18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.

sgm LaTeX2e(2002/01/18) P1: GCE

10.1146/annurev.psych.55.090902.141537

Annu. Rev. Psychol. 2004. 55:519–44

doi: 10.1146/annurev.psych.55.090902.141537
Copyright °c 2004 by Annual Reviews. All rights reserved
First published online as a Review in Advance on September 29, 2003

GENDER IN PSYCHOLOGY
Abigail J. Stewart and Christa McDermott
Psychology Department and Women’s Studies Program, University of Michigan,
Ann Arbor, Michigan 48109-1109; email: abbystew@umich.edu, mcdc@umich.edu
Access provided by Michigan State University Library on 02/02/15. For personal use only.
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Key Words sex differences, within-sex differences, social power, intersectionality

■ Abstract Gender is increasingly understood as defining a system of power re-
lations embedded in other power relations. Psychological research on gender—which
has most often focused on analysis of sex differences, within-sex variability, and gen-
der roles—has begun to incorporate this new understanding. By drawing on three
resources, psychologists can make more rapid progress in understanding gender’s sig-
nificance for psychological processes: social science theories that link the individual
and social levels of analysis; constructs (such as identity) that bridge the social and
individual levels; and conceptual tools generated in feminist theory, perhaps especially
intersectionality. We review these resources, cite active research programs that have
employed them, and conclude by offering some practical suggestions about how to
incorporate these resources into our research.

CONTENTS
GENDER IN PSYCHOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
A Model for Theorizing Gender in Psychology: Linking Social
Structure and Personality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
Conceptual Tools for Psychology from Feminist Theory . . . . . . . . . . . . . . . . . . . . . 526
Gender as Analytic Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
Gender as a Set of Embedded Power Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Multiplicity and Instability of Selves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
Intersectionality: Theory and Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
Psychological Studies that Inform Our Study of Gender
from an Intersectional Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
How Can Psychologists Use Intersectionality to Improve Research? . . . . . . . . . . . . 537
Gender’s Diverse Meanings in Psychology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538

GENDER IN PSYCHOLOGY
Gender is widely recognized to be an important empirical factor (or variable)
in understanding many aspects of behavior. In psychology gender is often used
empirically, without much consciousness of its social or conceptual significance.
In this chapter we focus on the use of gender as an analytic tool in psychology. Of
0066-4308/04/0204-0519$14.00 519
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

520 STEWART ¥ MCDERMOTT

course analytic tools must prove empirically useful (or not), but recognizing the
conceptual significance of different ways of using gender in empirical research
can, we think, help psychologists identify newer and more powerful ways to use
gender to study psychological phenomena. Ordinarily psychologists use gender in
empirical research in at least three wholly different ways: to signify sex differences,
within-sex variability, and the gender-linked power relations that structure many
social institutions and interactions.
First, and perhaps still most often, gender is used to think about ways in which
boys and girls or men and women differ. According to this “sex differences ap-
proach,” psychologists consider how and why average differences in personality,
Access provided by Michigan State University Library on 02/02/15. For personal use only.

behavior, ability, or performance between the sexes might arise (see, e.g., Block
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

1984, Buss 1995, Eagly 1994, Levy & Heller 1992, Maccoby & Jacklin 1974, and
Maccoby 1998). This approach often appears to assume, or actually does assume,
that these differences arise from preexisting “essential” differences between male
and female human organisms. In its strongest form, the sex differences approach
assumes not only will there be group differences between men and women on
key traits but also nonoverlapping distributions. Early on it was noted that few—
but some—behavioral differences have nearly nonoverlapping distributions, no-
tably: ejaculation, pregnancy, childbirth, and lactation. These derive directly from
demonstrable biological differences between men and women. In most other cases,
perceived and demonstrated sex differences on some characteristic or behavior oc-
cur with overlapping distributions, sometimes with highly gender-differentiated
“tails” of the distribution (e.g., some sex-linked developmental disorders, or ex-
tremely high scores in math ability, both of which are disproportionately found in
boys). Even in these cases, and more markedly in cases with less differentiated
tails on the distribution, there are simply average sex differences with wide and
overlapping distributions (Favreau 1993, 1997).
Psychologists have entertained several kinds of hypotheses to account for these,
ranging from biological differences (e.g., in behavior-related hormones), differen-
tial treatment or socialization (which could, of course, be grounded in biological
differences but enhanced by differential treatment), and differential social roles or
social situations. All three kinds of explanations are aimed at accounting for overall
sex differences, or differences between groups of men and women or girls and boys
(see, e.g., Maccoby 1998). They focus, then, on the sharply demarcated differences
(in the very few instances of nonoverlapping distributions), or on the aggregate
differences in the distributions, or on the gender-differentiated tails. For example,
research on gender differences in the rate of depression has explored the potential
role of gender-linked hormones, differential socialization toward inhibition, and
differential exposure to childhood sexual abuse (see Nolen-Hoeksema et al. 1999).
Some psychologists have noted that in many cases a focus on sex differences ig-
nores the large variance within gender on many characteristics (Martin 1994), and
therefore may tend to exaggerate sex differences, or even reinforce or create them
in the mind of the public (Hare-Mustin & Marecek 1990, Unger 1989). To avoid
such exaggerations, some psychologists have selected particularly “gendered”
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 521

