer timed each group with a stopwatch, beginning when they touched their pieces
and ending when they finished their assembly. If the instructor determined that a group's LegoTM Person
was not an exact replica of the model, the observer restarted the stopwatch while the group made another
Measures. Dependent measures were taken immediately after each group had finished the task.
Each group member completed a peer evaluation form, evaluating every other member of their group on
leadership, skills, and effort. Leadership was defined as “how much the person contributed either to
strategic direction to define the team's goal and/or process management to move the team toward its
goal.” Skills were defined as “how much competence the person possessed on skills that the team needed
to accomplish its goal.” Effort was defined as “how hard the person worked at the task.” Scale anchors
were 0 (did not contribute) and 9 (worked extremely hard, had extremely useful skills, contributed
extremely helpful leadership). Because we were interested in performance evaluations as a measure of
stereotyping, we used only men's evaluations of women. The number of male peers who evaluated each
woman ranged from four to seven. These were combined for each woman to form mean male peer
evaluations of leadership, of skills, and of effort. We measured cohesion with a five item scale modified
from Smith et al. (1994). A sample item is "Our team was united in trying to reach its goals for
performance." Alphas were .90 for men and .91 for women. To assess whether women felt stereotyped,
participants also responded to the statement "In my Lego Assembly team, I felt stereotyped because of my
gender," on a scale from 1 (strongly disagree) to 7 (strongly agree). Finally, participants indicated their
gender, age, ethnicity, and national origin.
Analyses. Our independent variable was the number of women in the group, and our unit of
analysis was each individual woman. Because women who were members of duos and trios were nested
in groups, we used the MIXED procedure in SPSS 16.0 to estimate effects of group composition on our
Duo Status and Stereotyping 20
dependent measures of peer evaluations and men's perceptions of cohesion, as well as women's
perceptions of stereotyping. Duo female groups were dummy coded as the reference condition, so that we
could report estimates for the difference between duo and solo, and the difference between duo and trio
conditions. In all hypothesis tests using the MIXED procedure, the unstandardized estimate b is given as
the effect size estimate, and directional one-tailed tests are reported. The Satterthwaite approximation is
used to estimate degrees of freedom.
Preliminary analyses. The sample consisted of 37 women and 128 men nested in 23 groups. Two
groups did not finish the task: a solo female group and a duo female group. Final assembly time for the 21
groups that completed the task ranged from 1.93 to 14.58 minutes, M = 5.25, SD = 3.03. Group
composition did not significantly predict assembly time, F(2, 18) = 0.35, p = .71, but in our sample
groups with one woman (M = 5.74, SD = 0.96, n = 13) and groups with three women (M = 4.76, SD =
1.99, n = 3) took longer than groups with two women (M = 4.29, SD = 1.54, n = 5). Descriptive statistics
and correlations appear in Table 2. Although they did not reach statistical significance, a woman's age
was negatively correlated with her peer evaluations, and her international status was positively correlated
with her perception of being stereotyped; these were retained as control variables. In addition, we
included group size as a control variable, because it approached a negative correlation with cohesion, and
we included a dummy variable to represent session, as the overall sample characteristics were slightly
different between the two sessions. We also controlled for each group's completion time.
Peer evaluations. We predicted that male peers would evaluate women as contributing less
leadership and fewer skills, although not necessarily less effort, when there were two women in the group,
compared to when there were one or three. To test this, we submitted each woman's mean male peer
evaluations of leadership, skills, and effort to separate MIXED models as described above. Compared to
women in duo female groups, both solo women, b = 1.44, SE = .54, t(19.50) = 2.69, p < .01 and women in
trio female groups, b = 1.57, SE = .69, t(12.50) = 2.28, p = .02, received higher evaluations of leadership
from their male peers. Compared to women in duo female groups, both solo women, b = .72, SE = .33,
Duo Status and Stereotyping 21
t(21.38) = 2.21, p < .01, and women in trio female groups, b = .80, SE = .42, t(14.79) = 1.90, p = .04, also
received higher evaluations of skills from their male peers. And compared to women in duo female
groups, solo women, b = .66, SE = .39, t(21.28) = 1.69, p = .05, received higher evaluations of effort from
their male peers, but women in trio female groups did not, b = .57, SE = .53, t(18.24) = 1.07, p = .15. The
pattern of effects suggests that the differences between women in duo female groups and other women in
the study was greatest for male peer evaluations of leadership, intermediate for evaluations of skills, and
smallest for evaluations of effort, providing support for Hypotheses 1 and 2.
