ArticlePDF Available

Gender Dichotomization at the Level of Ingroup Identity: What It Is, and Why Men Use It More Than Women

Authors:

Abstract and Figures

In 5 studies (N = 756), we show that men's relative to women's gender ingroup identities are characterized by greater levels of gender dichotomization, a tendency to distance masculine from feminine traits. We demonstrate further that men's gender dichotomization is motivated, in part, by a need to eschew femininity from their ingroup identity to bolster a precarious gender status. Studies 1-3 establish and replicate the basic effect, and rule out alternative explanations (positivity, projection, status striving) for men's tendency to dichotomize more than women. Studies 4 and 5 demonstrate the motivated nature of gender dichotomization by establishing that men, but not women, dichotomize more strenuously when reminded of the precariousness of their gender status, and report stronger motivation to restore their gender status upon learning that their ingroup is becoming less dichotomized. Across 3 studies, strength of identification with their gender group moderates men's dichotomization tendencies. Discussion considers the implications of these findings for understanding the precarious nature of manhood and identifies practical applications of gender dichotomization in the interpersonal realm. (PsycINFO Database Record (c) 2013 APA, all rights reserved).
Content may be subject to copyright.
Gender Dichotomization at the Level of Ingroup Identity:
What It Is, and Why Men Use It More Than Women
Jennifer K. Bosson and Kenneth S. Michniewicz
University of South Florida
In 5 studies (N756), we show that men’s relative to women’s gender ingroup identities are
characterized by greater levels of gender dichotomization, a tendency to distance masculine from
feminine traits. We demonstrate further that men’s gender dichotomization is motivated, in part, by a
need to eschew femininity from their ingroup identity to bolster a precarious gender status. Studies 1–3
establish and replicate the basic effect, and rule out alternative explanations (positivity, projection, status
striving) for men’s tendency to dichotomize more than women. Studies 4 and 5 demonstrate the
motivated nature of gender dichotomization by establishing that men, but not women, dichotomize more
strenuously when reminded of the precariousness of their gender status, and report stronger motivation
to restore their gender status upon learning that their ingroup is becoming less dichotomized. Across 3
studies, strength of identification with their gender group moderates men’s dichotomization tendencies.
Discussion considers the implications of these findings for understanding the precarious nature of
manhood and identifies practical applications of gender dichotomization in the interpersonal realm.
Keywords: ingroup identity, gender dichotomization, precarious manhood, male role norms
“A woman simply is, but a man must become. Masculinity is
risky and elusive. It is achieved by a revolt from woman” (Camille
Paglia, 1992, p. 82).
The notion that masculinity “is achieved by a revolt from
woman” lies at the core of several gender role theories. Some
theorists claim that the antifemininity mandate—the rule that boys
and men must avoid feminine behaviors, tendencies, and prefer-
ences—is the most pervasive and salient norm of the male gender
role (Thompson, Grisanti, & Pleck, 1985). Here, we consider how
the antifemininity mandate shapes men’s ingroup identity, or their
beliefs about the traits associated with men.
1
We propose that men
eschew femininity from their ingroup identity as a means of
protecting a relatively precarious, easy-to-lose gender status (Van-
dello, Bosson, Cohen, Burnaford, & Weaver, 2008). In contrast,
because girls and women have less need than boys and men to
protect their relatively assured gender status, women should be less
inclined to reject masculinity from their ingroup identity.
The purposes of this research are threefold. First, we introduce
the construct of gender dichotomization and establish that men’s,
relative to women’s, ingroup identity is characterized by more
pronounced levels of it. We define gender dichotomization as the
extent to which one views same-gender-typical traits as more
central to her or his gender group than other-gender-typical traits.
Moreover, we propose that men dichotomize feminine and mas-
culine traits in their ingroup identity more strenuously than women
do, at least partly as a means of distancing the group “men” from
traits reflecting femininity, and thereby coping with a relatively
precarious gender status. Consistent with this, dichotomization
tendencies should be especially high among men who identify
more strongly with their gender group. Second, we rule out several
alternative explanations for men’s gender dichotomization tenden-
cies. Third, we establish the motivated nature of gender dichoto-
mization by demonstrating that men, but not women, endorse more
dichotomized ingroup identities when reminded of the precarious-
ness of their gender status and report stronger motivation to restore
their gender status following information that their ingroup iden-
tity is becoming less dichotomized (i.e., increasing in other-
gender-typical traits).
The issue of gender dichotomization in ingroup identities has
implications for understanding a variety of self- and group-level
phenomena. Ingroup identities, like personal identities, constitute
an important part of the self-concept (Brewer & Gardner, 1996;
Swann & Bosson, 2010). According to self-categorization theory,
people’s ingroup identities (i.e., beliefs about the qualities that
1
Ingroup identities are beliefs about a group held by its own members
(e.g., Gómez, Selye, Huici, & Swann, 2009;Swann & Bosson, 2010). In
this sense, they are comparable to ingroup stereotypes. We prefer the term
ingroup identity over ingroup stereotype, however, because the former
more clearly conveys our assumption that beliefs about one’s ingroup are
part of the self-concept.
This article was published Online First June 10, 2013.
Jennifer K. Bosson and Kenneth S. Michniewicz, Department of Psy-
chology, University of South Florida.
We thank Jamie Goldenberg and Joe Vandello for their helpful com-
ments on an earlier draft of this manuscript. We also thank Harlee Good-
less, Andrew Krajewski, Daniel Rynn, and Serena Yeager for their assis-
tance coding essays.
Correspondence concerning this article should be addressed to Jennifer
K. Bosson, Department of Psychology, PCD 4118G, University of South
Florida, 4202 East Fowler Avenue, Tampa, FL 33620-7200. E-mail:
jbosson@usf.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Personality and Social Psychology © 2013 American Psychological Association
2013, Vol. 105, No. 3, 425–442 0022-3514/13/$12.00 DOI: 10.1037/a0033126
425
characterize their ingroups) exert a powerful influence on their
personal identities (i.e., beliefs about the qualities that define them
as individuals), and reminders of group memberships can elicit
behavioral conformity to group norms (e.g., Jetten, Spears, &
Manstead, 1996;Turner, Hogg, Oakes, Reicher, & Wetherell,
1987). However, the specific contents of ingroup identities and
personal identities need not always overlap: People may hold
stereotypical beliefs about their groups while perceiving their own
stable self-views in less stereotypical terms (Biernat, Vescio, &
Green, 1996). Yet even when personal and ingroup identities do
not match, people will work to confirm their ingroup identities
(Gómez, Selye, Huici, & Swann, 2009). For example, high school
students from Madrid who were highly certain of their Madrilenian
ingroup identity preferred to interact with outgroup members who
shared their beliefs about Madrilenians’ traits, even when these
group-relevant traits were inconsistent with the students’ personal
identities (Gómez et al., 2009). These students also evaluated
outgroup members who verified their ingroup identities as more
competent than those who did not, and they felt better understood
by the former outgroup members. Thus, ingroup identities act as
self-guides that shape behavior, cognition, and affect, just as
personal identities do.
In what follows, we define gender dichotomization in detail and
clarify its links to men’s precarious gender status. Building from
the literatures on gender role norms and social identity, we then
make a case for the motivated nature of gender dichotomization.
Finally, we distinguish our proposed mechanism—men’s need to
eschew femininity—from other mechanisms before outlining our
hypotheses.
Gender Dichotomization and Precarious Manhood
Gender dichotomization is the extent to which one views her or
his gender group as possessing same-gender-typical traits and
lacking other-gender-typical traits. For example, a man high in
gender dichotomization might view independent and assertive
(stereotypically masculine traits) as very central to the group
“men” and warm and nurturing (stereotypically feminine traits) as
very peripheral to the group “men.” The more one dichotomizes
masculine and feminine traits in one’s ingroup identity, the more
psychological distance one perceives between masculine and fem-
inine aspects of the identity.
Stereotypes consist of traits and qualities that are strongly as-
sociated with given social groups (Brewer, Dull, & Lui, 1981;
Cantor & Mischel, 1977). Moreover, gender stereotypes tend to
fall along a bipolar dimension such that persons high in masculine
traits are assumed to be low in feminine ones (Deaux & Lewis,
1984;Foushee, Helmreich, & Spence, 1979). Thus, it is unsurpris-
ing to propose that people view gender-typical traits as more
central to their ingroup identity than gender-atypical traits. What is
novel is our suggestion that men, on average, exhibit more gender
dichotomization in their ingroup identity than women do. We
propose that this gender difference in dichotomization stems from
and reflects men’s continual need to protect and bolster a gender
status that is relatively easy to lose.
According to work on precarious manhood, cultures around the
world conceptualize manhood (relative to womanhood) as an
achieved rather than ascribed social status that is difficult to earn
but easy to lose (Bosson & Vandello, 2011;Vandello et al., 2008).
In many cultures, manhood status—the social recognition that one
is a “real man”—must be earned via the passage of difficult,
dangerous, and physically risky manhood rituals (Gilmore, 1990).
Conversely, the same requirements of proof rarely exist for wom-
anhood status. Although women are judged negatively for failing
to meet cultural standards of femininity, their status as “real
women” is not questioned as frequently as is men’s status as “real
men.”
Even in cultures that do not prescribe formalized manhood
rituals, people generally conceptualize manhood as a status that is
difficult to achieve and easy to lose. For example, U.S. college
students agreed with, understood, and liked statements about pre-
carious manhood (e.g., “Manhood is hard won and easily lost”)
more than comparable statements about precarious womanhood
(Vandello et al., 2008). They also characterized the transition from
boyhood to manhood as requiring more social achievements than
the transition from girlhood to womanhood. And when confronted
with statements about lost manhood or womanhood (e.g., “I used
to be a man [woman]; now I am no longer a man [woman]”),
participants generated more social than physical reasons for lost
manhood (e.g., “He lost his job”), whereas they generated more
physical than social reasons for lost womanhood (e.g., “She had a
hysterectomy”).
The implications of occupying a precarious gender status are
numerous. When men are reminded of the precariousness of their
gender, they exhibit heightened anxiety and physically aggressive
cognitions (Vandello et al., 2008), more aggressive posturing
(Bosson, Vandello, Burnaford, Weaver, & Wasti, 2009), riskier
financial decisions (Weaver, Vandello, & Bosson, 2013), and
heightened cortisol reactivity (Caswell, Bosson, Vandello, & Sell-
ers, 2013). More generally, men appear especially sensitive to the
requirements of proof and action that accompany their gender
status. For example, men more than women interpret other men’s
aggression—when elicited by gender threats—as situationally
rather than dispositionally caused (Weaver, Vandello, Bosson, &
Burnaford, 2010). Men also overestimate the blow to their gender
status that will result from job loss and its concomitant inability to
provide for family (Michniewicz, Vandello, & Bosson, 2013).
Here, we propose that men’s tendency to dichotomize masculine
and feminine traits in their ingroup identity both reflects their
chronic need to uphold a precarious gender status and bolsters their
manhood status when it has been challenged. To explicate our
logic, we turn to research on the antifemininity mandate.
The Antifemininity Mandate
According to theories of hegemonic masculinity (e.g., Herek,
1986;Kimmel, 1997), one of the core directives of the male gender
role is the antifemininity mandate: Boys and men must avoid
femininity in their behavior, personality, appearance, and interests.
Whereas both men and women receive pressure to avoid acting
like the other gender, feminine male targets are evaluated more
negatively than are masculine female targets (e.g., Feinman, 1981;
Levy, Taylor, & Gelman, 1995). In childhood, boys receive more
parental encouragement than girls for performing gender-typical
behavior (Lytton & Romney, 1991), and boys, as compared with
girls, who exhibit gender-atypical behaviors receive harsher criti-
cism from peers (Fagot, 1977). In adulthood, feminine men receive
more negative evaluations (McCreary, 1994) and are perceived as
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
426 BOSSON AND MICHNIEWICZ
more “different” from the rater (Sirin, McCreary, & Mahalik,
2004) than masculine women. Thus, men generally experience
powerful pressures to eschew femininity. In fact, publicly display-
ing femininity is one of the means by which men can lose their
manhood status in others’ eyes. Stereotypically feminine acts such
as styling hair and talking about emotions constitute gender threats
for men and, as such, raise men’s concerns about their status as
“real men” (Bosson, Prewitt-Freilino, & Taylor, 2005;Bosson,
Taylor, & Prewitt-Freilino, 2006).
Given this pressure to avoid femininity, gender dichotomization
at the level of the ingroup identity may buffer men from uncer-
tainty about how to achieve and maintain manhood. According to
self-categorization theory (Hogg & Abrams, 1993;Turner et al.,
1987), people’s ingroup identities quell uncertainty, in part, by
offering guidelines for self-definition and behavior in ambiguous
situations. Under conditions of uncertainty or stress, people cling
more fiercely to their group memberships as important sources of
meaning (Grieve & Hogg, 1999;Mullin & Hogg, 1999). Finally,
as noted earlier, people who are highly certain of their ingroup
identities will work to protect these identities because of the
meaning and coherence they afford (Gómez et al., 2009). In
general, then, people’s beliefs about their ingroups offer guidance
and information that promote a sense of meaning and predictabil-
ity. More specifically, men’s ingroup identities may offer a cog-
nitive roadmap for protecting and restoring threatened manhood.
When their gender status is under threat, men’s ingroup identities
may provide both prescriptions (e.g., “act rational”) and proscrip-
tions (e.g., “don’t act sensitive”) for reasserting manhood. Thus, to
the extent that men more often than women experience uncertainty
about their gender status, they should also maintain a relatively
large distance between masculine and feminine traits at the level of
the ingroup identity.
Alternatively, gender differences in people’s tendency to dichot-
omize masculine from feminine traits may reflect prior learning
experiences. After all, if boys learn from an early age that they
must avoid femininity, then perhaps they develop the belief that
men, on average, are quite high in masculine traits and quite low
in feminine ones. Cognitive learning processes associated with the
adoption of cultural stereotypes most likely explain much of the
variance in people’s gender ingroup identities (e.g., Bem, 1981;
Biernat, 1991). However, if gender dichotomization also serves the
motivational purpose for men that we propose— quelling uncer-
tainty about gender status—then dichotomization should fluctuate
in response to challenges to men’s gender status. Reminders of the
precariousness of their gender status should temporarily heighten
men’s, but not women’s, need to eschew other-gender-typical
traits from their ingroup identity. Moreover, information that their
ingroup identity is becoming less dichotomized (i.e., increasing in
other-gender-typical traits) should motivate men, but not women,
to pursue activities that will restore their gender status. These
hypotheses, which we test in Studies 4 and 5, cannot be explained
solely by cognitive learning accounts of stereotyping, and instead
require a motivational mechanism.