phenomena (that is, phenomena that are defined by being different, on average,
between men and women), and have explored sources of the within-gender vari-
ance on those characteristics. Thus, for example, Eccles & Jacobs (1986) noted
that both math ability and math performance are “gendered” in that at certain ages
boys demonstrate higher average ability and higher performance (see also Eccles
et al. 1990). Even at those ages, though, there is enormous variance in math ability
and performance among both boys and girls. By examining the factors that predict
differential ability and performance within gender they showed that girls’ ability
and performance scores were predicted by parents’ expectations of them. Thus, to
the extent that parents form rigid ideas about the math ability and behavior of boys
Access provided by Michigan State University Library on 02/02/15. For personal use only.

and girls, and act on those ideas, differences between boys and girls are likely to
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

be exaggerated, if not actually created.

Research of this kind can raise further questions about the role of social
structure—in addition to social interactions like those between parents or teach-
ers and children—and gender. Some psychologists have suggested that “gender”
operates not merely at the level of sex differences, or as the result of social inter-
actions in which beliefs about gender are expressed in actions that actually create
confirming evidence for those beliefs, but also in the social structures that define
power relationships throughout the culture (Fiske & Stevens 1993, Goodwin &
Fiske 2001, Stewart 1998). According to this viewpoint, “gender” describes a set
of power relations in which—absent other cues and definitions—maleness signals
authority, status, competence, social power and influence, and femaleness signals
lack of authority, low status, incompetence, and little power and influence. [This
perspective is perfectly compatible with recognition that gender also signals more
positive associations with femaleness (e.g., nurturance) and more negative ones
with maleness (e.g., violence). From this perspective, the point is that social status
and authority are not associated with these characteristics in the same way.]
Psychologists who begin by noticing that gender signifies a set of power rela-
tions often focus on how a particular behavior of interest (e.g., leadership, marital
conflict, or performance of a task) takes place within social relationships (dyads,
organizational hierarchies, broader social structures) that are themselves gendered.
For example, a partner may stay in an abusive relationship because she lacks the
economic resources to leave. If marital conflict, and perhaps especially violence,
takes place in dyads in which one individual has total or relative control of financial
and extrafamilial resources, it is important to consider that fact in understanding
the factors that maintain both the violence and the relationship. Recognizing the
importance of the power structure of the dyad may open up potential avenues
for change; shifting some financial and resource power to equalize the balance
may actually change the pattern of conflict behavior (White et al. 2001). Equally,
scholars who have examined the efficacy of organizational mechanisms for re-
porting sexual harassment have discovered that those mechanisms rarely produce
systematic reporting by women in organizations in which men hold most of the
most powerful positions. Fears of retaliation are, under these conditions, quite
rational and widely held. Widespread underreporting of these incidents is difficult
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

522 STEWART ¥ MCDERMOTT

to understand without an understanding of the gendered power relationships in

organizations (Riger 1991).
To summarize, psychologists have found “gender” to be a powerful conceptual
tool in at least three ways: (a) in sorting individuals into male and female and explor-
ing the ways in which differences in behavior, performance, and characteristics
are associated with that individual difference (whether the hypothesized causal
mechanism is biological, socialization, or social location); (b) in understanding
how gender might relate to individual differences among men and among women;
and (c) in understanding how gender structures social institutions within which
men and women operate. These three approaches have generally been viewed as
Access provided by Michigan State University Library on 02/02/15. For personal use only.

alternatives, and often are pursued in isolation from one another. One of the im-
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

portant recent advances has been an increased inclination to recognize that these
three approaches are in fact compatible and can be integrated.
It is fair to say that the first approach (sex differences) is the most well devel-
oped within psychology, and the most widely used in the field. At the same time,
advances in feminist theory, within interdisciplinary women’s studies as well as
within psychology, have raised serious questions about the degree to which this
approach—by itself—can offer much to our understanding of complex social be-
havior (see, e.g., Martin 1994, Riger 1992). Some areas of psychology—perhaps
especially developmental and clinical psychology—have examined the role of gen-
dered socialization pressures quite thoroughly and with considerable success in
adding to our understanding of both average gender differences and the within-
gender variance in certain phenomena (e.g., math achievement, emotions, PTSD,
and depression; see Brody 1999, Eccles et al. 1990, Fivush & Buckner 2000,
Nolen-Hoeksema et al. 1999, Pimlott-Kubiak & Cortina 2004, Powlishta et al.
2001, Zahn-Waxler 2000). Not surprisingly, it has been organizational, social, and
community psychology that have been most attuned to the issue of gender in social
structure (see, e.g., Gutek & Done 2001, Riger 2000). It is equally unsurprising—
given the highly individualistic focus of psychology as a field—that this approach
has been least prominent in the exploration of gender in psychology as a whole.
We hope, in this chapter, to show how adopting a framework that integrates social
structural and individual approaches to understanding gender can provide pow-
erful accounts of particular psychological phenomena. We understand that one
factor constraining psychologists in using this kind of framework is the need to
identify practical ways to incorporate these different levels of analysis in a single
design. We therefore offer some examples of successful strategies for doing so and
conclude with some suggestions about how to strengthen our research by using
gender more consistently as a conceptual tool.