Women's perceived stereotyping. Consistent with our arguments, we believed that women in duo
female groups would feel more stereotyped by their male peers than solo women or women in trio female
groups. Women's responses to the statement "I felt stereotyped because of my gender" were submitted as
the dependent measure to the MIXED procedure, as described above. Both solo women, b = -1.57, SE =
.74, t(19.20) = 2.13, p = .02, and women in trio female groups, b = -2.13, SE = 1.01, t(17.74) = 2.11, p =
.02, felt less stereotyped because of their gender than women in duo female groups. This is consistent
with Hypotheses 1 and 2 and supports our interpretation of the peer evaluations, above, as evidence of
Cohesion. We predicted that male peers would evaluate duo female groups as less cohesive than
solo or trio female groups. To test this, we submitted men's cohesion scale scores to the MIXED
procedure as described above. Men evaluated both solo female groups, b = .55, SE = .25, t(21.23) = 2.21,
p = .02, and trio female groups, b = .62, SE = .37, t(26.43) = 1.68, p = .05, as more cohesive than duo
female groups. Hypotheses 3 and 4 were supported.
Male-dominated, mixed-gender groups of seven or eight participants completed a masculine task
of building a model with interlocking construction pieces. Women received lower evaluations of their
leadership and skills from their male peers when they were members of groups that had exactly two
women, compared to groups with one or three women. For their part, women felt more stereotyped when
their group had two women, compared to when their group had one or three. Men reported that groups
Duo Status and Stereotyping 22
with two women were not as cohesive as groups with one or three women. Gender composition did not
predict a group's performance, and in fact the duo female groups in our sample performed better—they
took less time, on average, to assemble the model than the solo or trio female groups. Actual group
performance, therefore, cannot explain our results. The picture that emerges from these results is
consistent with Study 1. When there are exactly two women in a group, men evaluate the women more
We sought to shed light on how a member of a minority duo, in this case, one of two women, is
perceived by her majority peers in a group. In Study 1, men evaluated a woman more stereotypically
when she was one of two women in a group, than when she was the only woman or one of three women.
Study 2 replicated this finding for men’s evaluations of a woman’s leadership and skills on a masculine
task. Consistent with this, women felt more stereotyped when they were one of two than when they were
a solo or one of three women. Thus, Hypotheses 1 and 2 were supported. Furthermore, men in groups
with two women reported their groups were less cohesive than men in groups with one or three women,
supporting Hypotheses 3 and 4. Instead of a gradual easing of stereotyping as the number of women in a
group increases from one to two to three, we find stereotyping peaks when there are two women in a
Our paper contributes to the literature on diversity in work groups by demonstrating that the
relationship between the number of minority group members and stereotyping is not linear in a small
group. By incorporating theories of person perception, specifically the impression formation continuum
(Fiske and Neuberg 1990) and the concept of entitativity (Campbell 1958), into the group composition
literature, we argued that having two minority members in a group leads majority group members to view
the minority duo as highly entitative, which should lead to increased stereotyping and decreased
cohesiveness in groups with two minority members relative to one or three. Our results were consistent
with these predictions.
Our work has several important implications for research on gender diversity. First, this is the
Duo Status and Stereotyping 23
first empirical work to test the hypothesis that two represents a unique context that is distinct from one or
more than two. We document that a duo is subject to more pressure from the majority (with respect to
stereotyping) than is a solo or a member of a group with three. In spite of our lay assumptions, we find
that two heads may not be better than one in the way they are viewed by the majority. Second, our
findings suggest that three are a sufficient number of women to begin to reduce the automatic processes
of categorization and stereotyping in a small group. We provide further empirical support for the idea that
in relatively small groups, three represents the critical mass of women needed to overcome the negative
pressures associated with being in the minority in a group (Konrad et al. 2008). Third, we extend the
literature on numeric distinctiveness and biased performance evaluations by showing that when they
represent a subgroup within a group, women are more likely to be stereotyped and given lower
performance evaluations by men on a masculine task when there are two women in the group than when
there are one or three.