Strength of Gender Identification
Thus far, we have posited a mean gender difference in dichot-
omization tendencies. Logically, however, there should also exist
within-group differences in men’s dichotomization. The more
strongly people identify with a given group, the more they invest
psychologically in it. For instance, those who identify more
strongly with a given group also rely more heavily on the group as
a source of shared reality, stress reduction, and well-being (Has-
lam, O’Brien, Jetten, Vormedal, & Penna, 2005;Haslam &
Reicher, 2006). Here, we treat strength of gender identification as
a moderator of the link between participant gender and dichoto-
mization in several studies. By gender identification, we mean the
importance or centrality of gender group membership to one’s
overall self-concept (referred to as centrality by Leach et al., 2008,
and identity by Luhtanen & Crocker, 1992). We predict that men
who are more deeply invested in their gender group will exhibit a
stronger tendency to eschew femininity from their mental repre-
sentations of men, because these men care more about both main-
taining their own gender status and protecting the identity of the
group as a whole (Doosje & Ellemers, 1997). In contrast, women’s
gender identification may predict dichotomization less reliably
than men’s given that women experience relatively low pressure to
eschew masculinity from their ingroup identities.
Alternative Possible Mechanisms Underlying
Dichotomization
We realize that motivations other than a desire to eschew
femininity may drive men’s gender dichotomization. Here, we
detail several of these possible alternatives and explain how we
address them.
Projection
People tend to project their own traits onto others (e.g., Cadinu
& Rothbart, 1996;Heider, 1958), especially when the others being
evaluated are ingroup members (Robbins & Krueger, 2005)or
similar to the self (Ames, 2004). It is thus possible that men’s
tendency to dichotomize in their ingroup identities simply reflects
their tendency to perceive themselves as possessing high levels of
masculine traits and low levels of feminine traits (Bem, 1974;
Spence & Helmreich, 1978). To address this possibility, we mea-
sure people’s personal identities on the same traits used to assess
ingroup identities, and we control for these in Study 1.
We do not necessarily expect men to dichotomize their mascu-
line and feminine personal identities to the same degree that they
dichotomize their ingroup identities. Why not? After all, if people
incorporate the qualities associated with valued groups into the
self-concept (Turner et al., 1987), then men might endorse mas-
culine and reject feminine personal identities just as they endorse
masculine and reject feminine ingroup identities. However, as
noted earlier, there are limits to the extent to which people self-
stereotype, or embrace group-relevant traits as self-descriptive
(Biernat et al., 1996), and men, on average, self-stereotype less
strongly and consistently than women do (Latrofa, Vaes, Cadinu,
& Carnaghi, 2010). This may be because gender is (at least in
some contexts) a less salient and accessible category for men than
it is for women (Cadinu & Galdi, 2012;Casper & Rothermund,
2012). Moreover, personal identities typically comprise more in-
dividuated, less inclusive sets of traits than ingroup identities
(Brewer & Gardner, 1996), and they derive largely from self-
relevant feedback and other self-evaluative experiences (e.g., Pel-
ham, 1995;Pelham & Swann, 1989). Social reality may not allow
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
427
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
men to maintain the same level of gender dichotomization in their
stable, personal identities that they can maintain in their ingroup
identities, which derive not from personal experience but from
cognitive representations of the group prototype (Turner et al.,
1987). We examine this issue in Study 1.
Social Status Striving
Social status (i.e., dominance, power, and influence over others)
is more central to the male than the female gender role. Across
time and cultures, gender stereotypes associate boys and men more
strongly than girls and women with traits that convey social status,
such as assertiveness, dominance, and action (e.g., Ashmore, Del
Boca, & Wohlers, 1986;Spence & Helmreich, 1978;Williams &
Best, 1990). Gender stereotypes that link men with social status are
not merely descriptive; they are prescriptive as well (Prentice &
Carranza, 2002). Thus, boys and men learn that they ought to
display high-status traits and behaviors (Rudman, Moss-Racusin,
Phelan, & Nauts, 2012). Not surprisingly, men are more motivated
than women to pursue social status (Buss, 1999;Hoyenga, 1993).
For instance, men tend to display a sociopolitical orientation—
social dominance orientation characterized by the endorsement
of group-based social inequities and a desire for one’s ingroups to
dominate outgroups (Dambrun, Duarte, & Guimond, 2004;Pratto,
Stallworth, & Sidanius, 1997). As such, being a “real man” means,
to some degree, being high in social status. If so, men’s gender
dichotomization in their ingroup identity might reflect a desire to
claim, on behalf of the ingroup, traits associated with high social
status and to dissociate from traits associated with low social
status.
Although gender status and social status are confounded for
men, however, they are not synonymous. The male gender role
encapsulates numerous traits and behaviors that are nondiagnostic
of social status such as chivalry,honesty, and boisterousness, for
example. Moreover, eschewing femininity does not entail eschew-
ing only low-status traits and behaviors. Feminine traits such as
sensitive to others,cooperative, and polite are viewed as both
prescriptive for women and status neutral (Rudman et al., 2012).
Thus, embracing masculinity and avoiding femininity involves
embracing and avoiding a wider array of traits than merely those
that connote high and low social status, respectively. We expect
men to assiduously dichotomize masculine and feminine traits—
and thereby protect their ingroup identity from femininity—re-
gardless of both the perceived social status associated with mas-
culine and feminine traits and men’s attitudes toward social status.
We test this hypothesis in Study 2.
Positivity
Dichotomizing masculine and feminine traits in one’s ingroup
identity might reflect a desire for positivity (e.g., Alicke &
Sedikides, 2009;Tajfel & Turner, 1986), particularly if one per-
ceives same-gender-typical traits as favorable and other-gender-
typical traits as unfavorable. However, to the extent that unfavor-
able traits are highly central to an ingroup identity, people who
identify strongly with that group tend to seek feedback and inter-
action partners that confirm and verify these negative traits (Chen,
Chen, & Shaw, 2004;Gómez et al., 2009). Moreover, if gender
dichotomization serves the function of eschewing femininity, then
men should dichotomize along negative as well as positive mas-
culine and feminine traits. In the studies reported here, we address
this possibility by using sets of positive and negative traits that are
matched for valence, and we examine dichotomization tendencies
separately by trait valence.
The Present Studies
We propose that men’s motivation to eschew femininity mani-
fests in men’s, compared with women’s, greater tendency to di-
chotomize masculine and feminine traits in their ingroup identity.
In the present article, we detail three correlational studies and two
quasi-experiments that test this idea and its implications. In Studies
1–3, we present men and women with lists of gender-typed traits
and ask them to indicate how central each trait is to their gender
group. We expect men relative to women to perceive a larger
distance between masculine and feminine traits in their ingroup
identity, regardless of trait valence. We also expect men who
identify more strongly with their gender to dichotomize more
strenuously (Studies 1, 2, and 4), and we rule out competing
explanations for the gender dichotomization effect, including sta-
tus striving and projection. In Study 4, we evince the motivational
component of gender dichotomization by reminding some partic-
ipants of a time when they lost their gender status. We expect
men’s, but not women’s, dichotomization tendencies to increase
following this reminder because gender status loss activates, for
men, the need to eschew other-gender-typical traits. In Study 5, we
take the opposite approach and threaten some people’s ingroup
identities by telling them that their gender group is increasing in
other-gender-typical traits. We expect this information to heighten
men’s, but not women’s, motivation to demonstrate their gender
status via gender-typical behaviors.
Study 1
The purpose of Study 1 was to establish the basic phenomenon
by demonstrating that men exhibit more gender dichotomization in
their ingroup identities than women do. Moreover, we addressed
the alternative possibilities that gender differences in dichotomi-
zation reflect positivity striving and projection. Participants rated
the centrality to their gender group of 20 traits that varied on
dimensions of genderedness (feminine, masculine) and desirability
(positive, negative). We hypothesized that men would perceive a
larger distance between same-gender-typical and other-gender-
typical traits than women and that this difference would emerge on
both positive and negative traits. We also predicted an interaction
of gender and gender identification strength such that identifica-
tion should predict dichotomization more strongly among men
than women. Finally, we expected these effects to emerge even
when controlling for personal identities on the same 20 traits.
In this and later studies, our measure of gender dichotomization
derives from ratings of 20 gender-typed traits (see Table 1). Of
these, 18 are from Williams and Best’s (1990) cross-cultural study
of gender stereotypes, and two were added to represent updated
versions of William and Best’s original traits (graceful instead of
poised and rowdy instead of disorderly). Williams and Best quan-
tified each trait in terms of femininity/masculinity and desirability;
we used their data (along with pilot ratings of the two new traits)
to identify sets of feminine negative, feminine positive, masculine
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
428 BOSSON AND MICHNIEWICZ
negative, and masculine positive traits matched on genderedness
and desirability. Specifically, the feminine traits were as feminine
(M87.6, SD 9.02) as the masculine traits were masculine
(M86.5, SD 5.95), paired samples t(9) 1, and the positive
traits (M98.9, SD 26.94) were as positive as the negative ones
(M⫽⫺110.0, SD 37.25) were negative, paired samples t(9) 1.
Method
Participants and procedure. One hundred eighty-six under-
graduate students (56% women) from a large southeastern univer-
sity completed the measures described below online in exchange
for partial course credit. Participants’ ages ranged from 18 to 54
(Mdn 20 years), and they described themselves as White
(56.5%), Black (17.7%), Latino/a (15.6%), Asian (6.5%), and
“other” (3.7%). Participants completed all measures as well as
several additional scales not relevant to this study, in the order
listed; the 20 gendered traits always appeared in the same ran-
domly determined order.
Measures.
Gender identification strength. We adapted the four-item
Identity subscale of Luhtanen and Crocker’s (1992) Collective
Self-Esteem scale so that items referred to gender (e.g., “Being a
woman [man] is an important part of my self-image”). To these we
added five items adapted from Branscombe, Kobrynowicz, and
Owen’s (1996) Gender Group Identification scale (e.g., “My gen-
der is central to my identity,” “I am ashamed to be a woman
[man]”). All items were rated on scales ranging from 1 (strongly
disagree)to7(strongly agree). After reverse coding negatively
worded items, we averaged all items (␣⫽.88).
Personal identities. On 19-point scales ranging from 5% (way
below average) to 95% (way above average), participants rated
themselves, compared with other members of their gender, on the
20 traits. We computed composites for feminine positive (␣⫽
.70), masculine positive (␣⫽.67), feminine negative (␣⫽.77),
and masculine negative (␣⫽.77) traits.
Ingroup identities. Written instructions explained that “some
qualities are very central or typical to some social groups, whereas
other qualities are less central. . .. For each trait below, please
indicate how central that trait is to your gender group.” Participants
rated the 20 traits on scales ranging from 1 (not at all central)to
9(extremely central). We created feminine positive, masculine
positive, feminine negative, and masculine negative trait compos-
ites by averaging the five traits in each category (s.84).
Results
Gender dichotomization. We submitted participants’ in-
group identity scores to a 2 (gender: female, male) 2 (trait type:
feminine, masculine) 2 (trait valence: positive, negative) anal-
ysis of variance (ANOVA), with repeated measures on the last two
factors. This revealed main effects of gender, F(1, 184) 20.28,
p.001, f.33; trait type, F(1, 184) 30.79, p.001, f.41;
and trait valence, F(1, 184) 104.34, p.001, f.75, as well
as interactions of Gender Trait Type, F(1, 184) 428.36, p
.001, f1.53, and Gender Trait Valence, F(1, 184) 13.15,
Table 1
Ratings of Traits on Genderedness and Desirability
Positive traits Negative traits
Trait set Trait
Genderedness
(Desirability) Trait
Genderedness
(Desirability)
Studies 1, 2, and 4
Masculine Assertive 85 (541) Egotistical 84 (389)
Capable 73 (626) Loud 85 (428)
Independent 94 (612) Rowdy 93 (391)
Rational 90 (591) Rude 89 (342)
Stable 88 (585) Show-off 84 (384)
Feminine Affectionate 95 (611) Complaining 89 (359)
Graceful 97 (588) Dependent 90 (463)
Sensitive 89 (592) Moody 69 (407)
Sympathetic 89 (603) Nagging 98 (344)
Warm 78 (640) Worrying 82 (393)
Studies 3 and 5
Masculine Adventurous 2.45 (4.09) Arrogant 2.28 (1.67)
Competitive 2.37 (3.62) Boastful 2.24 (2.08)
Daring 2.38 (3.73) Coarse 2.35 (2.07)
Enterprising 2.51 (4.22) Reckless 2.23 (1.91)
Feminine Appreciative 3.43 (4.55) Fussy 3.67 (2.01)
Emotionally Expressive 4.13 (3.82) Insecure 3.55 (1.89)
Enthusiastic 3.33 (4.39) Melodramatic 4.08 (1.94)
Humble 3.32 (4.37) Weak 3.52 (1.91)
Note. First set of traits: Genderedness and desirability ratings for all traits (other than Graceful and Rowdy) are from Williams and Best (1990); ratings
for Graceful and Rowdy are based on pilot testing and were calculated following the procedures described in Williams and Best. Genderedness reflects the
percentage of respondents who rated masculine traits as more masculine than feminine, and feminine traits as more feminine than masculine. Desirability
ratings were made on 5-point scales and then converted to standard scores, with a mean of 500 and a standard deviation of 100; thus, scores above 500
are favorable and those below 500 are unfavorable. Second set of traits: Genderedness and desirability ratings are based on pilot testing and were made
on 5-point scales; ratings of feminine traits and unfavorable traits were reverse coded before ttests were conducted.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
429
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
p.001, f.27.
2
These effects were further qualified by a
three-way interaction of gender, trait type, and valence, F(1,
184) 4.28, p.05, f.15. As shown in Figure 1, both men and
women rated same-gender-typical traits as more central to their
ingroup identity than other-gender-typical traits, thereby driving
the significant three-way interaction. Most important for our pur-
poses, the very large Gender Trait Type interaction effect
showed that men displayed much more gender dichotomization
than women did: Regardless of trait valence, men (M
diff
2.46 for
positive traits, 2.53 for negative traits) perceived larger distances
between masculine and feminine traits in their ingroup identity
than women did (M
diff
1.27 for positive traits, 1.61 for negative
traits).