A Model for Theorizing Gender in Psychology: Linking Social

Structure and Personality
An important precondition for using gender as a conceptual tool to full advantage
in psychology is having a developed understanding of how to bring factors at
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 523

different levels of analysis—social structural, individual, and even biological—

into a single model. The field of personality theory offers us a remarkably diverse
set of models for doing this, but they have been used infrequently in other subfields
of psychology and even in empirical personality research. We suspect this is largely
due to the isolation of subfields from one another, and propose that these models
have wide applicability within psychology (and even in our neighboring disciplines
in both the social and biological sciences). There are also, of course, difficulties
in operationalizing models at multiple levels; these pose later, and less difficult
problems than the absence of adequate, or fully developed, theory.
Psychoanalytic theory offered early ideas about the relationship between the
Access provided by Michigan State University Library on 02/02/15. For personal use only.

individual and society (Freud 1930/1961). Early critiques of psychoanalytic the-

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

ory often offered elaborations of the theory designed to address its exclusively
intrapsychic and interpersonal focus. Some branches of psychoanalytic theory de-
veloped even more extensively the intrapsychic focus (e.g., Klein 1965), while
others developed the interpersonal focus (e.g., Sullivan 1953). But a vocal and
persistent branch elaborated a “psychosocial” understanding of the individual that
incorporated social structure into a model that included attention to individual
intrapsychic and interpersonal experiences and their consequences. Erich Fromm
(1947) argued, for example, that the dominant economic system in which individ-
uals are embedded shapes personality development. Karen Horney (1939) argued,
even more relevantly for thinking about gendered social structure, that women
envy not the biological penis, but what it signifies—social power and authority.
Many contemporary feminist psychoanalytic theorists have argued for the powerful
influence of social structure in creating and constraining the range of emotional
and other associations individuals may have to particular social, interpersonal,
and individual objects and symbols (Benjamin 1988; Chodorow 1978, 1994; Flax
1990).
During the post-World War II period, a branch of interdisciplinary social sci-
ence flourished that aimed at studying “culture and personality.” Scholars in this
tradition focused on understanding how cultural differences—in social struc-
ture, practices, and ideology—might shape individual psychology (Honigman
1967, Kluckhohn et al. 1953). Anthropologists including the Kluckhohns and the
Whitings (Whiting 1963) exemplified this approach, and examined a range of
individual-level behaviors (prosocial behavior, child-rearing practices) in differ-
ent cultures. Equally, sociologists (Inkeles & Smith 1974; Parsons 1955, 1967)
argued that features of the social structure that varied in different contexts (e.g.,
the gendered differentiation of social roles, the degree of modernization) are con-
sequential for individual personality. Finally, psychologists in this tradition, such
as David McClelland, focused on the ways in which transcultural phenomena—
like achievement motivation—might nevertheless be predicted by features of the
economic system and religion. These approaches required scholars to collect data
across different cultures, and to specify direct and indirect connections between
the social level and the individual level. For example, McClelland (1967) argued
that under capitalism, child-rearing focuses on increasing individual initiative (and
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

524 STEWART ¥ MCDERMOTT

hence achievement motivation), but under precapitalist systems it focused on mini-

mizing initiative, while enhancing willingness to submit to authority. In turn, these
systematic differences partly account for the relatively higher levels of achieve-
ment motivation in capitalist versus precapitalist systems. These theorists began
to articulate the specific cultural and societal features that might be pertinent for
individual development, and identified mechanisms that might link those features
with individual personality development. The approach depended on finding av-
erage personality differences across cultures and then linking those differences to
differences in social structure.
The culture and personality approach was widely criticized, partly by those who
Access provided by Michigan State University Library on 02/02/15. For personal use only.