Implications for Managers
Our results have important implications for organizational efforts to manage diversity and
diversify organizations. First, we found that men in groups with two women stereotyped them more and
gave them lower performance evaluations than those in groups with one or three women. If the males in
the majority see the performance of two women as worse than that of a solo, they may be disinclined to
continue to diversify the group with respect to gender as they may incorrectly believe that if two was
worse than one, three may be worse than two. Second, we found that men in groups with two women
reported less group cohesion than those in groups with one or three women. Although we did not find that
groups with two women performed worse than those with one or three, cohesion has been tied to group
performance and intentions to continue to work together (Menon and Phillips in press; Mullen and
Copper 1994). This suggests that more diverse groups (i.e., those with two women) may be in greater
jeopardy of negative performance and disbanding than less diverse groups (i.e., those with one woman).
Finally, we found that women felt stereotyped when there were two women in the group, consistent with
Konrad et al. (2008). Concerns about sexism have been linked to a woman’s intention to leave a position
Duo Status and Stereotyping 24
(Laband and Lentz 1998), which suggests that women may be more inclined to leave the organization
when there are finally two of them than when there was only one. This may represent something of a
revolving door that as two come in one (or both) are more likely to leave. Taken together, this work
suggests that organizations concerned with increasing diversity and minimizing stereotyping and sexism
should make a serious effort to move beyond having only two women in a group, department, or on a
board. These efforts at further diversification are vital as they are likely to have important positive
consequences at the individual, group, and organizational levels.
Our theoretical contribution also has implications for managers. We proposed that person
perception in a small group is an important antecedent of stereotyping and cohesiveness. This theory
assumes that the underlying psychological processes for individuating solo and trio women and for
stereotyping duo women are largely automatic. However, the mediating processes of attention and
motivation in the impression formation continuum (Fiske and Neuberg 1990) are at least partly volitional.
That means that majority group members who consciously pay attention to members of duos and are
motivated not to stereotype them can overcome some of the negative effects of duo status in their groups.
Managers who wish to improve group dynamics in groups with minority duos can call attention to this
theory and our results and implement strategies for group members to learn more about each other and
strengthen interpersonal ties. This should help alleviate some of the problems described above.
This work represents an initial exploration into a complex phenomenon and as such there are both
limitations and many opportunities for future research. One limitation is that we restricted our studies to
women in stereotypically male contexts. Thus, we cannot say whether duo status would be experienced in
the same way for men and women. Low status duos may be more at risk for negative consequences like
stereotyping than high status duos. A second limitation in Study 2 is that we cannot tell if being part of a
duo resulted in the women actually behaving differently and how much that impacted their performance
evaluations. The results from Study 1 suggest that part of the reason for the lower performance
evaluations of women duos by the men in their groups was that the men stereotyped them more than they
Duo Status and Stereotyping 25
stereotyped female solos or trios. However, in Study 2 the women also reported feeling more stereotyped
and as a result they may have performed worse on the task due to experiencing stereotype threat (Inzlicht
and Ben-Zeev 2000).
In addition to exploring how context impacts duo status, it is important to broaden the
examination of duo status beyond gender to other categories such as race and sexual orientation as well as
nationality and departmental affiliation. Doing so will help us understand how other factors, such as the
visibility and status of the category, impact stereotyping of duos. Future work should also move beyond
stereotyping of the minority by the majority to other important outcomes such as the extent to which the
majority provides support to the minority. The idea that a solo is otherwise isolated without the majority
group member’s support, suggests that the majority will go out of their way not only not to stereotype a
solo but also to include them and provide social support. Seeing a duo as a unit suggests that duos may
get much less support from the majority although they still need that support. The majority may leave the
duo alone assuming that they “have each other.” Because individual members of the duo are likely to be
seen as interchangeable, the majority may not solicit the opinions of both feeling that one can speak for
the other and they may not share information equally with both assuming that one will pass the
information along to the other.
This paper does not explore the dynamics within duos. In addition to being seen as a unit, duos
may believe that others see them as such. They may even feel some obligation and pressure to provide
support to a similar other. On the other hand, duos may also feel concern about having their category
highlighted due to the presence of a similar other and thus may try to distance themselves from each
other, setting up a negative dynamic (Loyd et al. 2008). Konrad et al. (2008) described two women on a
board as "the conspiracy stage" in which the women thought twice about sitting next to one another, lest
the men think they were up to something. With respect to the dynamics within the subgroup, a minority
group of three becomes another interesting case to explore. Because of the opportunities for coalitions in
three person groups, a subgroup of three is likely to experience more complicated dynamics of trust and
Duo Status and Stereotyping 26
support between group members.