Moderation by gender identification strength. We expected
gender identification strength to moderate the link between gender
and dichotomization. To simplify the conduct and presentation of
analyses, we created separate positive and negative dichotomiza-
tion scores for each gender by subtracting positive and negative
other-gender-typical traits from positive and negative same-
gender-typical traits, respectively. In separate analyses, we re-
gressed these dichotomization indices onto gender (dummy
coded), gender identification strength (centered on zero), and the
interaction term. The model predicting positive dichotomization
produced a main effect of gender (␤⫽⫺0.44), t(182) ⫽⫺6.59,
p.001, f
2
.24, that was qualified by a Gender Strength
interaction (␤⫽⫺0.20), t(182) ⫽⫺2.36, p.02, f
2
.03. As
shown in Figure 2a, men who identified more strongly with their
gender also dichotomized more strenuously on positive traits (␤⫽
0.36, p.001, f
2
.09), whereas identification was unrelated to
dichotomization among women (␤⫽0.04, p.67). In the model
predicting negative dichotomization, there was a main effect of
gender (␤⫽⫺0.32), t(182) ⫽⫺4.61, p.001, f
2
.12, and a
marginally significant Gender Strength interaction (␤⫽
0.17), t(182) ⫽⫺1.87, p.06, f
2
.02. Again, stronger
identification among men predicted a greater tendency to dichot-
omize negative traits (␤⫽0.29, p.01, f
2
.05) (see Figure 2b).
Among women, identification was unrelated to dichotomization
(␤⫽0.02, p.82).
Controlling for projection. Did men also dichotomize their
personal identities more strenuously than women? To answer this,
we submitted personal identity scores to a 2 (gender: female,
male) 2 (trait type: feminine, masculine) 2 (trait valence:
positive, negative) ANOVA, with repeated measures on the last
2
Following Cohen (1988), we report effect sizes using ffor ANOVA
output and f
2
for multiple regression output. For f, values of .10, .25, and
.40 correspond to small, medium, and large effects. For f
2
, values of .02,
.15, and .35 correspond to small, medium, and large effects. All fand f
2
effect sizes were obtained using G
Power (Faul, Erdfelder, Lang, & Bu-
chner, 2007).
Figure 1. Gender ingroup identity ratings on masculine and feminine,
positive and negative, traits. Y-error bars are the standard errors associated
with each condition mean.
Figure 2. Dichotomization on positive (top panel) and negative (bottom
panel) traits as a function of gender and gender identification strength.
Predicted values are estimated at one standard deviation from the mean
(M5.57, SD 1.03) on gender identification strength.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
430 BOSSON AND MICHNIEWICZ
two factors. Overall, people rated positive traits as more self-
descriptive than negative traits, F(1, 183) 611.04, p.001, f
1.83, and men endorsed all traits more than women did, F(1, 183)
6.28, p.02, f.18. A Trait Type Trait Valence interaction
indicated that people rated themselves higher on masculine than
feminine positive traits, whereas they rated themselves higher on
feminine than masculine negative traits, F(1, 183) 47.23, p
.001, f.51; moreover, a significant three-way interaction re-
vealed that this pattern was more pronounced among women than
men, F(1, 183) 6.47, p.02, f.19. No other effects were
significant (Fs1.98, ps.16), thus indicating that men’s
tendency to dichotomize masculine from feminine traits was not
evident at the level of their personal identities.
Nonetheless, to control for the role of projection in gender
dichotomization, we reran the two regression models described
previously but entered the four personal identity composites as a
group in Step 1; in Step 2 we entered gender, gender identification
strength, and the interaction term. In the analysis on positive
dichotomization, personal identities were not significant (F1),
but both gender (␤⫽⫺0.47), t(177) ⫽⫺6.97, p.01, f
2
.27,
and the Gender Strength interaction (␤⫽⫺0.22), t(177)
2.57, p.02, f
2
.04, were significant. As above, identification
predicted dichotomization among men (␤⫽0.36, p.001, f
2
.10) but not women (␤⫽0.01, p.91). In the analysis on
negative dichotomization, personal identities were significant, F(4,
179) 2.92, p.03, as were gender (␤⫽⫺0.39), t(176)
5.72, p.001, f
2
.19, and the Gender by Strength interaction
(␤⫽⫺0.22), t(177) ⫽⫺2.52, p.02, f
2
.04. Again, men
higher in identification also dichotomized more (␤⫽0.29, p
.01, f
2
.06), whereas identification was unrelated to dichotomi-
zation among women (␤⫽⫺0.05, p.61). Thus, although
personal identities accounted for some of the variance in dichot-
omization on negative traits, the main effect of gender and the
Gender Strength interaction still emerged when we controlled
for these identities.
3
Summary
Men as compared with women more strongly dichotomized
feminine from masculine traits in their ingroup identity, whether
the traits were positive or negative in valence. Although we found
evidence of positivity striving—with most respondents endorsing
positive traits more strongly than negative traits in their personal
and ingroup identities—the desire for positivity could not fully
explain men’s tendency to distance feminine and masculine as-
pects of their ingroup identity. Moreover, these findings cannot be
explained as mere projection of personal identities onto the group
given that personal identities did not follow the same pattern as
ingroup identities. Thus, we found no evidence that men maintain
the same level of gender dichotomization in their personal identi-
ties as they do in their ingroup identities. Finally, men who
identified more strongly with their gender group tended to dichot-
omize more (on both positive and negative traits) than men who
identified weakly with their gender group.
Study 2
Our primary goal in Study 2 was to replicate the findings from
Study 1 while controlling for various indices of social status.
Although we view the social status hypothesis as a reasonable
explanation for some of the between-group variance in men’s
versus women’s gender dichotomization tendencies, we expect to
find that men dichotomize more than women do even when con-
trolling for individual differences in their motivation to pursue
social status and their perceptions of the social status associated
with each of the 20 traits. That is, we view men’s need to eschew
femininity from their ingroup identity as broader than a need
merely to avoid low-status traits. As in Study 1, men should
perceive a larger distance between feminine and masculine traits in
their ingroup identities than women, and this difference should
emerge on both positive and negative traits. Moreover, an inter-
action of gender and gender identification strength should reveal
that identification predicts dichotomization tendencies more
strongly among men than women. Finally, these effects should
emerge when controlling for all of the social status measures.
Method
Participants. Two hundred twenty-four college students at a
large university in the southeastern United States completed the
study materials online in exchange for course credit. We deleted
the data from 14 participants whose duration times (9 min)
indicated that they did not give careful consideration to their
responses (mean completion time for all measures was 17.35 min).
The final sample of 210 participants (56% women; Mdn 20
years) described themselves as White (58.1%), Latino/a (14.3%),
Black (13.3%), Asian (5.7%), and “other” (8.6%).
Procedure. Participants completed the following measures, in
the same random order, during one of two online sessions. Mea-
sures of gender identification strength and social dominance ori-
entation were completed during an online mass-testing session,
and the remaining measures were completed during a separate
online session that included several other scales not relevant here.
Measures.
Gender identification strength. We assessed strength of gen-
der identification with the same scale used in Study 1 (␣⫽.87).
Ingroup identities. Participants rated the centrality to their
gender of the 20 traits from Study 1. We created composites of
positive feminine, positive masculine, negative feminine, and neg-
ative masculine traits (s.81).
Status ratings of traits. Participants indicated the extent to
which each of the 20 traits was associated with social status,
defined as “a person’s social influence or power over others.”
Traits were rated on scales ranging from 1 (strongly associated
with low status)to9(strongly associated with high status), and we
3
Given stereotypes that Black women are less feminine than White
women (e.g., Collins, 2004), we examined race differences in gender
dichotomization at the level of ingroup identity. Looking only at those
racial/ethnic groups for whom we had enough data (Whites, Blacks,
Latinos/as), we found a main effect of race/ethnicity, F(2, 161) 8.15,
p.001, such that Blacks (M1.28, SE 0.22) dichotomized signifi-
cantly less than both Whites (M2.88, SE 0.12) and Latinos/as (M
2.04, SE 0.23; p.03), the latter two of whom did not differ (p.87).
Among Whites, men dichotomized more than women on both positive and
negative traits (Fs23.75, ps.001); among Latinos/as, men dichoto-
mized more than women on positive traits (p.04), but not on negative
traits (p.64); and among Blacks, men dichotomized marginally signif-
icantly more than women on positive traits (p.075), but not negative
ones (p.23).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
431
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
averaged them to obtain positive feminine (␣⫽.75), positive
masculine (␣⫽.81), negative feminine (␣⫽.70), and negative
masculine (␣⫽.66) status composites.
Social dominance orientation. We used the 16-item Social
Dominance Orientation Scale (SDO; Pratto, Sidanius, Stallworth,
& Malle, 1994) to measure people’s preferences for group-based
social hierarchies. Sample items include, “It’s OK if some groups
have more of a chance in life than others” and “No one group
should dominate society.” Respondents indicate how each state-
ment makes them feel on a scale ranging from 1 (very negative)to
7(very positive). After reverse coding half of the statements so that
higher ratings indicate more positive attitudes toward hierarchies,
we averaged the items (␣⫽.94).
Desire for social status. We wrote six items to assess partici-
pants’ personal desires for status: “I seek out positions of high social
status,” “It is important to me to belong to social groups that have
power relative to other groups,” “I prefer to belong to social groups
that are high in status,” “I do not care whether the social groups to
which I belong have more status than other social groups,” “I tend to
avoid social status,” and “It is not important to me that my social
groups have power in society.” Items were rated on scales ranging
from 1 (strongly disagree)to7(strongly agree). After reverse coding
the latter three items, we averaged them (␣⫽.87).
Results
Not surprisingly, masculine traits (M6.04) were associated with
higher social status than feminine traits (M4.72), F(1, 208)
258.56, p.001, f1.11. As shown in Table 2, four of the social
status variables correlated marginally or significantly with at least one
of the dichotomization indices. The SDO correlated with desire for
social status (r.34, p.001), but neither of these measures
correlated with status ratings of the traits (average r.08).
Gender dichotomization. We hypothesized that men would
exhibit more gender dichotomization in their ingroup identities than
women, regardless of their standing on status-relevant variables and
their perceptions of the status associated with each trait. To simplify
the conduct and presentation of analyses, we followed the procedures
described in Study 1 for creating separate positive and negative
dichotomization scores for each gender. Specifically, we subtracted
positive and negative other-gender-typical traits from positive and
negative same-gender-typical traits, respectively. Before proceeding
to our primary analyses, we first examined whether either of the
status-relevant individual-difference variables—SDO or desire for
social status—moderated any of our effects. No significant effects
emerged (ps.12), thereby indicating that it was appropriate to treat
these measures as covariates.
In the first analysis of covariance (ANCOVA), we treated gender as
the independent variable, positive dichotomization scores as the de-
pendent variable, and the set of four relevant control variables (SDO,
status of positive masculine traits, status of positive feminine traits,
desire for social status) as covariates. Of the covariates, status of
positive masculine traits reached significance, F(1, 202) 21.52, p
.001, and status of positive feminine traits was marginally significant,
F(1, 202) 3.53, p.07. The remaining covariates were not
significant (ps.10). Most importantly, as shown on the left-hand
side of Figure 3, men dichotomized more than women did when
rating the centrality to their ingroup identity of positive traits, F(1,
202) 44.22, p.001, f.47, after controlling for the covariates.
In the second ANCOVA, we treated gender as the independent
variable, negative dichotomization scores as the dependent variable,
and the set of relevant social status variables as covariates. In this
analysis, no covariates approached significance (all Fs2.31, ps
.13). As shown on the right-hand side of Figure 3, after controlling for
the covariates, men dichotomized more than women did when rating
the centrality to their ingroup identity of negative traits, F(1, 202)
4.83, p.03, f.15.
Moderation by gender identification strength. We next
tested whether gender identification strength moderated the relation
between gender and dichotomization. To do this, we regressed the
positive and negative dichotomization scores separately onto the set of
relevant status variables in Step 1, and gender (dummy coded),
zero-centered gender identification strength scores, and the interaction
term in Step 2. In the model predicting positive dichotomization, the
set of social status variables was significant, F(4, 203) 6.07, p
.001, f
2
.12. In Step 2, there was a main effect of gender (␤⫽
0.45), t(200) ⫽⫺7.07, p.001, f
2
.25, that was qualified by a
Gender Strength interaction (␤⫽⫺0.19), t(200) ⫽⫺2.14, p
.04, f
2
.02. Among men, identification predicted dichotomization of
positive traits (␤⫽0.25, p.01, f
2
.04); among women, there was
no association between identification and dichotomization (␤⫽
0.02, p.85). In the model predicting negative dichotomization,
the set of social status variables did not reach significance, F(4,
203) 1.04, p.38. In Step 2, there was a significant main effect of
gender (␤⫽⫺0.22), t(200) ⫽⫺3.09, p.01, f
2
.05, but the
Gender Strength interaction did not reach significance (␤⫽
0.10, t1). Nonetheless, the pattern was consistent with that
obtained on positive dichotomization scores: Men’s gender iden-
tification predicted dichotomization of negative traits (␤⫽0.29,
p.01, f
2
.04), whereas women’s gender identification did not
(␤⫽0.15, p.12).
4
4
As in Study 1, we examined the role of race/ethnicity (White, Black,
Latino/a) in dichotomization tendencies. This time, there was no main
effect of race, F(2, 174) 1.66, p.19, although the gender comparisons
yielded a similar pattern to that found in Study 1: Among Whites, men
dichotomized more than women on both positive and negative traits (Fs
6.06, ps.02); among Latinos/as, men dichotomized more than women on
positive traits (p.01), but not negative traits (p.39); and among
Blacks, men dichotomized more than women on positive traits (p.01),
but not negative traits (p.51). We did not conduct analyses looking at
race/ethnicity in subsequent studies due to insufficient numbers of non-
Whites in our samples.
Table 2
Correlations of Status Variables With Gender Dichotomization
Indices, Split by Participant Gender (Study 2)
Gender dichotomization
Positive traits Negative traits
Status variable Women Men Women Men
Social dominance orientation .09 .21
.16
.02
Desire for social status .09 .15 .08 .00
Status positive feminine .02 .19
.08 .15
Status negative feminine .15 .13 .02 .05
Status positive masculine .27
ⴱⴱ
.31
ⴱⴱ
.10 .24
Status negative masculine .06 .08 .04 .19
p.10.
p.05.
ⴱⴱ
p.01.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
432 BOSSON AND MICHNIEWICZ
Summary
Although people’s motivations for and perceptions of social
status accounted for some of the variance in their gender dichot-
omization tendencies, the predicted effect of gender on dichoto-
mization still emerged when we controlled for all of the social
status-relevant variables. Thus, men’s greater dichotomization ten-
dencies cannot be explained by their relatively strong motivation
to pursue and claim social status, or by their tendency to perceive
masculine traits as higher in social status than feminine traits.
Finally, men who identified more strongly with their gender also
tended to dichotomize more on both positive and negative traits.
Study 3
Our primary aim in Study 3 was to replicate the basic gender
dichotomization effect using a different set of traits. Although the
traits used in Studies 1 and 2 were normed across cultures, these
ratings were published more than 30 years ago. To replicate our
effect using more recently normed traits, we piloted 68 traits culled
from contemporary works on gender stereotypes (Guimond et al.,
2007;Prentice & Carranza, 2002;Rudman et al., 2012). Pilot
participants (37 men, 42 women) rated each trait, in a different
random order, on both genderedness (from 1 very masculine to
5very feminine) and desirability (from 1 very unfavorable to
5very favorable). On the basis of these ratings, we selected 16
traits that met the criteria laid out in Study 1: The masculine traits
were as masculine as the feminine ones were feminine, paired
samples t(77) 1, and the positive traits were as positive as the
negative ones were negative, paired samples t(77) 1. These 16
traits and their ratings are presented in Table 1.