felt that features of the culture or social structure were often identified in a fairly
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

arbitrary way, and by anthropologists who felt that the approach tended to cre-
ate over-general, even stereotypical images of cultures (Inkeles & Levinson 1954,
1969). Moreover, it depended on viewing different cultures with the same general
conceptual categories, an approach in conflict with anthropology’s fundamental
effort to take different cultures on their own terms, rather than imposing preex-
isting categories imported from a different culture (Shweder 1991). Nevertheless,
psychologists and sociologists have continued to examine links between personal-
ity and social structure within large, complex cultures like our own. The reasoning
here has been that individuals occupy different locations within the social structure,
and those different locations are consequential for the development of personality.
Sociologist James S. House (1981) and psychologist Carol Ryff (1987) are par-
ticularly prominent exponents of these views. Psychologist Joseph Veroff (1983)
specified [as did Bronfenbrenner (1986) for developmental psychology] that indi-
vidual personality must be understood within a variety of social contexts that are
increasingly removed from the person (dyad, family, neighborhood or community,
and wider society). None of these theorists focused particularly on gender, but
all of them provide specific guidance on how to link individual-level phenomena
(personality, behavior, attitudes, and performance) with social-level phenomena.
One critical tool for making the connection is the concept of identity.
Identity emerged as an important concept in psychology during the same post-
war period as culture and personality. Erik Erikson—one of the psychosocial psy-
choanalytic theorists searching for ways to link the intrapsychic, interpersonal, and
social levels of analysis—articulated a relatively broad theory of identity that also
provided that link. According to Erikson (1950/1963, 1968), identity is an emer-
gent structure of personality that develops throughout the lifespan but takes on
organization and significance in adolescence. He argued that identity is composed
of many disparate elements, including one’s identification with one’s own past
experience, with particular characteristics and traits, with ideas and ideologies,
and with a defined place (often an occupation, but also other roles, e.g., family
and gender) in the social structure. Adolescence is critical to this theory because
identity is a self-conscious structure that depends on a certain degree of cognitive
understanding of both the self and the social structure; it is also critical because it
is a period during which certain kinds of social affirmation (including awarding of
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 525

degrees, paying of salary, etc.) begin to accrue. These affirmations in turn support
some identity developments and preclude others. Erikson’s notion of identity is
extremely broad and inclusive, which may suggest that one’s identity may only
have a modest level of coherence derived both from the individual’s desire to per-
ceive her/himself as coherent and the social pressures that encourage it. Erikson’s
“identity” is quite malleable, or subject to continued change over the adult life
course, based on both individual experiences and input from the social structure.
Erikson (1968, see also Erikson 1974) viewed identity as gendered in sev-
eral senses. First, he assumed that identification with a gendered body and with
gendered roles was part of identity. Second, he understood that social structures
Access provided by Michigan State University Library on 02/02/15. For personal use only.

constrained and encouraged identity developments of different sorts in a variety of

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

ways. Finally, he knew that one of these constraints was the operation of gender
in heterosexual couples. While feminist psychologists have found many specific
observations of Erikson to be grounded in a limited view of gender possibili-
ties (Franz & White 1985, Strouse 1974), the broader theory of identity offers a
specific vision of how social structural features are internalized by individuals,
and thus how social structure and the individual may be linked. Stewart & Healy
(1989) suggested that in fact sociological ideas of generation and individual the-
ories of development such as Erikson’s could be integrated—using the concept of
identity—to help us examine empirically how and why generations differ. Cole
et al. (2001), Helson et al. (1995), and Stewart (2003) have demonstrated how this
approach can be used to study changes in gendered phenomena over time.
During the same period when Erikson was developing his personality theory
and subsequent scholars were building on it, social psychologists developed a quite
separate and significant line of theoretical and empirical understanding of social
identity. This approach actually grew out of research on perceptual misestimation
(Tajfel 1957), which Tajfel (1981) later generalized to the emotions, values, and
stresses attached to group memberships. He showed that these attachments to
groups (defined in a variety of ways—by religion, nation, occupation, race, gender,
etc.) are consequential for a variety of perceptions and behaviors. Since Tajfel’s first
powerful observations, many social psychologists have demonstrated that social
identities based on group identifications are indeed consequential. Retrospectively
this approach can be seen as an elaboration of one element of identity (identification
with groups) in Erikson’s theory. It offers us a systematic way of understanding how
that element of identity operates. It is important to note that social identity theorists
define as “groups” some features of the social structure that other theorists would
define differently. Thus, some theorists would view “groups” such as occupations,
gender, etc., as roles (e.g., Parsons 1955, 1967), while others would view them as
social structures (e.g., Gurin et al. 1980). There may be important consequences to
these different ways of thinking about these identities (as attached to groups or to
social structures), but for many scholars the notion of social identity has offered an
approach to linking social structure and individual behavior. More narrowly some
psychologists have found connections between social structures and gendered
behavior through gender identities.
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

526 STEWART ¥ MCDERMOTT

Gurin and her colleagues (Gurin et al. 1980) noted that social identities could
be attached to social strata as well as collections of individual persons. They ex-
amined how identification with a social stratum (class, gender, and age) predicted
political beliefs (Gurin 1985, 1987). Gurin & Markus (1989) explored how gender
consciousness (or identification with a gender stratum) related to other ideas about
politics and the family. More recently, Frable (1997) reviewed the literature on a
variety of social identities, including gender. She noted an increasing focus in this
literature on the development of social identities. Separate developmental mod-
els have been offered for the development of racial identities (Cross 1991, 1995;
Phinney 1989), feminist identities (Downing & Roush 1985, Downing Hansen
Access provided by Michigan State University Library on 02/02/15. For personal use only.