The changing demographics in the workforce have focused increasing attention on whether and
how organizations can successfully diversify their ranks. In spite of the demographic trends, however, the
change in organizations has been slow, especially in the upper ranks. Thus, in many cases rather than
understanding how sweeping demographic changes are impacting organizations and those who are
demographically different, we are dealing with the impact of incremental change as women and
underrepresented minorities go from being the only one in their work group to one of two or more. Our
findings suggest that more attention should be paid to the effects of these incremental changes and that
organization’s efforts to move beyond small numbers of women in groups can have great potential
Duo Status and Stereotyping 27
Biernat, M., C.S. Crandall, L.V. Young, D. Kobrynowicz, S.M. Halpin. 1998. All that you can be:
Stereotyping of self and others in a military context. J. Personality Soc. Psych. 75(2) 301-317.
Brenner, O.C., J. Tomkiewicz, V.E. Schein. 1989. The relationship between sex-role stereotypes and
requisite management characteristics revisited. Acad. Management J. 32(3) 662-669.
Brewer, M.B., A.S. Harasty. 1996. Seeing groups as entities: The role of perceiver motivation. R.M.
Sorrentino, H.E. Tory, eds. Handbook of motivation and cognition. Guilford Press, New York, 347-370.
Buhrmester, M.D., T. Kwang, S.D. Gosling. in press. Amazon's Mechanical Turk: A new source of
inexpensive, yet high-quality, data. Perspectives Psych. Sci.
Campbell, D. 1958. Common fate, similarity, and other indices of the status of aggregates of persons as
social entities. Behav. Sci. 3(1) 14-25.
Chatman, J., A. Boisnier, S. Spataro, C. Anderson, J. Berdahl. 2008. Being distinctive versus being
conspicuous: The effects of numeric status and sex-stereotyped tasks on individual performance in
groups. Organ. Behav. Human Decision Processes 107(2) 141-160.
Crawford, M., S. Sherman, D. Hamilton. 2002. Perceived entitativity, stereotype formation, and the
interchangeability of group members. J. Personality Soc. Psych. 83(5) 1076-1094.
Devine, P.G. 1989. Stereotypes and prejudice: Their automatic and controlled components. J. Personality
Soc. Psych. 56(1) 5-18.
Eagly, A. 1987. Sex differences in social behavior: A social-role interpretation. Lawrence Erlbaum.
Eagly, A.H., S.J. Karau. 2002. Role congruity theory of prejudice toward female leaders. Psych. Rev.
Fisk, S., S. Neuberg. 1990. A continuum of impression formation, from category-based to individuating
processes: Influences of information and motivation on attention and interpretation. Advances in
experimental social psychology 23 1-74.
Fiske, S.T., A.J.C. Cuddy, P. Glick, J. Xu. 2002. A model of (often mixed) stereotype content:
Competence and warmth respectively follow from perceived status and competition. J. Personality Soc.
Psych. 82(6) 878-902.
Fiske, S.T., J. Xu, A.C. Cuddy, P. Glick. 1999. (Dis)respecting versus (dis)liking: Status and
interdependence predict ambivalent stereotypes of competence and warmth. J. Soc. Issues 55(3) 473-489.
Fuegen, K., M. Biernat. 2002. Reexamining the effects of solo status for women and men. Personality
Soc. Psych. Bull. 28(7) 913-925.
Greenwald, A.G., M.R. Banaji. 1995. Implicit social cognition: Attitudes, self-esteem, and stereotypes.
Psych. Rev. 102(1) 4-27.
Hackman, J.R., R. Wageman. 2005. A theory of team coaching. Acad. Management Rev. 30(2) 269-287.
Duo Status and Stereotyping 28
Hamilton, D., S. Sherman. 1996. Perceiving persons and groups. Psych. Rev. 103(2) 336-354.
Harrison, D.A., K.H. Price, M.P. Bell. 1998. Beyond relational demography: Time and the effects of
surface- and deep-level diversity on work group cohesion. Acad. Management J. 41(1) 96-107.
Heilman, M.E. 2001. Description and prescription: How gender stereotypes prevent women's ascent up
the organizational ladder. J. Soc. Issues 57(4) 657-674.
Heilman, M.E., C.J. Block, R.F. Martell. 1995. Sex stereotypes: Do they influence perceptions of
managers? J. Soc. Behav. Personality 10(6) 237-252.
Hilton, J.L., W. von Hippel. 1996. Stereotypes. Annual Review of Psychology 47 237-271.