A second goal of Study 3 was to examine whether men engage
in greater gender dichotomization than women when rating the
other gender group. In doing so, we addressed the possibility that
men simply perceive a larger psychological distance than women
do between masculine and feminine traits in their schemas for both
genders. Thus, we had participants rate their gender ingroup and
the other gender group on the same 16 traits. We predicted a
Gender Group interaction such that men should display more
dichotomization than women in their ingroup identity, but not in
their stereotypes of the other group. For purposes of brevity, we
did not consider gender identification strength in this study.
Method
Participants and procedure. Eighty-three participants
(50.6% women; Mdn
age
28), recruited through Amazon’s Me-
chanical Turk (MTurk) website, received $0.25 apiece for rating
the centrality to their own and the other gender group of 16 traits.
MTurk respondents tend to be more diverse on basic demographic
variables (age, race, country of origin) than are typical American
college samples, and the data they provide are similar in quality
and psychometric soundness to that obtained in college samples
(Buhrmester, Kwang, & Gosling, 2011). Participants described
themselves as White (80.5%), Asian (14.6%), Black (2.4%), La-
tino/a (1.2%), and “other” (1.2%).
Order of rating own versus other gender group was counterbal-
anced, and the 16 traits were presented in a different random order
for each participant. Centrality ratings were made on scales rang-
ing from 1 (not at all central to group)to9(very central to group),
and we averaged the traits to create eight 4-item composites
representing all levels of genderedness, favorability, and group (all
s.79).
Results and Summary
Following the procedures from Studies 1 and 2, we created
positive and negative dichotomization scores for each gender. We
also created scores that reflected the extent to which people di-
chotomized masculine and feminine traits in the other gender
group (other gender’s positive and negative gender-typical traits
minus other gender’s positive and negative gender-atypical traits).
We then entered these scores into a 2 (gender: male, female) 2
(trait valence: positive, negative) 2 (group: own, other)
ANOVA, with repeated measures on the last two factors. Results
revealed a main effect of group, F(1, 80) 4.11, p.05, f.23,
that was qualified by the expected Gender Group interaction,
F(1, 80) 56.88, p.001, f.84. No other effects emerged
(Fs1.52, ps.22). As predicted, men dichotomized more than
women did when rating their ingroup identity on both positive
(Ms2.75 vs. 1.53) and negative (Ms3.12 vs. 1.58) traits
(Fs13.74, ps.001, fs.41). Interestingly, however, men
dichotomized less than women did when rating the other group on
positive (Ms1.44 vs. 2.49) and negative (Ms1.42 vs. 2.36)
traits (Fs5.43, ps.03, fs.26).
These findings indicate that the gender dichotomization effect
observed in Studies 1 and 2 was not unique to the set of gendered
traits used in those studies. Indeed, the effect replicated using traits
that were normed in a contemporary sample of men and women.
Moreover, men’s tendency to dichotomize in their ingroup identity
cannot be explained as part of an overall tendency to perceive large
distances between masculinity and femininity in their gender sche-
mas in general; when it came to their stereotypes about women,
men dichotomized less than women did in their stereotypes about
men.
Figure 3. Men’s and women’s dichotomization on positive and negative
traits, controlling for status relevant variables. Y-error bars are the standard
errors associated with each condition mean.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
433
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
Study 4
Having established the basic phenomenon and ruled out several
alternative interpretations of it in Studies 1–3, we designed Study
4 to demonstrate the motivational nature of gender dichotomiza-
tion. We have argued that men are motivated to dichotomize
masculine from feminine aspects of their ingroup identity as a
means of coping with a precarious gender status. That is, men
eschew femininity from their ingroup identity more strongly than
women eschew masculinity because men experience greater un-
certainty than women do surrounding the stability of their gender
status, and gender dichotomization quells this uncertainty. If our
logic is sound, then we should observe that reminders of the
precariousness of their gender status temporarily heighten men’s
need to eschew femininity from their ingroup identity. Specifi-
cally, reminding men of a time when they lost their own manhood
status in other people’s eyes should cause them to dichotomize
their ingroup identity more strenuously. Thus, we propose a dy-
namic link between the integrity of men’s gender status on an
individual level (i.e., their presumed standing, at any moment, as
a “real man” in others’ eyes) and their eschewal of femininity on
the group level. A threat on one of these levels should elicit a
compensatory reaction on the other.
Men and women rated the strength of their gender identification,
wrote a brief essay that served as the gender status loss manipu-
lation, and then rated the centrality to their gender group of the 20
traits from Studies 1 and 2. We manipulated gender status loss by
asking some people to write about a time when they lost their
womanhood [manhood] “in other people’s eyes”; control partici-
pants wrote about time management. We expected a two-way
interaction of Gender Gender Status Loss, such that men should
dichotomize more than women in general, but especially so fol-
lowing a reminder of gender status loss. Given women’s relatively
more stable gender status and their weaker pressures to avoid
masculinity, we did not expect them to display compensatory
dichotomization following a reminder of gender status loss. We
also examined whether gender identification strength moderated
these effects. If men who identify more strongly with their gender
react particularly strongly to a reminder of manhood loss, then we
would expect these men to show the largest increase in gender
dichotomization following such a reminder.
Method
Participants and design. Participants were recruited through
MTurk and completed our survey online in exchange for $0.50.
We received 187 completed surveys; of these, we deleted data
from seven people who did not follow instructions (e.g., their
essays were uninterpretable). This left a final sample of 180
participants (50% women; Mdn 31 years) who described them-
selves as White (80.0%), Black (7.2%), Latino/a (5.0%), Asian
(4.4%), and “other” (3.4%). Participants were randomly assigned
to gender status loss condition in a 2 (gender: female, male) 2
(gender status loss: yes, no) design.
Procedure. Interested account holders in Amazon’s MTurk
followed a link to the study at the Surveygizmo website (www
.Surveygizmo.com). After giving their informed consent, partici-
pants completed the following measures and tasks in the order
described below.
Gender identification strength. We administered an abbrevi-
ated measure of gender identification strength consisting of the
five items from the scale used in Studies 1 and 2 with the highest
item-total correlations (rs.53). To minimize the salience of
gender, we embedded these items among items from a Big Five
personality scale (John & Srivastava, 1999). All items were rated
on scales ranging from 1 (strongly agree)to7(strongly agree),
and we averaged the five gender identification items after reverse
coding one that was negatively worded (␣⫽.87).
Writing task. Participants in the gender status loss condition
were asked to
please think about a time in your life when you did something (or
something happened to you)—in front of other people—that made
you feel bad about your status as a “real woman” [“real man”] and
perhaps even insecure as a member of your gender. This should be
something that would make most other people see you as a “not a real
woman” [“not a real man”].
We reasoned that this task would serve, for men, as a reminder of
the precariousness of their gender status. In control conditions,
participants were asked to “Please consider how you structure your
days. Then describe, in as much detail as possible, how activities,
chores, work, and family time are divided up during a typical day.”
In both writing conditions, participants were asked to spend at least
5 min on this task and to type their responses directly into a blank
text box.
Immediately following the writing task, participants in the threat
condition answered four questions on scales ranging from 1 (not at
all)to9(very much): “How masculine did you feel during the
experience you just described?”; “How feminine did you feel
during the experience you just described?”; “To what extent do the
events you described make you feel positive emotions?”; and “To
what extent do the events you described make you feel negative
emotions?” For purposes of comparison, participants in the control
condition also answered these latter two questions, although they
did not answer the first two because we did not want to prime
gender in the control condition.
Ingroup identities. Participants rated the centrality to their
gender group of the 20 traits from Studies 1 and 2, and we
averaged them to create positive feminine, positive masculine,
negative feminine, and negative masculine trait composites (s
.81).
Results
Manipulation checks. Among participants who wrote about
gender status loss, we compared men’s ratings of how “masculine”
they felt and women’s ratings of how “feminine” they felt with the
scale midpoint (5) using one-sample ttests. As intended, men
(M3.11) felt less masculine than the scale midpoint, t(44)
5.02, p.001, d1.51, and women (M3.52) felt less
feminine than the scale midpoint, t(44) ⫽⫺4.53, p.001, d
1.37. The control writing task elicited less negative emotion (M2.96)
than the gender status loss writing task (M6.13), t(178)
9.91, p.001, d1.49, and it elicited more positive emotion
(M6.80) than the gender status loss writing task (M3.54),
t(178) 10.05, p.001, d1.51. To rule out concerns that the
effects reported below were driven by condition or gender differ-
ences in mood, we used positive and negative affect as covariates
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
434 BOSSON AND MICHNIEWICZ
in follow-up analyses. We also controlled for word count of
essays, as an index of effort.
Two judges who were naïve to our hypotheses also read and
coded 16 randomly selected essays from each of the four condi-
tions (for a total of 64 essays). On 7-point scales, judges rated “to
what extent are the behaviors, activities, cognitions, or emotions in
the essay feminine/masculine?” Judges also rated “the extent to
which the writer describes a gender role violation” on a scale
ranging from 1 (no violation)to7(a clear violation). Judges
showed good interrater reliability (ICCs .72), so we averaged
their ratings. Men’s gender status loss (vs. control) essays con-
tained more masculine content (Ms1.88 and 1.16; d0.96),
more feminine content (Ms3.03 and 1.28; d1.76), and more
evidence of gender role violations (Ms4.63 and 1.31; d3.61;
all Fs5.36, ps.03). Women’s gender status loss (vs. control)
essays contained more masculine content (Ms2.69 and 1.28;
d1.34), more feminine content (Ms2.69 and 1.59; d0.93),
and more evidence of gender role violations (Ms4.47 and 1.06;
d5.50; all Fs6.16, ps.02). Finally, men’s and women’s
essays described similar degrees of gender role violation in both
the gender status loss (t1) and control (t1.24, p.22)
conditions. Thus, although both men’s and women’s gender status
loss essays contained more gender-typical content than control
essays, they also contained more gender-atypical content and much
stronger evidence of gender role violations. The writing instruc-
tions thus appeared to have their intended effects.
Gender dichotomization. We computed positive and nega-
tive dichotomization scores, as in previous studies, and submitted
them to a 2 (gender: female, male) 2 (gender status loss: yes, no)
2 (trait valence: positive, negative) ANOVA, with repeated
measures on the last factor. This analysis revealed a main effect of
gender, F(1, 176) 26.10, p.001, f.38, that was qualified
by the predicted Gender Gender Status Loss interaction, F(1,
176) 4.48, p.04, f.16. Moreover, although there was a
significant Gender Trait Valence interaction, F(1, 176) 47.25,
p.001, f.52, the three-way interaction did not approach
significance (F1). This indicates that the Gender Gender
Status Loss interaction pattern did not differ as a function of trait
valence. Therefore, we collapsed across valence to yield the means
presented in Figure 4. The Gender Gender Status Loss interac-
tion was driven by a tendency for men to dichotomize their
ingroup identity more strenuously in the status loss condition than
in the control condition, F(1, 176) 4.39, p.04, f.16,
whereas women’s tendency to dichotomize was unaffected by a
gender status loss reminder (F1). Men, moreover, dichotomized
more than women did in both the gender status loss and control
conditions (Fs4.52, ps.04, fs.16). We conducted a
follow-up ANCOVA, with positive affect, negative affect, and
word count as covariates. None of the covariates reached signifi-
cance (all Fs1.30, ps.25), and the gender main effect, F(1,
172) 28.58, p.001, f.41, and Gender Gender Status
Loss interaction, F(1, 172) 5.85, p.02, f.18, both remained
significant.
Moderation by gender identification strength. We next
tested whether gender identification strength further moderated the
Gender Gender Status Loss interaction. To do this, we regressed
positive and negative dichotomization scores onto gender (dummy
coded), gender status loss condition (dummy coded), zero-centered
gender identification strength, all two-way interactions, and the
three-way interaction term. In the model predicting positive di-
chotomization, there was a main effect of gender (␤⫽0.38),
t(172) 4.10, p.001, f
2
.10; a Gender Gender Status Loss
interaction (␤⫽0.24), t(172) 2.11, p.04, f
2
.03; and a
three-way interaction (␤⫽⫺0.22), t(172) ⫽⫺2.04, p.05, f
2
.02. This three-way interaction remained significant when we
entered positive affect, negative affect, and word count as control
variables (␤⫽⫺0.39), t(168) ⫽⫺2.61, p.02, f
2
.04.
To deconstruct the three-way interaction, we examined the
Gender Status Loss Strength interactions separately among men
and women. Among men, the Status Loss Strength interaction
approached significance (␤⫽⫺0.19), t(172) ⫽⫺1.74, p.08,
f
2
.02; among women, it did not (␤⫽0.19), t(172) 1.27, p
.21. Simple slope tests revealed that strength of gender identifica-
tion did not predict positive dichotomization among men or
women in either condition (all s0.22, ps.14). Instead, and
contrary to prediction, the interaction pattern was driven by an
increase in dichotomization among weakly identified men in the
gender status loss condition relative to the control condition (␤⫽
0.35, p.01, f
2
.06). Conversely, writing about a loss of
gender status did not increase dichotomization tendencies among
strongly identified men (␤⫽⫺0.06, p.66) or among women
(s.23, ps.18). In the model predicting negative dichoto-
mization, the three-way interaction was not significant (␤⫽
0.04, t1, p.81).
Summary
Consistent with our motivational account of gender dichotomi-
zation, men’s tendency to eschew feminine traits from their in-
group identity increased following a reminder of the precarious-
ness of their gender status. Conversely, a reminder of gender status
loss had no effect on women’s tendency to eschew masculine traits
from their ingroup identity. This pattern is consistent with our
logic that men dichotomize in their ingroup identity to quell
uncertainty about their own precarious gender status: As uncer-
tainty about their gender status increases, so does men’s need to
Figure 4. Men’s and women’s dichotomization as a function of experi-
mental condition. Y-error bars are the standard errors associated with each
condition mean.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
435
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
eschew femininity from their ingroup identity. We view this effect
as analogous to other commonly observed compensatory threat
effects, such as the phenomenon whereby reminding people of
their own (future) death causes them to cling more vociferously to
cultural stereotypes for the anxiety-buffering protection they pro-
vide (e.g., Schimel et al., 1999).
The findings with gender identification strength were complex.
Reminders of a loss of gender status exaggerated the already-
present gender differences in dichotomization of positive traits, but
most of the movement was among men who identified weakly with
their gender group. Neither strongly identified men nor women,
regardless of their gender identification strength, showed an in-
crease in dichotomization when reminded of a gender status loss.