2002, Hyde 2002, Moradi & Mezydlo Subich 2002), and sexual identities (Cass
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

1979, Diamond & Savin-Williams 2000). These developmental models are de-
signed to articulate the different kinds of origins and pathways that individuals
may follow in becoming identified with a particular social stratum. In addition,
these models sometimes specify different identity outcomes. Thus, Sellers and his
colleagues (Sellers et al. 1998) are interested in the substantively different racial
identities (e.g., nationalist, assimilationist) that African Americans may identify
with, while Henley and her colleagues are interested in the different kinds of femi-
nist identities (essentialist, liberal, etc.) to which feminists may attach importance
(Henley et al. 1998, 2000). These approaches assume that while each social iden-
tity may be multidimensional, each serves to define fundamentally or internally
represent one’s relationship to a collectivity or group (Brewer & Gardner 1996).
Thus identity—both in the Eriksonian and the social identity tradition—provides
a potential conceptual link between the psychology of the individual and social
structures, including gender.

Conceptual Tools for Psychology from Feminist Theory

Although identity provides a critical linking construct, psychologists must work
out how to think of gender within each empirical research program. We have seen
that in psychology these programs are dominated by notions of gender as sex
differences, as within-sex variations, and as social structures. Feminist theorists
have offered parallel accounts of gender that allow us to examine gender in terms of
the gendered meanings attached to social phenomena, as a social system of power
relations embedded in others, and as identities. These accounts in turn have proven
useful to some psychological research efforts, and we believe they can enhance
others.

Gender as Analytic Tool

First, historian Joan Scott (1988) argued that gender is not merely an empirical
“fact” attached to persons or even symbols, but an “analytical tool.” That is (among
other things), we can use the idea of gender to examine what other categories of
analysis may signify—their extra baggage or surplus meaning. For example, we
may (in history) think about how particular occupations (such as the military,
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 527

or medicine or nursing) are gendered, and how that gendering influences public
discourse, as well as individuals’ possibilities (Anker 1997). When, for example,
medical expertise is gendered male (and associated with doctors, as in the contem-
porary United States), it is viewed as requiring high levels of intellect, training,
and authority, but when it is gendered female (and associated with nurses, as in the
contemporary United States) it is viewed as requiring high levels of compassion,
nurturance, and patience. Opportunities for medical training for women depend
on the terms in which the occupation is viewed, as do the salaries for nurses and
doctors (see Harden 2001 for a discussion of the very different situation in Russia,
where the medical profession is gendered female, and is both low in status and in
Access provided by Michigan State University Library on 02/02/15. For personal use only.

pay).
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

In psychology, we can see the use of gender as an analytic tool in research that ex-
plores how “traits” may be themselves “gendered”—that is, viewed as strongly as-
sociated with masculinity and being male, or femininity and being female. Stewart
(1998) argued that the ‘big five” personality traits (extraversion, conscientious-
ness, agreeableness, emotional stability or neuroticism, and openness to experi-
ence), introduced as universal descriptors of personality, were “gendered”—that
is, associated with masculinity (extraversion or surgency, including dominance)
and femininity (particularly agreeableness, but perhaps also neuroticism and con-
scientiousness), and should be understood in that way. It is equally possible that
traits subsumed within these larger traits are also gendered—e.g., “openness to
experience” includes aspects of tolerance (probably gendered female), as well as
openness to risk-taking (probably gendered male). This notion has been explored
from the perspective of sex differences using both meta-analyses (Feingold 1994)
and empirical analysis of a large, representative sample (Goldberg et al. 1998).
In a series of studies, Twenge has gone further in using gender as an analytic
tool to understand these traits. First, under the assumption that changes in social
definitions of gender over time should be represented in changes in average levels
of certain individual personality variables, she has examined (Twenge 2001; see
also Twenge 1997) changes over time in psychological indicators of assertiveness
(which has increased for both men and women in the United States from 1931
to 1993). She notes, though, that the increases for women have been much more
substantial than those for men, and now assertiveness produces little sex difference
in most studies. In this research Twenge uses gender as an analytic tool, both in
the sense of considering the gendered social meanings of particular traits (such
as assertiveness) and in the sense of changes in those meanings over time. At
the same time she also examines gender in the sense of “sex differences” that is
more familiar to most psychologists. There is clearly much more to be done in
this area; many of the most familiar personality traits are deeply gendered in their
social meanings and symbolism, and psychologists have done little to consider the
implications of those meanings within and over time.
Some social psychologists have examined how social behaviors may be gen-
dered in the laboratory. In fact, we can see how certain laboratory paradigms are
themselves crucially gendered. For example, many psychologists have argued that
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