Insko, C., J. Schopler, J. Kennedy, K. Dahl, K. Graetz, S. Drigotas. 1992. Individual-group discontinuity
from the differing perspectives of Campbell's realistic group conflict theory and Tajfel and Turner's social
identity theory. Social Psychology Quarterly 55(3) 272-291.
Inzlicht, M., T. Ben-Zeev. 2000. A threatening intellectual environment: Why females are susceptible to
experiencing problem-solving deficits in the presence of males. Psych. Sci. 11(5) 365-371.
Kanter, R.M. 1977. Men and women of the corporation. Basic Books, New York.
Karakowsky, L., J.P. Siegel. 1999. The effects of proportional representation and gender orientation of
the task on emergent leadership behavior in mixed-gender work groups. J. Appl. Psych. 84(4) 620-631.
Klauer, K.C., I. Wegener, K. Ehrenberg. 2002. Perceiving minority members as individuals: the effects of
relative group size in social categorization. Eur. J. Soc. Psych. 32(2) 223-245.
Koenig, A.M., J.A. Richeson. 2010. The contextual endorsement of sexblind versus sexaware ideologies.
Soc. Psych. 41(3) 186-191.
Konrad, A., V. Kramer. 2006. How many women do boards need. Harvard Bus. Rev. 84 12-22.
Konrad, A., V. Kramer, S. Erkut. 2008. Critical mass: The impact of three or more women on corporate
boards. Organ. Dynamics 37(2) 145-164.
Laband, D., B. Lentz. 1998. The effects of sexual harassment on job satisfaction, earnings, and turnover
among female lawyers. Ind. Lab. Relat. Rev. 51(4) 594-607.
Lortie-Lussier, M., N. Rinfret. 2002. The proportion of women managers: Where is the critical mass? J.
Appl. Soc. Psych. 32(9) 1974-1991.
Loyd, D., J. White, M. Kern. 2008. Duo status: Disentangling the complex interactions within a minority
of two. K.W. Phillips, ed. Research on Managing Groups and Teams Volume 11: Diversity and Groups.
Emerald Group Bingley, UK, 75-92.
Marques, J.M., V.Y. Yzerbyt. 1988. The black sheep effect: Judgmental extremity towards ingroup
members in inter- and intra-group situations. Eur. J. Soc. Psych. 18(3) 287-292.
McLeod, P.L., R.S. Baron, M.W. Marti, K. Yoon. 1997. The eyes have it: Minority influence in face-to-
face and computer-mediated group discussion. J. Appl. Psych. 82(5) 706-718.
Duo Status and Stereotyping 29
Menon, T., K. Phillips. in press. Getting even or being at odds? Cohesion in even-and odd-sized small
groups. Organ. Sci.
Monin, B., D.T. Miller. 2001. Moral credentials and the expression of prejudice. J. Personality Soc.
Psych. 81(1) 33-43.
Mullen, B., C. Copper. 1994. The Relation Between Group Cohesiveness and Performance: An
Integration. Psych. Bull. 115(2) 210-227.
O'Leary, M.B., M. Mortensen. 2010. Go (con)figure: Subgroups, imbalance, and isolates in
geographically dispersed teams. Organ. Sci. 21(1) 115-131.
O'Reilly, C., D. Caldwell, W. Barnett. 1989. Work group demography, social integration, and turnover.
Admin. Sci. Quart. 34(1) 21-37.
Paolacci, G., J. Chandler, P. Ipeirotis. 2010. Running experiments on Amazon Mechanical Turk.
Judgment and Decision Making 5(5) 411-419.
Pazy, A., I. Oron. 2001. Sex proportion and performance evaluation among high-ranking military officers.
J. Organ. Behav. 22(6) 689-702.
Randel, A.E. 2002. Identity salience: A moderator of the relationship between group gender composition
and work group conflict. J. Organ. Behav. 23(6) 749-766.
Reddy, W.B., O. Kroeger. 1972. Intergroup model-building: The Lego Man. J.W. Pfeiffer, J.E. Jones, eds.
1972 Annual Handbook for Group Facilitators. University Associates, San Diego, 36-42.
Rudenstine, D. Summer 2000. An interview with Associate Justice Sandra Day O'Connor Cardozo Life,
Rudman, L.A., A.G. Greenwald, D.E. McGhee. 2001. Implicit self-concept and evaluative implicit gender
stereotypes: Self and ingroup share desirable traits. Personality Soc. Psych. Bull. 27(9) 1164-1178.