On negative traits, however, both strongly and weakly identified
men increased their dichotomization following the gender status
loss reminder. One possibility is that strongly identified men are
chronically high in gender dichotomization on positive traits and
thus have little room for additional increases, even when reminded
of their precarious gender status. We return to this finding in the
General Discussion.
Study 5
Study 4 established that challenges to men’s manhood elicited a
compensatory increase in their eschewal of feminine traits from
their ingroup identity. In Study 5, we examine this process from
the opposite angle: We challenge men’s ingroup identity and
examine their compensatory efforts to demonstrate manhood on a
personal level. If men eschew femininity from their ingroup iden-
tity as a means of protecting their precarious gender status, then
information that their group is becoming more (relative to less)
feminine should temporarily increase men’s anxiety about their
own manhood status. Thus, this particular group-level threat (e.g.,
“men are becoming more feminine”) should have a similar effect
on men’s gender status concerns as does a person-level manhood
threat, because it reminds men of the ease with which their gender
status can be lost. As noted earlier, men often react to challenges
to their personal gender status with public, compensatory actions
such as aggressive posturing (Bosson et al., 2009) and risky
financial decisions (Weaver et al., 2013). We therefore expected
men to express greater motivation to enact a series of manhood-
restoring actions following information that their gender ingroup is
becoming less dichotomized. Doing so should allow men to cope
with their gender status anxiety by demonstrating their own “real
man” status in the face of uncertainty. Thus, we view men’s
intentions to prove manhood via gender-typical behaviors as a
means of coping with the anxiety elicited by a decrease in dichot-
omization of their ingroup identity. Conversely, information that
their group is becoming more (relative to less) masculine should
have comparatively little effect on women’s anxiety about their
gender status and ensuing motivation to prove their womanhood.
Women and men read a bogus article from the journal Science
that detailed their own gender’s standing on masculine and femi-
nine traits. The article described women [men] as remaining con-
stant across time in same-gender-typical traits, but as either in-
creasing, decreasing, or remaining stable in other-gender-typical
traits. After reading the article, participants rated their likelihood
of performing several masculine and feminine behaviors to dem-
onstrate their womanhood [manhood]. We expected a Gender
Article interaction, such that men (but not women) should become
more likely to perform gender-typical behaviors following infor-
mation that their ingroup identity is becoming less dichotomized,
and they should become less likely to perform such behaviors upon
learning that their ingroup identity is becoming more dichoto-
mized. Note that in this study, we did not address the issue of
gender identification strength as a moderator.
Method
Participants and design. One hundred six participants were
recruited through MTurk and completed the survey online in
exchange for $0.50. We deleted data from nine participants who
failed a manipulation check, leaving 97 participants (51.5% wom-
en; Mdn 31.5 years) in the final sample. Participants described
themselves as White (74.2%), Black (4.1%), Latino/a (8.2%),
Asian (9.3%), and “other” (4.1%). We randomly assigned partic-
ipants to article condition in a 2 (gender: female, male) 3
(article: less dichotomized, more dichotomized, control) design.
Procedure. Interested account holders in Amazon’s MTurk
followed a link to the study at the Qualtrics website (www.qual-
trics.com). After reading that the study was about news, memory,
and judgment, and giving their informed consent, participants
completed the following measures and tasks in the order described
below.
Science article. Participants read one of three fictitious arti-
cles, ostensibly “in press” at the journal Science, that described a
longitudinal, cross-cultural analysis of their own gender’s traits.
Each article began by providing several examples of positive and
negative masculine and feminine traits (taken from the list used in
Study 3; see Table 1) and briefly describing the methods of a
large-scale, cross-cultural study that measured over 18,000 wom-
en’s [men’s] traits across 10 decades. All article versions stated
that the participant’s gender group had remained stable and fairly
high across time on same-gender-typical traits, such that between
60% and 70% of all women [men] displayed feminine [masculine]
traits at each decade from the 1920s through 2010. Information
about the group’s standing on other-gender-typical traits differed
by condition. In less dichotomized conditions, women (for exam-
ple) learned that women were low on masculine traits in the 1920s
(with about 25% of women displaying masculine traits), but had
increased steadily over time until 2010, at which point 60% of
women displayed masculine traits. In more dichotomized condi-
tions, women learned that 60% of women displayed masculine
traits in the 1920s, but only 25% of women displayed masculine
traits in 2010. In control conditions, women learned that women
had remained stable and somewhat low on masculine traits (about
40% of women displayed masculine traits) from the 1920s through
2010. A large figure at the bottom of the page graphed these trends
across 10 decades. For men, all article details were identical but
reversed (referring to men’s changes in feminine traits across
time).
5
Gender-typical behaviors. Participants read and rated a list of
18 behaviors, half of which were masculine and half feminine. We
selected sets of comparable masculine and feminine behaviors that
would be familiar and easy to envision and that ranged in terms of
how effectively they would allow a person to restore his or her
5
All study materials are available from the first author.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
436 BOSSON AND MICHNIEWICZ
gender status. Pilot testing confirmed that the behaviors were
stereotypically masculine and feminine, respectively, and that both
sets of behaviors varied (from “somewhat” to “very”) in how
effectively they demonstrated gender status to onlookers. Mascu-
line behaviors included taking out the garbage,playing video
games,going for a run,shaving your face,mowing the lawn,
checking fantasy football league stats,picking a fight at a bar,
playing basketball with men from the neighborhood, and lifting
weights. Feminine behaviors included washing dishes,shopping
for clothes online,doing yoga,shaving your legs,planting flowers,
checking caloric intake,gossiping about someone,attending a
book club meeting with women from the neighborhood, and doing
aerobics.
Participants rated each behavior on two dimensions. The order
in which they made these ratings was counterbalanced and ran-
domly assigned, and the behaviors appeared in a different random
order for each participant. Our dependent measure was partici-
pants’ motivation to perform the behaviors to restore their gender
status. Written instructions asked participants to
please imagine that you did something that made other people ques-
tion your status as “a real woman [man].” If you could do any of the
following behaviors to restore your womanhood [manhood], how
likely is it that you would do each one?
Ratings were made on scales ranging from 1 (Very unlikely)to9
(Very likely), and we averaged them to create separate motivation
scores for each gender (s.91). Participants also rated, on scales
ranging from 1 (Not at all)to9(Very much), how effectively each
behavior “would allow a man [woman] to restore his manhood [her
womanhood],” and we averaged these (s.93).
Manipulation check. Participants first selected one of six
different statements that reflected the article’s main conclusion
(e.g., “women are becoming LESS masculine over time”); nine
participants were deleted for failing this manipulation check. Par-
ticipants also rated the article’s conclusion on a continuous scale
ranging from 1 (women [men] are becoming MORE masculine
[feminine] over time)to5(women [men] are becoming LESS
masculine [feminine] over time) with a midpoint of 3 (women
[men] have remained stable in their traits over time).
Results
Manipulation check. We submitted ratings of the article’s
conclusion to a 2 (gender: female, male) 3 (article: less dichot-
omized, more dichotomized, control) ANOVA. There was a large
main effect of article, F(2, 91) 401.57, p.001, f2.97, but
no effect of gender and no interaction (Fs1.31, ps.27).
Follow-up Tukey’s honestly significant difference tests indicated
that all three article conditions differed (ps.001): Participants in
the less dichotomized condition (M1.14, SE 0.08) rated their
gender group as becoming more masculine [feminine] than those
in the control condition (M3.00, SE 0.08), who rated their
gender group as remaining more stable than those in the more
dichotomized condition (M4.78, SE 0.10). Confident that
participants interpreted the article as intended, we proceeded to our
primary analyses.
Motivation to perform gender-typical behaviors. Did infor-
mation that their ingroup identity was becoming less dichotomized
(i.e., more feminine) heighten men’s motivation to enact manhood-
restoring behaviors? Before testing this, we submitted effective-
ness scores to a 2 (gender: female, male) 3 (article: less
dichotomized, more dichotomized, control) ANOVA to ensure that
the perceived effectiveness of the behaviors for restoring gender
status did not differ by condition. No effects emerged (Fs1).
We therefore submitted motivation scores to a 2 (gender) 3
(article) ANCOVA, with effectiveness ratings as a highly signifi-
cant covariate, F(1, 90) 123.35, p.001. This analysis yielded
the predicted Gender Article interaction, F(1, 90) 3.67, p
.03, f.28, and no other effects (Fs1).
As illustrated in Figure 5, among men, those who read that their
ingroup identity was becoming less dichotomized subsequently
reported stronger motivation to perform manhood-restoring behav-
iors, regardless of the behaviors’ presumed effectiveness, than did
men who read that their ingroup identity was becoming more
dichotomized (p.03, f.24). Men in the control condition did
not differ from men in either of the other two conditions (ps
.20). Among women, article condition had no effect on motivation
to perform womanhood-restoring behaviors (ps.12). Looking at
these data another way, men reported stronger motivation to per-
form gender-typed behaviors than women did after learning that
their ingroup identity was becoming less dichotomized (p.05,
f.21), but men reported marginally significantly weaker moti-
vation than women to perform these behaviors after learning that
their ingroup identity was becoming more dichotomized (p.06,
f.20). Men and women did not differ in the control condition
(p.72).
Removing the effectiveness covariate caused the Gender
Article interaction to drop to nonsignificance (p.24), but the
pairwise comparisons produced the same results: Men in the less
dichotomized condition (M5.64, SE 0.50) expressed greater
motivation to perform the manhood-restoring behaviors than those
in the more dichotomized condition (M4.09, SE 0.53; p
.04, f.23), and neither condition differed from the control
condition (ps.12). Among women, article condition was unre-
lated to motivation (ps.72).
Figure 5. Men’s and women’s motivation to perform gender-typed be-
haviors as a function of experimental condition. Y-error bars are the
standard errors associated with each condition mean.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
437
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
Summary
If maintaining a highly dichotomized ingroup identity buffers
men against uncertainty about their own gender status, then infor-
mation that their ingroup identity is becoming more feminine
should motivate men to action—specifically, manhood-restoring
action. Study 5 demonstrated this compensatory effect. Confronted
with compelling evidence that men, as a group, are becoming more
(vs. less) feminine over time, men reported a greater likelihood of
performing a range of behaviors that broadcast, to varying degrees,
“real man” status to onlookers. In contrast, women remained
relatively unaffected by information that their ingroup identity was
becoming more (vs. less) masculine over time.
Two points about these findings merit attention. First, the pre-
dicted pattern emerged among men regardless of their perceptions
of how effectively the behaviors would prove their manhood; thus,
the pattern cannot be explained by gender or article differences in
how useful participants thought it would be to perform the behav-
iors. Indeed, men’s motivation to pursue masculine behaviors in
the less dichotomized condition seemed driven by impulse rather
than reason, which is consistent with our logic that these reactions
reflected an attempt to cope with anxiety about their gender status.
That is, coping with gender status anxiety may motivate immediate
action, with little thought to its effectiveness. Second, the means in
the control condition—in which men learned that their ingroup
identity remained stably high in masculine traits and relatively low
in feminine ones— did not differ significantly from either of the
other two conditions. Thus, although the overall interaction effect
was medium in size, and the pattern of means among men was
consistent with predictions, it is somewhat surprising that the
control condition did not differ significantly from the less dichot-
omized condition. One possibility is that this study was underpow-
ered to detect such a difference. It is also possible, however, that
people do not have particularly strong or reliable reactions to
unchanging stimuli. Whereas information about increases or de-
creases in trait levels elicits strong reactions, information that
one’s traits have “stayed the same” may not evoke especially
meaningful responses. If so, perhaps a “no information” condition
would have served as a more useful control here.
General Discussion
This work documents the beginnings of a systematic examina-
tion of gender dichotomization by introducing this construct, high-
lighting its links to men’s precarious gender status, and establish-
ing its motivational properties. The results from five studies—
using both college student and nonstudent samples, as well as two
different sets of gendered traits—provide converging evidence that
men distance feminine from masculine aspects of their ingroup
identity more strenuously than women do and that they do so in
part to protect and maintain a gender status that is “hard won and
easily lost” (e.g., Vandello & Bosson, 2013).
We found in two studies that men who identify more strongly
with their gender also dichotomize their ingroup identity more
vociferously. Thus, men who are especially inclined to value and
derive meaning from their gender group membership are also more
inclined to eschew feminine traits, and embrace masculine traits, in
their ingroup identity. This effect emerged when we controlled for
the influence of variables including a desire for positivity, a
tendency to project personal identities onto groups, attitudes about
social status, and perceptions of the social status associated with
gendered traits. Men’s tendency to dichotomize their ingroup
identity more strenuously than women, moreover, could not be
explained by a general tendency to perceive gender groups as
highly dichotomized. Finally, a reminder of personal gender status
loss heightened men’s (but not women’s) gender dichotomization
at the level of their ingroup identity, and information that their
ingroup identity was becoming less dichotomized increased men’s
(but not women’s) motivation to restore their personal gender
status via action. Taken together, these findings suggest that men
experience unique concerns about their gender status that compel
them to eschew femininity at the level of their ingroup identity and
that a challenge to either of these—personal gender status or
gender ingroup identity— elicits a compensatory reaction in the
other.
This work adds to our understanding of men’s reactions to
gender-threatening experiences. According to work on precarious
manhood, the tenuousness of their gender status requires men
regularly to validate their masculinity via action, and this require-
ment is especially pronounced following gender threats (Vandello
et al., 2008). Whereas past investigations of men’s reactions to
gender threats focused on overt behaviors, however, the present
work documents a cognitive reaction. Cognitively representing
one’s ingroup identity as highly dichotomized may help men to
quickly and efficiently identify manhood-restoring traits and be-
haviors when the need arises. Consistent with this logic, we found
that men’s tendency to dichotomize at the level of ingroup identity,
although chronically higher than women’s, was heightened further
following a reminder of gender status loss. Thus, gender dichoto-
mization may serve as a cognitive roadmap that directs attention
quickly and effortlessly to the sorts of overt behaviors and traits
that will demonstrate and/or restore manhood. The tendency for
reminders of gender status loss to exaggerate perceptually the
distance between masculine and feminine aspects of the ingroup
identity may, in fact, be adaptive given that men become preoc-
cupied with self-conscious discomfort when reminded of the ten-
uousness of their manhood (Bosson et al., 2005;Vandello et al.,
2008).