528 STEWART ¥ MCDERMOTT

mathematics performance is gendered male in our culture (Eccles & Jacobs 1986,
Eccles et al. 1990). Claude Steele and his colleagues (Spencer et al. 1999; see also
Aronson et al. 1998) actually manipulated the degree to which it was gendered,
by telling students that their performance on a laboratory task was diagnostic of
sex differences. Doing so differentially affected male and female students’ perfor-
mance, specifically causing female students to perform more poorly.
Finally, many social psychologists have demonstrated that being a solo group
representative (a woman, for example) in a group results in being treated as a
“token” (Kanter 1977, Spangler et al. 1978, Yoder & Sinnett 1985). In this research,
token status was found to be associated with higher levels of stigma [negative
Access provided by Michigan State University Library on 02/02/15. For personal use only.

stereotyping of the “minority” group by the majority group members (Kanter 1977,
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Taylor et al. 1978) as well as poorer performance by tokens (Sackett et al. 1991)].
Interestingly, when this paradigm was complicated by considering the usual social
status of group members, it turned out that gender was a crucial analytic tool. Thus,
Sekaquaptewa & Thompson (2002) have shown that individuals who are “token”
or “solo” in a group perform differently depending on their social status in general.
Men who are “solo” or “tokens” in otherwise female groups perform as well as
the women, while women solos perform less well than the men in their groups.
Performance suffers from the combination of circumstances involving both solo
status and lower social status, and not from only one of these factors. Only by
taking account of the social status associated with being male or being female
could this be made clear.

Gender as a Set of Embedded Power Relations

Recognition that gender is critically linked with social status opens up opportunities
to think about gender as a set of power relations rather than merely as characteristics
or features of individuals. Feminist theorist Catherine MacKinnon (1987) argued,
in a powerful essay on “Difference and Dominance,” that,

On the first day that matters, dominance was achieved, probably by force. By
the second day, division along the same lines had to be relatively firmly in
place. On the third day, differences were demarcated, together with social sys-
tems to exaggerate them in perception and in fact, because the systematically
differential delivery of benefits and deprivations required making no mistake
about who was who. Comparatively speaking, man has been resting ever since.
Gender might not even code as difference, might not even mean distinction
epistemologically, were it not for its consequences for social power. (p. 40)

Using MacKinnon’s reasoning would lead us to conclude that the presence of a

female token in an otherwise all-male group introduces a dimension of dominance
that can be easily misread as simply “difference.” MacKinnon’s perspective on
gender would lead us to think more carefully about how gendered meanings of so-
cial phenomena (including persons) and behavior include relations of dominance.
In her early account of the importance of social status in social identity formation,
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 529

social psychologist Erika Apfelbaum (1979/1999) showed how dominant groups

create conditions among subordinates both for obedience or compliance and for the
formation of resistant social identities. Equally, Aida Hurtado (1996) has articu-
lated how complex gender, race, and class relations create group identities for white
women and women of color, women and men of color, and men of color from differ-
ent ethnic group backgrounds. Her point in these analyses is to show how different
relationships to privilege (often held by white men, but also held by white women,
and men of color in varying ways and degrees) complicate the experience of sub-
ordination for members of these groups. These analyses build on an understanding
of how both gender and racial-ethnic identities are grounded in power relations.
Access provided by Michigan State University Library on 02/02/15. For personal use only.

Susan Fiske’s (1993) analysis of how gender operates within organizations

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

begins from an understanding of this kind. Fiske argues that in any relationship
defined by differential power (like gender), the dominant group (men in this case)
can afford to be oblivious to certain kinds of social cues, while the subordinate
group (women) cannot. Fiske shows that this means that dominants and subor-
dinates have very different levels and kinds of information about each other; in
this sense their asymmetrical power relation drives further asymmetries in their
experience and knowledge.
Frable (1990) made a similar kind of argument about the greater effort re-
quired of subordinates or “marginals” in social interaction with dominants. She
demonstrated that marginality or subordinate status, while associated with greater
knowledge or information, also requires considerably greater attention and effort,
thus drawing away effort from other areas of performance. This kind of analysis
suggests that the performance of dominants and subordinates on tasks for which
they are equally qualified might end up being different because of the different
number and kind of demands they are facing when they are performing in the
context of relationships with each other (see also Frable 1993, 1997).
In a study of patterns and consequences of workplace incivility in the federal
courts, Cortina et al. (2002) noted that dominant or subordinate status based on
occupation and gender played a role in who instigated and who was the target
of uncivil behavior. Incivility directed at female attorneys mostly came from their
male peers, as well as from judges of both sexes. Female attorneys were subordinate
to both groups, one in terms of gender (in the case of their male peers), the other
in terms of occupation (in the case of the judges). Men tended to experience
incivility primarily from judges, who were above them in terms of a hierarchy based
on occupation status; they did not experience incivility from their fellow female
attorneys. Consequently, women were more frequently targets, and experienced
more of the severe forms of incivility and almost all of the gendered uncivil acts
(in contrast to general incivility).
Examining a more commonplace gendered phenomenon, the objectification
of the female body, Fredrickson & Roberts (1997) demonstrated that a particu-
larly gendered everyday experience of power inequity is expressed physically as
well as psychologically. Their objectification theory contends that women have
learned to see themselves from an outsider’s perspective, that of a more powerful
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