Sackett, P.R., C.L. DuBois, A.W. Noe. 1991. Tokenism in performance evaluation: The effects of work
group representation on male-female and White-Black differences in performance ratings. J. Appl. Psych.
Schein, V.E. 1973. Relationship between sex role stereotypes and requisite management characteristics. J.
Appl. Psych. 57(2) 95-100.
Sherman, M. August 2010. Three women on Supreme Court: How big a difference? Associated Press.
http://www.msnbc.msn.com/id/38606618/ns/politics-supreme_court Accessed November 10 2010.
Simmel, G., K. Wolff. 1950. The sociology of Georg Simmel. Free Press, Glencoe, IL.
Smith, K., K. Smith, J. Olian, H. Sims Jr, D. O'Bannon, J. Scully. 1994. Top management team
demography and process: The role of social integration and communication. Admin. Sci. Quart. 39(3).
Spangler, E., M.A. Gordon, R.M. Pipkin. 1978. Token women: An empirical test of Kanter's hypothesis.
Amer. J. Sociology 84(1) 160-169.
Duo Status and Stereotyping 30
Taylor, S.E., S.T. Fiske, N.L. Etcoff, A.J. Ruderman. 1978. Categorical and contextual bases of person
memory and stereotyping. J. Personality Soc. Psych. 36(7) 778-793.
Uhlmann, E.L., G.L. Cohen. 2007. "I think it, therefore it's true": Effects of self-perceived objectivity on
hiring discrimination. Organ. Behav. Human Decision Processes 104(2) 207-223.
Duo Status and Stereotyping 31
Table 1. Means, Standard Deviations, Correlations of Study 1 Variables
Mean SD 1 2 3 4 5 6 7 8 9 10 11
1 Age 30.75 10.67
2 Non-native speaker 1.06 .23 -.07
3 Minority 1.20 .40 -.20**
4 Total errors .23 .57 -.11 .06 .15*
5 Educational level 4.06 1.43 .29***
-.06 -.11 -.21**
6 Employed 2.13 .90 .28***
-.01 -.09 -.04 .36***
7 Student 1.27 .45 -.50***
.19* .15* .09 -.28***
8 Number of women 1.99 .81 -.04 .00 .03 -.03 .05 .03 -.03
9 Group size 8.88 3.40 .01 .15* .01 .09 -.02 -.03 .03 .02
10 Gender warmth 2.64 3.12 .08 .03 .09 -.03 -.10 .08 -.07 -.03 -.10
11 Gender potency 4.21 2.35 .16* .06 .02 -.09 .12 .01 -.05 -.13* .03 .03
Note. Non-native speaker (1 = no, 2 = yes); Minority (1 = no, 2 = yes); Student (1 = no, 2 = yes).
*p < .05. **p < .01. *** p < .001.
Duo Status and Stereotyping 32
Table 2. Means, Standard Deviations, Correlations of Study 2 Variables
Mean SD 1 2 3 4 5 6 7 8 9 10 11 12
1 Age 21.51 .96
2 International 1.12 .30 .12
3 Minority 1.19 .39 .07 -.13
4 Session 1.47 .51
5 Group size 7.38 .60 .25 .24 .00 .68***
6 Number of women 1.88 .81 .10 -.31 .03 .14 .16
7 Attempts 1.32 .60 -.29 .25 -.13 .24 .28 .07
8 Completion time 5.01 2.98 .02 .33 -.10 -.09 .12 -.15 .27
9 Leadership 6.82 1.39 -.21 -.15 -.17 .16 .03 .04 -.27 -.25
10 Skills 7.02 1.01 -.13 .12 -.02 .13 .05 -.05 -.34 -.31 .68***
11 Effort 7.70 .93 -.29 -.10 -.13 -.10 -.20 -.05 -.27 -.30 .76***
12 Cohesion 5.80 .63 -.08 -.17 -.06 -.26 -.22 .12 -.60***
13 Stereotyped 2.52 1.82 -.07 .35* .04 -.16 -.06 -.16 .50**
-.39* -.53** -.46**
Note. International (1 = no, 2 = yes); Minority (1 = no, 2 = yes); Session (1, 2).
*p < .05. **p < .01. *** p < .001.
Duo Status and Stereotyping 33
ACME Management Team (3 women, 10 member condition)