Speaking more broadly, psychologically distancing one’s in-
group from qualities associated with a salient outgroup satisfies
important psychological needs. According to several theories, a
lack of certainty about the self is aversive (Brewer, 1991;Hogg,
2007). Optimal distinctiveness theory, for example, posits that
when people feel too deindividuated (e.g., similar to others), the
need for differentiation becomes aroused and motivates identifi-
cation with a more exclusive or distinct group (Brewer, 1991,
2003). Thus, one strategy for coping with an ill-defined sense of
self involves highlighting an ingroup’s distinctiveness from out-
groups. The clearly defined ingroup, then, facilitates categorizing
the self as a group member, which allays uncertainty by offering
consensual validation of reality (e.g., Hogg, 2007;Pickett &
Brewer, 2001). If the precariousness of manhood ensures that men,
relative to women, experience more frequent self-challenges in the
form of gender threats, then men’s relatively stronger dichotomi-
zation tendencies might reflect a chronic need to restore self-
certainty after such threats. Put another way, gender differences in
dichotomization tendencies may reflect a special case of optimal
distinctiveness needs wherein men’s “optimal” level is substan-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
438 BOSSON AND MICHNIEWICZ
tially higher than women’s with regard to their gender ingroup, but
not necessarily other groups to which they belong.
Limitations and Directions for Future Work
Although we collected data from both college student and
nonstudent samples, our participants tended to be predominantly
White and defined by a somewhat narrow age range. Future
research should thus investigate gender dichotomization tenden-
cies using more diverse samples. In particular, cross-cultural com-
parisons of gender differences in dichotomization might be illu-
minating. Although beliefs about gender and the content of gender
stereotypes show cross-cultural consistency (e.g., Gilmore, 1990;
Williams & Best, 1990), there are subtle cultural differences in
how these beliefs manifest at the level of the self-concept. For
example, men tend to define themselves in less extremely gender-
stereotyped terms in cultures that are lower in gender equality,
presumably because of differences in the social comparison pro-
cesses that fuel self-ratings: In cultures lower in gender equality
(i.e., non-Western cultures), cross-gender comparisons are consid-
ered inappropriate, and so men compare themselves primarily with
other men when rating the self on gendered traits (Guimond et al.,
2007). The result is a dampening of the tendency to perceive the
self as highly typical of one’s gender. Although this research
examined men’s personal identities and not their ingroup identi-
ties, it is possible that similar cultural differences might emerge at
the level of ingroup identities as well. That is, men from non-
Western versus Western cultures may view men in general as
relatively less dichotomized along gendered traits. Alternatively,
given that personal and ingroup identities derive from different
types of learning experiences (cf. Brewer & Gardner, 1996), in-
group identities may well be immune to the social comparison
processes that shape personal identities.
An examination of racial/ethnic differences in gender dichoto-
mization should also be illuminating. As described in Footnotes 3
and 4, we found less extreme gender differences in dichotomiza-
tion among Black and Latino/a respondents relative to White
respondents, and Blacks overall exhibited less gender dichotomi-
zation in their ingroup identities relative to other racial/ethnic
groups. For Black women, these relatively weaker dichotomization
tendencies might reflect cultural stereotypes that depict Black
women as less feminine and more masculine than White women
(e.g., Collins, 2004;Donovan, 2011). If ingroup identities derive
largely from cognitive representations of the group prototype, it
makes sense that Black women’s gender ingroup identities are
characterized by a relatively small psychological distance between
same-gender-typical and other-gender-typical traits. Consistent
with this notion, there is some evidence that Black women’s views
of womanhood contain more stereotypically masculine qualities
than White women’s (Settles, Pratt-Hyatt, & Buchanan, 2008).
What remain puzzling, however, are Black men’s relatively weak
dichotomization tendencies. Stereotypes of Black men often depict
them as hypermasculine (Collins, 2004), and some work suggests
that perceivers automatically equate “Blackness” with maleness
and masculinity (e.g., Goff, Thomas, & Jackson, 2008). Thus, why
Black men in our samples dichotomized their ingroup identities
less vigorously than White men, and only barely more vigorously
than Black women, is not clear. More work is needed to under-
stand how the intersection of race and gender shapes people’s
ingroup identities, as well as the compensatory links between
gender status and gender dichotomization at the level of ingroup
identity.
Future work should also clarify the links between gender iden-
tification strength and dichotomization tendencies. We found in
two correlational studies (Studies 1 and 2) that men who identified
more strongly with their gender group also dichotomized feminine
and masculine aspects of their ingroup identities more strenuously.
This makes sense because these men’s relatively strong investment
in their gender group should ensure that eschewing femininity
from the ingroup identity is a chronic concern for them. In Study
4, however, we manipulated gender status loss and observed
inconsistent effects of our manipulation on men who were weakly
versus strongly identified with their gender group: A reminder of
gender status loss heightened dichotomization tendencies among
both weakly and strongly identified men on negative traits, but it
only increased dichotomization among weakly identified men on
positive traits. As noted, it is possible that the dichotomization
tendencies of strongly identified men are already as extreme as
reality allows (at least on positive traits), thus rendering additional
dichotomization difficult. It is also possible, however, that the
three-way interaction obtained in Study 4 is unreliable and there-
fore requires replication before confident conclusions can be
drawn. After all, the strong situation used in Study 4 —reminders
of a time when gender status was lost—may have overwhelmed
individual differences in men’s chronic dichotomization tenden-
cies. Consistent with this interpretation, the Status Loss Strength
interaction among men was only marginally significant, thus rais-
ing questions about its reliability.
To complicate matters further, although gender identification
strength did not predict women’s dichotomization tendencies in
any of the studies reported here, it is not entirely clear why. Our
preferred explanation is that, for women, maintaining gender status
does not depend on eschewing masculinity from the ingroup
identity. It is possible, however, that the link between women’s
gender identification strength and their dichotomization tendencies
is more complex than we considered. Both traditional and nontra-
ditional women, for example, may invest equally strongly in their
gender group membership, but differ strikingly in their dichoto-
mization tendencies. Whereas strongly identified traditional
women may invest in binary conceptualizations of gender (and
thus dichotomize a lot), strongly identified nontraditional women
may instead hold androgynous views of gender (and thus dichot-
omize very little).
Given the unexpected and unreliable moderation pattern ob-
served among men in Study 4, and the interpretation difficulties
surrounding the null effects found for women, we hesitate to draw
firm conclusions about the role of gender identification strength in
dichotomization. One thing seems certain: In the absence of a
gender threat or salient reminder of gender status loss, men who
invest more deeply in their gender group also dichotomize their
ingroup identity more strenuously. Other conclusions, however,
will require additional research.
We end by considering two possible interpersonal applications
of this work. First, given that people prefer interaction and affili-
ation with those who verify their ingroup identities (Gómez et al.,
2009), our findings lead to the prediction of gender differences in
these verification processes. Men, whose ingroup identities are
characterized by more dichotomization than women’s, should pre-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
439
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
fer interaction partners who also perceive men in a highly dichot-
omized manner. In turn, increased affiliation with those who verify
a dichotomized view of men could suggest one route through
which men’s (often erroneous) beliefs about gender-normative
expectations are propagated. Men often display pluralistic igno-
rance about the extent to which others expect them to uphold
gender role norms (e.g., Felson, 1982;Michniewicz et al., 2013;
Vandello, Ransom, Hettinger, & Askew, 2009), and these errors
may be fueled by men’s exaggerated perceptions of most other
men’s masculinity. Future research might therefore examine
whether men indeed seek affiliation partners who hold dichoto-
mized stereotypes about men, and if so, whether this fuels men’s
erroneous beliefs about others’ gendered expectations.
Finally, men’s tendency to dichotomize at the level of their
ingroup identity may broaden our understanding of people’s dis-
approval of nonprototypical group members. For instance, the
tendency to eschew feminine traits from the ingroup identity may
reduce men’s tolerance of perceived femininity in other men,
especially under conditions of gender status threat. Consistent with
this reasoning, men whose gender status is threatened subsequently
evaluate effeminate and nonprototypical men more negatively
(Glick, Gangl, Gibb, Klumpner, & Weinberg, 2007;Schmitt &
Branscombe, 2001). From our perspective, this effect could be
driven by temporary changes in how men define their gender
ingroup identity. In a state of heightened gender dichotomization,
men may perceive highly feminine men as “black sheep”
(Marques, Yzerbyt, & Leyens, 1988), a perceptual shift that should
yield numerous consequences for interpersonal dynamics. We
view this potential application—the role of gender dichotomiza-
tion in men’s defining and policing of their gender group bound-
aries—as a particularly important and timely direction for future
work.
References
Alicke, M. D., & Sedikides, C. (2009). Self-enhancement and self-
protection: What they are and what they do. European Review of Social
Psychology, 20, 1– 48. doi:10.1080/10463280802613866
Ames, D. R. (2004). Strategies for social inference: A similarity contin-
gency model of projection and stereotyping in attribute prevalence
estimates. Journal of Personality and Social Psychology, 87, 573–585.
doi:10.1037/0022-3514.87.5.573
Ashmore, R. D., Del Boca, F. K., & Wohlers, A. J. (1986). Gender
stereotypes. In R. D. Ashmore & F. K. Del Boca (Eds.), The social
psychology of female–male relations (pp. 69 –119). Orlando, FL: Aca-
demic Press.
Bem, S. L. (1974). The measurement of psychological androgyny. Journal
of Consulting and Clinical Psychology, 42, 155–162. doi:10.1037/
h0036215
Bem, S. L. (1981). Gender schema theory: A cognitive account of sex
typing. Psychological Review, 88, 354 –364. doi:10.1037/0033-295X.88
.4.354
Biernat, M. (1991). Gender stereotypes and the relationship between mas-
culinity and femininity: A developmental analysis. Journal of Person-
ality and Social Psychology, 61, 351–365. doi:10.1037/0022-3514.61.3
.351
Biernat, M., Vescio, T. K., & Green, M. L. (1996). Selective self-
stereotyping. Journal of Personality and Social Psychology, 71, 1194 –
1209. doi:10.1037/0022-3514.71.6.1194
Bosson, J. K., Prewitt-Freilino, J. L., & Taylor, J. N. (2005). Role rigidity:
A problem of identity misclassification? Journal of Personality and
Social Psychology, 89, 552–565. doi:10.1037/0022-3514.89.4.552
Bosson, J. K., Taylor, J. N., & Prewitt-Freilino, J. L. (2006). Gender role
violations and identity misclassification: The roles of audience and actor
variables. Sex Roles, 55, 13–24. doi:10.1007/s11199-006-9056-5
Bosson, J. K., & Vandello, J. A. (2011). Precarious manhood and its links
to action and aggression. Current Directions in Psychological Science,
20, 82– 86. doi:10.1177/0963721411402669
Bosson, J. K., Vandello, J. A., Burnaford, R. M., Weaver, J. R., & Wasti,
S. (2009). Precarious manhood and displays of physical aggression.
Personality and Social Psychology Bulletin, 35, 623– 634. doi:10.1177/
0146167208331161
Branscombe, N. R., Kobrynowicz, D., & Owen, S. (1996, August). Gender
group identification: Implications for coping with prejudice and self-
esteem in women and men. Paper presented at the 104th Annual Con-
vention of the American Psychological Association, Toronto, Ontario,
Canada.
Brewer, M. B. (1991). The social self: On being the same and different at
the same time. Personality and Social Psychology Bulletin, 17, 475–
482. 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: Guilford Press.
Brewer, M. B., Dull, V., & Lui, L. (1981). Perceptions of the elderly:
Stereotypes as prototypes. Journal of Personality and Social Psychol-
ogy, 41, 656 – 670. doi:10.1037/0022-3514.41.4.656
Brewer, M. B., & Gardner, W. (1996). Who is this “we”? Levels of
collective identity and self representations. Journal of Personality and
Social Psychology, 71, 83–93. doi:10.1037/0022-3514.71.1.83
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s Mechan-
ical Turk: A new source of inexpensive, yet high-quality, data? Per-
spectives on Psychological Science, 6, 3–5. doi:10.1177/
1745691610393980
Buss, D. M. (1999). Human nature and individual differences: The evolu-
tion of human personality. In L. A. Pervin & O. P. John (Eds.), Hand-
book of personality: Theory and research (2nd ed., pp. 31–56). New
York, NY: Guilford Press.
Cadinu, M., & Galdi, S. (2012). Gender differences in implicit gender
self-categorization lead to stronger gender self-stereotyping by women
than by men. European Journal of Social Psychology, 42, 546 –551.
doi:10.1002/ejsp.1881
Cadinu, M., & Rothbart, M. (1996). Self-anchoring and differentiation
processes in the minimal group setting. Journal of Personality and
Social Psychology, 70, 661– 677. doi:10.1037/0022-3514.70.4.661
Cantor, N., & Mischel, W. (1977). Traits as prototypes: Effects on recog-
nition memory. Journal of Personality and Social Psychology, 35,
38 – 48. doi:10.1037/0022-3514.35.1.38
Casper, C., & Rothermund, K. (2012). Gender self-stereotyping is context
dependent for men but not for women. Basic and Applied Social Psy-
chology, 34, 434 – 442. doi:10.1080/01973533.2012.712014
Caswell, T. A., Bosson, J. K., Vandello, J. A., & Sellers, J. G. (2013). The
high cost of low status: Testosterone and men’s stress responses to
gender threats. Psychology of Men & Masculinity. Advance online
publication. doi:10.1037/a0031394
Chen, S., Chen, K. Y., & Shaw, L. (2004). Self-verification motives at the
collective level of self-definition. Journal of Personality and Social
Psychology, 86, 77–94. doi:10.1037/0022-3514.86.1.77
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Collins, P. H. (2004). Black sexual politics: African Americans, gender,
and the new racism. New York, NY: Routledge. doi:10.4324/
9780203309506
Dambrun, M., Duarte, S., & Guimond, S. (2004). Why are men more likely
to support group-based dominance than women? The mediating role of
gender identification. British Journal of Social Psychology, 43, 287–
297. doi:10.1348/0144666041501714
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
440 BOSSON AND MICHNIEWICZ
Deaux, K., & Lewis, L. L. (1984). Structure of gender stereotypes: Inter-
relationships among components and gender label. Journal of Person-
ality and Social Psychology, 46, 991–1004. doi:10.1037/0022-3514.46
.5.991
Donovan, R. A. (2011). Tough or tender: (Dis)similarities in White college
students’ perceptions of Black and White women. Psychology of Women
Quarterly, 35, 458 – 468. doi:10.1177/0361684311406874
Doosje, B., & Ellemers, N. (1997). Stereotyping under threat: The role of
group identification. In R. Spears, P. J. Oakes, N. Ellemers, & S. A.
Haslam (Eds.), The social psychology of stereotyping and group life (pp.
257–272). Oxford, England: Blackwell.