530 STEWART ¥ MCDERMOTT

male person evaluating a less powerful person as an object. Objectification affects

women’s perception of their physical selves, creating a habit of monitoring the
body and seeing the self as an object. That habit in turn can result in experiences of
shame and anxiety and contribute to serious mental health problems such as eating
disorders.
In a very different domain, Nolen-Hoeksema and her colleagues (Nolen-
Hoeksema et al. 1999) have argued that the observed gender difference in de-
pressive symptoms may well result substantially from several features of women’s
subordinate social status (along with other factors). These include women’s per-
sistent experience of relative social devaluation, as well as their exposure to and
Access provided by Michigan State University Library on 02/02/15. For personal use only.

fear of violence against them—that is, the very power relations that characterize
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

gender and have been illuminated by Fiske, Frable, and Fredrickson & Roberts.

Multiplicity and Instability of Selves

Postmodern and poststructural theorists have argued that the idea of “identity” is
itself a “modernist” notion—one that assumes a stable, coherent individual agent
whose behavior flows from an essential core identity or self. They have historicized
this idea (noting that this understanding of the self or identity emerged in the En-
lightenment and took on increasing force in the twentieth century). Moreover, they
point out that if we take full account of the capacity of social processes to construct
individuals, we must take seriously the possibility that the presence of this idea in
the culture shaped individuals’ self-understandings, as well as theorists’ notions
of persons. These ideas have been recognized by many feminist theorists as poten-
tially threatening to the notion of politicized social identities like “women” (see,
e.g., Bordo 1993). Nevertheless, some aspects of these theories have been identi-
fied by feminist psychologists as potentially helpful in studying the psychology of
gender, among other things (see, e.g., Bohan 2002, Gergen 2001, Hare-Mustin &
Marecek 1990).
One particularly valuable use of this kind of theorizing is in an area already
rich with activity in psychology: research that has focused on the complexity and
situational variability in individuals’ selves or identities. While many psycholog-
ical studies point to some degree or kind of stability and unity of identity, many
others have explored the ways in which identities are multiple and overlapping,
sometimes conflicting or even mutually contradictory, as well as ways in which
identities are salient and visible in some situations and invisible in others. For ex-
ample, Markus & Nurius (1986) initiated a line of research into “possible selves,”
arguing that individuals not only have ideas of themselves, but also ideas of the
selves they hope to be and of the selves they hope not to become. Moreover, they
argue that these different “possible selves”—like “actual” selves—have motiva-
tional implications. People strive to become the selves they imagine they could
be, and strive to avoid becoming the selves they fear being (see, e.g., Knox et al.
2000). Equally, Linville (1985, 1987) demonstrated that individual differences in
self-complexity can be measured, and predict affective reactions to performance
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 531

feedback, among other outcomes. Many other psychologists have also examined
aspects of the organization of selves or identities (see, e.g., Rosenberg & Gara
1985, Woolfolk et al. 1995). In more recent research, Roccas & Brewer (2002)
have examined the complexity of social identities and their relationships to tol-
erance of outgroup members. In a series of studies, Deaux and her colleagues
have examined the ways in which identities are negotiated over time and across
situations (Deaux & Ethier 1998, Ethier & Deaux 1994). Merely to study these
processes is, of course, to recognize that identities are not completely stable and
“essential”; rather, they are emergent in social interactions that are sought out and
responded to by individual persons.
Access provided by Michigan State University Library on 02/02/15. For personal use only.

Finally, considerable empirical attention has recently been paid to the issue
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

of “bicultural identities” among individuals who feel shaped by and loyal to at

least two different cultures (LaFromboise et al. 1993). Such individuals are often
immigrants from one culture to another, but include second- and third-generation
descendents who remain identified with a “home” culture they may have only
experienced in a “different” culture. Theorists (e.g., Berry 1988) have articulated
alternative types of bicultural identities (e.g., assimilating, accommodating, alter-
nating, etc.), and researchers have explored the prominence of different kinds of
bicultural identities in immigrant groups (Smith et al. 2004), as well as the ways in
which bicultural persons experience different situations in terms of identity (Hong
et al. 2000). Although these researchers do not necessarily focus on gender or
feminist identities, and do not explicitly relate their research to feminist theories
(see Deaux & Stewart 2001 for an exception), it seems to us (as to Howard 2000)
that the developments in these two domains are parallel, compatible, and could
usefully inform one another. Explicitly linking them would help psychology to en-
gage more directly in the interdisciplinary dialogue on identity—including gender
identity—that is taking place in many fields, both contributing from our empirical
knowledge and drawing from the empirical and theoretical resources in other fields
(see, e.g., Brubaker & Cooper 2000).