Fagot, B. I. (1977). Consequences of moderate cross-gender behavior in
preschool children. Child Development, 48, 902–907. doi:10.2307/
1128339
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G
Power 3: A
flexible statistical power analysis program for the social, behavioral, and
biomedical sciences. Behavior Research Methods, 39, 175–191. doi:
10.3758/BF03193146
Feinman, S. (1981). Why is cross-sex-role behavior more approved for
girls than for boys? A status characteristic approach. Sex Roles, 7,
289 –300. doi:10.1007/BF00287543
Felson, R. B. (1982). Impression management and the escalation of ag-
gression and violence. Social Psychology Quarterly, 45, 245–254. doi:
10.2307/3033920
Foushee, H. G., Helmreich, R. L., & Spence, J. T. (1979). Implicit theories
of masculinity and femininity: Dualistic or bipolar? Psychology of
Women Quarterly, 3, 259 –269. doi:10.1111/j.1471-6402.1979
.tb00544.x
Gilmore, D. D. (1990). Manhood in the making. New Haven, CT: Yale
University Press.
Glick, P., Gangl, C., Gibb, S., Klumpner, S., & Weinberg, E. (2007).
Defensive reactions to masculinity threat: More negative affect toward
effeminate (but not masculine) gay men. Sex Roles, 57, 55–59. doi:
10.1007/s11199-007-9195-3
Goff, P. A., Thomas, M. A., & Jackson, M. C. (2008). “Ain’t I a woman?”
Towards an intersectional approach to person perception and group-
based harms. Sex Roles, 59, 392– 403. doi:10.1007/s11199-008-9505-4
Gómez, A., Seyle, D. C., Huici, C., & Swann, W. B., Jr. (2009). Can
self-verification strivings fully transcend the self-other barrier? Seeking
verification of ingroup identities. Journal of Personality and Social
Psychology, 97, 1021–1044. doi:10.1037/a0016358
Grieve, P. G., & Hogg, M. A. (1999). Subjective uncertainty and inter-
group discrimination in the minimal group situation. Personality and
Social Psychology Bulletin, 25, 926 –940. doi:10.1177/
01461672992511002
Guimond, S., Branscombe, N. R., Brunot, S., Buunk, B. P., Chatard, A.,
Désert, M.,...Yzerbyt, V. (2007). Culture, gender, and the self:
Variations and impact of social comparison processes. Journal of Per-
sonality and Social Psychology, 92, 1118 –1134.
Haslam, S., O’Brien, A., Jetten, J., Vormedal, K., & Penna, S. (2005).
Taking the strain: Social identity, social support, and the experience of
stress. British Journal of Social Psychology, 44, 355–370. doi:10.1348/
014466605X37468
Haslam, S., & Reicher, S. (2006). Stressing the group: Social identity and
the unfolding dynamics of responses to stress. Journal of Applied
Psychology, 91, 1037–1052. doi:10.1037/0021-9010.91.5.1037
Heider, F. (1958). The psychology of interpersonal relations. New York,
NY: Wiley. doi:10.1037/10628-000
Herek, G. (1986). On heterosexual masculinity. Some psychical conse-
quences of the social construction of gender and sexuality. American
Behavioral Scientist, 29, 563–577. doi:10.1177/000276486029005005
Hogg, M. A. (2007). Uncertainty-identity theory. In M. P. Zanna (Ed.),
Advances in experimental social psychology (Vol. 39, pp. 69 –126). San
Diego, CA: Academic Press.
Hogg, M. A., & Abrams, D. (1993). Towards a single-process uncertainty-
reduction model of social motivation in groups. In M. A. Hogg & D.
Abrams (Eds.), Group motivation: Social psychological perspectives
(pp. 173–190). Hertfordshire, England: Harvester Wheatsheaf.
Hoyenga, K. B. (1993). Sex differences in human stratification: A biosocial
approach. In L. Ellis (Ed.), Social stratification and socioeconomic
inequality:Vol. 1.A comparative biosocial analysis (pp. 139 –157).
Westport, CT: Praeger.
Jetten, J., Spears, R., & Manstead, A. R. (1996). Intergroup norms and
intergroup discrimination: Distinctive self-categorization and social
identity effects. Journal of Personality and Social Psychology, 71,
1222–1233. doi:10.1037/0022-3514.71.6.1222
John, O. P., & Srivastava, S. (1999). The Big Five Trait taxonomy: History,
measurement, and theoretical perspectives. In L. A. Pervin & O. P. John
(Eds.), Handbook of personality: Theory and research (2nd ed., pp.
102–138). New York, NY: Guilford Press.
Kimmel, M. S. (1997). Masculinity as homophobia: Fear, shame and
silence in the construction of gender identity. In M. M. Gergen & S. N.
Davis (Eds.), Toward a new psychology of gender (pp. 223–242).
Florence, KY: Taylor & Frances/Routledge.
Latrofa, M., Vaes, J., Cadinu, M., & Carnaghi, A. (2010). The cognitive
representation of self-stereotyping. Personality and Social Psychology
Bulletin, 36, 911–922. doi:10.1177/0146167210373907
Leach, C. W., van Zomeren, M., Zebel, S., Vliek, M. L. W., Pennekamp,
S. F., Doosje, B.,...Spears, R. (2008). Group-level self-definition and
self-investment: A hierarchical (multicomponent) model of in-group
identification. Journal of Personality and Social Psychology, 95, 144 –
165. doi:10.1037/0022-3514.95.1.144
Levy, G. D., Taylor, M. G., & Gelman, S. A. (1995). Traditional and
evaluative aspects of flexibility in gender roles, social conventions,
moral rules, and physical laws. Child Development, 66, 515–531. doi:
10.2307/1131594
Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-
evaluation of one’s social identity. Personality and Social Psychology
Bulletin, 18, 302–318. doi:10.1177/0146167292183006
Lytton, H., & Romney, D. M. (1991). Parents’ differential socialization of
boys and girls: A meta-analysis. Psychological Bulletin, 109, 267–296.
doi:10.1037/0033-2909.109.2.267
Marques, J. M., Yzerbyt, V. Y., & Leyens, J. P. (1988). The “black sheep
effect”: Extremity of judgements towards ingroup members as a function
of group identification. European Journal of Social Psychology, 18,
1–16. doi:10.1002/ejsp.2420180102
McCreary, D. R. (1994). The male role and avoiding femininity. Sex Roles,
31, 517–531. doi:10.1007/BF01544277
Michniewicz, K. S., Vandello, J. A., & Bosson, J. K. (2013). Men’s
(mis)perceptions of the gender threatening consequences of unemploy-
ment. Manuscript submitted for publication.
Mullin, B. A., & Hogg, M. A. (1999). Motivations for group membership:
The role of subjective importance and uncertainty reduction. Basic and
Applied Social Psychology, 21, 91–102.
Paglia, C. (1992). Sex, art, and American culture: Essays. New York, NY:
Vintage Books.
Pelham, B. W. (1995). Self-investment and self-esteem: Evidence for a
Jamesian model of self-worth. Journal of Personality and Social Psy-
chology, 69, 1141–1150. doi:10.1037/0022-3514.69.6.1141
Pelham, B. W., & Swann, W. B., Jr. (1989). From self-conceptions to
self-worth: The sources and structure of self-esteem. Journal of Person-
ality and Social Psychology, 57, 672– 680. doi:10.1037/0022-3514.57.4
.672
Pickett, C. L., & Brewer, M. B. (2001). Assimilation and differentiation
needs as motivational determinants of perceived in-group and out-group
homogeneity. Journal of Experimental Social Psychology, 37, 341–348.
doi:10.1006/jesp.2000.1469
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
441
GENDER DICHOTOMIZATION IN INGROUP IDENTITY
Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social
dominance orientation: A personality variable predicting social and
political attitudes. Journal of Personality and Social Psychology, 67,
741–763. doi:10.1037/0022-3514.67.4.741
Pratto, F., Stallworth, L. M., & Sidanius, J. (1997). The gender gap:
Differences in political attitudes and social dominance orientation. Brit-
ish Journal of Social Psychology, 36, 49 – 68. doi:10.1111/j.2044-8309
.1997.tb01118.x
Prentice, D. A., & Carranza, E. (2002). What women and men should be,
shouldn’t be, are allowed to be, and don’t have to be: The contents of
prescriptive gender stereotypes. Psychology of Women Quarterly, 26,
269 –281. doi:10.1111/1471-6402.t01-1-00066
Robbins, J. M., & Krueger, J. I. (2005). Social projection to ingroups and
outgroups: A review and meta-analysis. Personality and Social Psychol-
ogy Review, 9, 32– 47. doi:10.1207/s15327957pspr0901_3
Rudman, L. A., Moss-Racusin, C. A., Phelan, J. E., & Nauts, S. (2012).
Status incongruity and backlash effects: Defending the gender hierarchy
motivates prejudice against female leaders. Journal of Experimental
Social Psychology, 48, 165–179. doi:10.1016/j.jesp.2011.10.008
Schimel, J., Simon, L., Greenberg, J., Pyszczynski, T., Solomon, S.,
Waxmonsky, J., & Arndt, J. (1999). Stereotypes and terror management:
Evidence that mortality salience enhances stereotypic thinking and pref-
erences. Journal of Personality and Social Psychology, 77, 905–926.
doi:10.1037/0022-3514.77.5.905
Schmitt, M. T., & Branscombe, N. R. (2001). The good, the bad, and the
manly: Threats to one’s prototypicality and evaluations of fellow in-
group members. Journal of Experimental Social Psychology, 37, 510 –
517. doi:10.1006/jesp.2001.1476
Settles, I. H., Pratt-Hyatt, J. S., & Buchanan, N. T. (2008). Through the
lens of race: Black and White women’s perceptions of womanhood.
Psychology of Women Quarterly, 32, 454 – 468. doi:10.1111/j.1471-
6402.2008.00458.x
Sirin, S. R., McCreary, D. R., & Mahalik, J. R. (2004). Differential
reactions to men’s and women’s gender role transgressions: Perceptions
of social status, sexual orientation, and value dissimilarity. The Journal
of Men’s Studies, 12, 119 –132. doi:10.3149/jms.1202.119
Spence, J. T., & Helmreich, R. L. (1978). Masculinity and femininity: Their
psychological dimensions, correlates, and antecedents. Austin: Univer-
sity of Texas Press.
Swann, W. B., Jr., & Bosson, J. K. (2010). Self and identity. In D. T.
Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychol-
ogy (5th ed., pp. 589 628). Hoboken, NJ: Wiley.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of inter-group
behavior. In S. Worchel & L. W. Austin (Eds.), Psychology of inter-
group relations (pp. 7–24). Chicago, IL: Nelson-Hall.
Thompson, E. H., Grisanti, C., & Pleck, J. H. (1985). Attitudes toward the
male role and their correlates. Sex Roles, 13, 413– 427. doi:10.1007/
BF00287952
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S.
(1987). Rediscovering the social group: A self-categorization theory.
Oxford, England: Basil Blackwell.
Vandello, J. A., & Bosson, J. K. (2013). Hard won and easily lost: A
review and synthesis of theory and research on precarious manhood.
Psychology of Men & Masculinity, 14, 101–113.
Vandello, J. A., Bosson, J. K., Cohen, D., Burnaford, R. M., & Weaver,
J. R. (2008). Precarious manhood. Journal of Personality and Social
Psychology, 95, 1325–1339. doi:10.1037/a0012453
Vandello, J. A., Ransom, S., Hettinger, V., & Askew, K. (2009). Men’s
misperceptions about the acceptability and attractiveness of aggression.
Journal of Experimental Social Psychology, 45, 1209 –1219. doi:
10.1016/j.jesp.2009.08.006
Weaver, J., Vandello, J. A., & Bosson, J. K. (2013). Intrepid, imprudent, or
impetuous? The effects of gender threats on men’s financial decisions.
Psychology of Men & Masculinity, 14, 184 –191.
Weaver, J., Vandello, J. A., Bosson, J. K., & Burnaford, R. (2010). The
proof is in the punch: Gender differences in perceptions of action and
aggression as components of manhood. Sex Roles, 62, 241–251. doi:
10.1007/s11199-009-9713-6
Williams, J. E., & Best, D. L. (1990). Sex and psyche: Gender and self
viewed cross-culturally. Thousand Oaks, CA: Sage.
Received July 31, 2012
Revision received April 19, 2013
Accepted April 19, 2013
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
442 BOSSON AND MICHNIEWICZ
... Taking the example of gender identity, consumers could signal their gender online by actively supporting or, on the As already mentioned in the introduction, gender is a complex and unique identity due to the number of approaches psychologists and sociologists developed to study this concept. Wood and Eagly (2015) argue that there are two main approaches to gender identity in psychology-(a) based on personality traits associated with femininity/masculinity (communion/agency) (e.g., Evers & Sieverding, 2014), and (b) based on social identity and self-categorization, i.e., gender group identity, or one's strength or importance of identification with their gender group (e.g., Bosson & Michniewicz, 2013). We introduce the two approaches to gender identity in psychology and connect them with a broader identity classification. ...
... Similarly, Van Breen et al. (2017) advocate for a multiple identity approach to gender in women, emphasizing that identification with women can predict attitudes towards self-ascribed femininity, whereas feminist identity is related to women's social position and attitudes towards sexism and gender equality. Men's strong gender identity leads to the process of gender dichotomization, a tendency to distance masculine traits from feminine (Bosson & Michniewicz, 2013), which might indicate that they would want to signal their gender in the marketplace We would like to point out that gender role beliefs, like identity strength and identity-related digital behaviors, exist on a continuum. To illustrate the differences, we use two extremes: egalitarian and traditional gender role beliefs. ...
... Spielmann et al. (2021) show that men with traditional gender role beliefs have a stronger connection to male brand representations than those with egalitarian beliefs. Moreover, according to Bosson and Michniewicz (2013), strongly identified men have the highest tendency to eschew femininity from their self-concept. Consequently, we suggest that men with strong gender identity and traditional gender role beliefs have an increased desire to signal traditional masculinity and agency in their digital behaviors. ...
Article
Full-text available
Nowadays consumers can express their identities not only through their possessions and buying behavior, but also using social media and digital networks. This article aims to understand these digital consumer behaviors by focusing on identity strength and the identity signaling phenomenon. We develop a conceptual model that combines internal and external factors to explain the intensity and content of digital identity-related behaviors. We use the example of gender identity to build our research propositions, as gender is one of the most frequently and intensely debated identities in online consumer discussions. Further, we propose how digital and offline identity signaling behaviors are intertwined, and discuss the online behaviors of trans consumers. In doing so, our conceptual work highlights the unique features of digital identity signaling behaviors as well as the complexity of identities, including gender, and provides useful insights for researchers and marketers.
... One possible reason is that compared to heterosexual women, heterosexual men may be less comfortable with gender nonconformity (Allen & Smith, 2011;Bosson & Vandello, 2011;Sloan et al., 2015;Vandello et al., 2008) and, therefore, feel more threatened by trans people (Harrison & Michelson, 2019). In addition, men are more likely than women to engage in gender dichotomization or to view femininity and masculinity as distinctly different categories (Bosson & Michniewicz, 2013) which may lead to stronger beliefs in gender binary thinking (see also Doan et al., 2019). For example, Broussard et al. (2018) found that being a cisgender heterosexual man predicted a preference for a tworesponse option when asking about gender (e.g., only Female and Male options). ...
... Many trans people are perceived as not conforming to norms in their gender expression, appearance, and behavior (Gazzola & Morrison, 2014). Compared to cisgender men, cisgender women tend to be more comfortable with gender nonconformity (Allen & Smith, 2011;Bosson & Vandello, 2011;Sloan et al., 2015), are less likely to dichotomize feminine and masculine characteristics (Bosson & Michniewicz, 2013), and they feel less threatened by individuals who challenge traditional norms of gender identity, such as trans people (Harrison & Michelson, 2019). These feelings of discomfort and threat to one's own gender identity, as well as a tendency to strictly dichotomize gender, may lead to prejudice and discrimination against trans people, especially among cisgender heterosexual men (Glotfelter & Anderson, 2017;Nagoshi et al., 2008). ...
... Again, this discrepancy may be due to our more diverse sample with regard to sexual orientation and educational level. However, it would be worthwhile to examine gender comparisons for associations between the gender identity theme in transgender definitions and the tendency to engage in gender dichotomization (Bosson & Michniewicz, 2013), gender binary thinking (Doan et al., 2019), and a preference for more than two gender responses (Broussard et al., 2018). Although participants' gender was not significant, participants' sexual orientation was, as evidenced by the finding that nonheterosexual participants referenced the gender identity theme more than heterosexual participants. ...
... Although not explicitly addressed in the study, another interesting finding was that compared to women, men showed significantly greater negative attitudes towards gay and trans people, even after introducing disgust sensitivity and RWA measures in the model. It has been suggested that the traditional masculine identity is sustained by a more profound endorsement of gender norms and sexist beliefs that distance masculine from feminine traits (Bosson & Michniewicz, 2013). Thus, it may be that gay and trans people threaten these ideals of masculinity and as a result men report significantly stronger levels of homonegativity and transnegativity compared to their women counterparts (Kiss et al., 2020;Nagoshi et al., 2008). ...
Article
The present study examined the associations between three forms of disgust sensitivity (i.e. moral, pathogen, and sexual) and homonegativity towards gay men and lesbian women, based on the behavioural immune system (BIS) theory. Two forms of homonegativity were assessed: old-fashioned (i.e. moral and religious objections to homosexuality) and modern (i.e. objections to homosexuality that are grounded in beliefs such as sexual minorities demand and receive ‘preferential’ treatment). Frequency and valence of contact with sexual minorities also was measured. An online survey was completed by 263 self-identified heterosexual participants, a majority of whom were White (n = 173) and cisgender women (n = 192). Sexual disgust was the strongest predictor of old-fashioned homonegativity towards lesbian women, and pathogen disgust was the only predictor of old-fashioned homonegativity against gay men. No measures of disgust were statistically significant predictors of modern homonegativity. Both frequency and quality of intergroup contact played a significant role in moderating different effects of sexual disgust on homonegativity. The limitations of this study and directions for future research are outlined.
... Females in Saudi Arabia may find it difficult at first to shop online, but once they have some online experience, they are more likely to view the local cultural and legal restrictions on their activities as being less valid or fair, thus causing them to raise their tendency to shop online. Consistently with this, Bosson and Michniewicz [17] contend that females face more difficulty when shopping online than males do. In this study, the researcher will examine the relationship between perceived ease of use and females' attitude toward online shopping. ...
Article
Full-text available
This research aims to determine the factors that influence females’ attitude toward online shopping in Saudi Arabia. The potential and main theory that is selected for this study is the Theory of Acceptance Model (TAM) which studies online shopping acceptance of new technology through two important variables which are perceived ease of use and perceived usefulness. Furthermore, this research has included the factors of site quality, site credibility, and customer fashion lifestyle. The self-administrated method has been used to distribute the questionnaire for the respondents, who are selected by utilizing the judgement sampling technique. 320 questionnaires were valid for analysing. SPSS and Smart PLS software have been employed for analysing the data of study. The finding of the present pointed out that perceived ease of use, site quality, and customer fashion lifestyle have a significant positive relationship with females’ attitude toward online shopping. In contrast, usefulness and site credibility were not statistically significant. This study has significantly contributed to literature on the shopper’s attitude toward online shopping. On the other hand, the findings of this study represented a guideline for e-retailers in Saudi Arabia to increase and maintain online shoppers for the long term and compete with other competitors. Finally, study limitations are discussed, and future study directions are proposed.
... To be a "real man" implies repressing traits or behaviors that are culturally coded as feminine. Since men feel continual pressure to prove their masculinity and demonstrate they are "real men" (Vandello & Bosson, 2013;Vandello et al., 2008), stereotyped feminine traits and behaviors are actively avoided, notably by those men who endorse such a conception of masculinity, in a way to affirm their masculine identity (Bosson & Michniewicz, 2013). This may lead them to exhibit sexist attitudes and behave aggressively against women (e.g., Glick & Fiske, 1996;Kilianski, 2003;Smith et al., 2015), especially when their masculinity is threatened (Bosson et al., 2009;Michniewicz & Vandello, 2015). ...
Article
Full-text available
The current research investigated the endorsement of hegemonic masculinity, sexism, and homophobia, and the perceptions of discrimination, among samples of heterosexual male and female, and LGB students who had entered into traditionally male-dominated and female-dominated fields of study. Specifically, students from vocational and educational training in Swiss upper-secondary schools were recruited. Results revealed that adherence to hegemonic masculinity, sexism, and homophobia is higher in male-dominated fields of study (vs. female-dominated). Furthermore, heterosexual female and LGB students enrolled in male-dominated fields of study have been found to experience and anticipate more discrimination than heterosexual male students. Implications of these results are discussed.
... A "good man" must also consistently demonstrate toughness, as evidenced by his ability to withstand physical and mental pain, and emotional distress, without showing signs of stereotypically feminine emotions (e.g., fear, anxiety) and behaviors (e.g., uncontrolled tears; Shields, 2002). Good men are also nothing like women (Bosson et al., 2009;Dahl et al., 2015) or feminized men (see also Bosson & Michniewicz, 2013;Kite & Deaux, 1987;Lehavot & Lambert, 2007;Pascoe, 2007). ...
Article
Full-text available
The present work examined whether men's and women's gender-identities and experiences of gender threats influenced their self-images. Findings across two studies (N = 567) revealed that masculinity in men appears to be more precarious than femininity is in women, but when similarly threatened in a given situation both men's and women's anger predicted their construction of gender compensatory self-images. Specifically, in Study 1, participants' definition of the self in terms of gender ingroup (vs. outgroup) traits (a) positively predicted the gender stereotypicality of men's and women's actual photographs and women's constructed self-images, but (b) negatively predicted the gender stereotypicality of men's self-images. Men whose self definitions least strongly prioritized gender ingroup (over outgroup) traits generated the most gender stereotypic self-images, as rated by independent judges. In addition, in Study 2, after being led to believe that they performed like average members of their gender outgroup (i.e., threat condition) on a gender knowledge test, men expressed more public discomfort and were angrier than women. Gender threat (vs. assurance) also indirectly predicted the generation of more gender stereotypic self-images for men, but not women; this effect was significant via serial mediation, through public discomfort and anger. However, extending prior findings, anger (but not public discomfort) was significantly associated with and predicted the construction of feedback contradicting self-images similarly. We discuss the implications of these findings for theory and research on gender-identity, self-image, and compensatory gender threat responses.
... previous research showing that gender norms for boys tend to be more restrictive than those for girls (Lytton and Romney, 1991;Sullivan et al., 2018), as well as on research in precarious masculinity theory. This theory suggests that masculinity is a precarious status but femininity is more stable, and that men are consequently more sensitive to gender prototypicality threats than women (Bosson and Michniewicz, 2013;Vandello and Bosson, 2013). For ability stereotypes, no specific gender differences are predicted in the strength of their relations with other variables. ...
Article
Full-text available
Despite progress, gender gaps persist in mathematical and language-related fields, and gender stereotypes likely play a role. The current study examines the relations between parents' gender-related beliefs and their adolescent child's motivation and career aspirations through a survey of 172 parent-child dyads. Parents reported their gendered beliefs about ability in mathematics and language arts, as well as their prescriptive gender role beliefs. Students reported their expectancies and values in these two domains, as well as their career aspirations The results of path models suggested that parents' ability stereotypes about language boosted girls' motivation for language arts, thereby nudging them away from STEM pathways. Girls' career aspirations stemmed not only from their valuation of the corresponding domain, but also from their valuation of competing domains. Such findings highlight the need to consider multiple domains simultaneously in order to better capture the complexity of girls' career decisions. For boys, parents' language ability stereotypes were directly related to mathematical career aspirations. These results suggest that stereotypes that language arts is not for boys push them instead toward mathematics. Our study also highlighted the unique role of parental beliefs in traditional gender roles for boys' motivation and career aspirations. Specifically, parents' gender role stereotypes directly related to less interest in language arts only among boys. This highlights that research into gender gaps in female-dominated fields should consider stereotypes related to appropriate behavior and social roles for boys.
... Although women may resist hegemonic femininity and adopt more masculine traits (e.g., butch, tomboy) without comparatively much hostility, men who adopt more feminine traits often experience social backlash from strangers, friends, and even family members-notably fathers [4]. Despite current social trends, social dichotomization between femininity and masculinity continues to exist [5,6]. This suggests a rigidity in male gender norms, as compared to female gender norms, whereby social trends favor divergences from traditional gender norms for females but not for males. ...
Article
Full-text available
Currently, research explicitly examining masculinity and internalized homonegativity is sparse, and even sparser studies are those using qualitative methods. To address this, this study aims to explore: how gender norms are constructed and experienced amongst gay men; and how gender and sexual identity are experienced in relation to masculine norms amongst gay men. A sample of 32 self-identified gay men aged 22–72 years (M = 34.34, SD = 12.94) participated in an online semi-structured interview on masculinity and homosexuality. The study used Zoom to facilitate the online interviews as it offered privacy, accessibility, ease of use, and voice recording, among other benefits. Thematic analyses revealed gay men’s understandings of masculinity, femininity, and sources of pressure to conform. Furthermore, gay men emphasize the conflict experienced between heteronormative gender and sexuality norms, which highlights the term homosexual male as an oxymoron.
Article
Full-text available
Social role theory posits that binary gender gaps in agency and communion should be larger in less egalitarian countries, reflecting these countries’ more pronounced sex-based power divisions. Conversely, evolutionary and self-construal theorists suggest that gender gaps in agency and communion should be larger in more egalitarian countries, reflecting the greater autonomy support and flexible self-construction processes present in these countries. Using data from 62 countries (N = 28,640), we examine binary gender gaps in agentic and communal self-views as a function of country-level objective gender equality (the Global Gender Gap Index) and subjective distributions of social power (the Power Distance Index). Findings show that in more egalitarian countries, gender gaps in agency are smaller and gender gaps in communality are larger. These patterns are driven primarily by cross-country differences in men’s self-views and by the Power Distance Index (PDI) more robustly than the Global Gender Gap Index (GGGI). We consider possible causes and implications of these findings.
Article
Disgust specific to sexual stimuli has been thought to be an adaptation that serves purposes of pathogen-avoidance, partner selection and social dominance. While the link between disgust responses and homonegative and transnegative attitudes has been relatively established, it is not yet clear why. Literature using evolutionary psychology perspectives of these phenomena is scarce in areas with substantial LGBT-related violence like Colombia. This research aimed to study the interplay of predispositional (e.g. sociodemographic, reported contact), affective (e.g. disgust sensitivity) and cognitive (e.g. Right-Wing Authoritarian) variables in homonegative and transnegative attitudes of Colombian adults. Participants (N = 272) had a mean age of 26.38 (SD = 9.47), women comprised 72% of the sample and men 28%. Hierarchical regression analyses showed that increased sexual specific disgust sensitivity and greater RWA predicted stronger homonegativity and transnegativity. The relationship between sexual disgust and prejudice was partially mediated by RWA. Findings suggest that sexual disgust sensitivity adaptations in homonegativity and transnegativity may respond to selection pressures that differ from pathogen-avoidance perspectives, and that are associated with maintaining social hierarchy and social dominance. Anti-prejudice initiatives would benefit from targeting emotional responses of sexual disgust, especially within communities and institutions that have historically endorsed conservative and traditional values.
Article
An experiment was conducted to investigate the idea that an important motive for identifying with social groups is to reduce subjective uncertainty, particularly uncertainty on subjectively important dimensions that have implications for the self-concept (e.g., Hogg, 1996; Hogg & Mullin, 1999). When people are uncertain on a dimension that is subjectively important, they self-categorize in terms of an available social categorization and, thus, exhibit group behaviors. To test this general hypothesis, group membership, task uncertainty, and task importance were manipulated in a 2 × 2 × 2 between-participants design (N = 128), under relatively minimal group conditions. Ingroup identification and desire for consensual validation of specific attitudes were the key dependent measures, but we also measured social awareness. All three predictions were supported. Participants identified with their group (H1), and desired to obtain consensual validation from ingroup members (H2) when they were uncertain about their judgments on important dimensions, indicating that uncertainty reduction motivated participants towards embracing group membership. In addition, identification mediated the interactive effect of the independent variables on consensual validation (H3), and the experimental results were not associated with an increased sense of social awareness and, therefore, were unlikely to represent only behavioral compliance with generic social norms. Some implications of this research in the study of cults and "totalist" groups and the explication of genocide and group violence are discussed.
Article
A meta-analysis of 172 studies attempted to resolve the conflict between previous narrative reviews on whether parents make systematic differences in their rearing of boys and girls. Most effect sizes were found to be nonsignificant and small. In North American studies, the only socialization area of 19 to display a significant effect for both parents is encouragement of sex-typed activities. In other Western countries, physical punishment is applied significantly more to boys. Fathers tend to differentiate more than mothers between boys and girls. Over all socialization areas, effect size is not related to sample size or year of publication. Effect size decreases with child's age and increases with higher qualify No grouping by any of these variables changes a nonsignificant effect to a significant effect. Because little differential socialization for social behavior or abilities can be found, other factors that, may explain the genesis of documented sex differences arc discussed.
Article
W. James (1890) argued that the importance people attach to their self-views determines the impact of these self-views on people's global feelings of self-worth. Despite the intuitive appeal of this position, most research on the relation between people's specific self-views and their global self-esteem has failed to support this assertion. B. W. Pelham and W. B. Swann (1989) provided evidence in support for W. James's assertion, but H. W. Marsh (1993) criticized this evidence. In this article, further evidence is presented for W. James's (1890) assertion. In addition, the favorability of people's specific self-views is identified as a moderator of the extent to which belief importance is related to self-esteem. The theoretical implications and limitations of these findings are discussed.