Intersectionality: Theory and Examples

Although multiple and bicultural identities theories can be used to enrich the
psychological study of gender, even these theories miss some key aspects of its
complexity: the interaction between one’s many social identities (i.e., ethnic, class,
and gender) and the influences of different social structures on the construction
of these identities and relations between members of diverse social groups. Sev-
eral feminist sociologists (Collins 1994, King 1988), psychologists (Deaux &
Stewart 2001, Fine & Weis 1998, Hurtado 1996), critical legal theorists (Crenshaw
1991/1997), and philosophers (Spelman 1988) have advanced intersectionality as
a perspective that includes these interactive and structural elements in theories of
gender.
Three central tenets of intersectionality are especially useful to the psycho-
logical study of gender: (a) no social group is homogenous, (b) people must be
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

532 STEWART ¥ MCDERMOTT

located in terms of social structures that capture the power relations implied by
those structures, and (c) there are unique, nonadditive effects of identifying with
more than one social group. The first idea, the recognition that women (and men)
are not a homogenous group, may seem obvious to psychologists, in that one of the
purposes of psychological study has been to tease out individual differences. By
taking group categories as given, and focusing only on individual-level variance
within those groups, however, we avoid attending to how we have constructed
these group categories themselves. For example, Hurtado (1996) illustrates how
statistical analyses of pay and education parity data misrepresent women of color
by assuming to describe “women.” By averaging the overall lower pay and ed-
Access provided by Michigan State University Library on 02/02/15. For personal use only.

ucation status of women of color (here black, Hispanic, and Native American)
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

with white women’s, researchers produce a group average much higher than the
average for women of color. A more complex picture continues to emerge when
we note that black women have educational attainment at a level comparable to
white women’s average education level and higher than black men’s, yet they are
paid less than both of these other groups.
A related line of research shows that women readily acknowledge that women
as a group are discriminated against in their place of work, yet tend to deny having
ever suffered discrimination, such as pay inequity, personally. This phenomenon
is known as denial of personal discrimination. It can be accounted for by cognitive
and emotional biases such as a need to justify others’ and our own misfortunes in
disparate ways, in order to avoid cognitive dissonance (Crosby 1984, Crosby et al.
1993). These processes likely apply to women of all ethnic backgrounds. How-
ever, women who experience both gender-based discrimination and discrimination
based on ethnicity, face additional interpretive ambiguities. Consider, for exam-
ple, that African American women, in attempting to make accurate attributions of
causality for discrimination experiences, may find it impossible and/or fruitless to
identify race or gender as unique sources of those experiences (Essed 1990, Gay
& Tate 1998). For this reason, African American women may employ different
cognitive coping mechanisms than European American women when dealing with
experiences of discrimination, adding another layer of complexity to the causes,
perceptions, and effects of discrimination.
By analyzing pay and education parity data in terms of either gender (women
versus men) or race (blacks versus whites), the unique pattern of black women’s
experience is lost along with critical knowledge of how both race and gender
“work” to have behavioral consequences. Cole & Stewart (2001) examined how the
study of differences such as these can lead to invidious (offensively discriminating)
comparisons. Methods that rely on invidious comparisons and seek to legitimate
popular stereotypes are “more likely to cause harm and produce distorted and
partial findings” (Cole & Stewart 2001, p. 295). In contrast, research that takes
into account people’s multidimensional experience of the world is more likely
to produce new knowledge that presents a complex picture of difference instead
of oversimplifying it. Thinking intersectionally encourages us to examine how
grouping categories are constructed and to disaggregate them when we must. It does
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 533

not mean that every dimension of social identity needs to be studied simultaneously,
but rather that psychologists must be conscious of the consequences of our analytic
strategies.
The second integral part of intersectionality is the notion that people must be
understood as located within social structures. Locating a person socially demands
consideration of her material reality and the social forces that shape that reality,
particularly social disparities and power dynamics. Considering these contexts,
we not only find “new” populations overlooked in psychological study but also
a fuller picture of the social influences on any person’s psychology. Access to
power, in particular, is a frequently neglected but strongly influential aspect of
Access provided by Michigan State University Library on 02/02/15. For personal use only.

social structure. Again turning to the study of discrimination, it should be noted

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

that a critical factor shaping experiences and consequences of discrimination is

the victim’s access to power. While white women may be excluded from various
forms of power, their usually close proximity to white men (as brothers, fathers,
lovers, husbands) gives them greater access to the dominant holders of power than
a black woman would have who does not regularly interact with white men outside
of formal public situations and roles (Hurtado 1996). These unequal relations to
privilege create differences, differences that often are, but should not be, attributed
to culture or personality.
The third principle, that individuals identify with more than one social group
simultaneously and this identification produces unique consequences for the indi-
vidual, calls for a shift to acknowledging the ways social systems work together
to affect gender identities (King 1988). When we stop seeing “race and gender
as exclusive or separable categories” we create more complex categories with
richer pictures of gender within them (Crenshaw 1991/1997, p. 551). This holds
true for all social categories. An individual may share identification with one or
more categories with another person, but those identities in combination with
the social contexts they do not hold in common create a diverse range of expe-
riences, worldviews, and psychological mechanisms for dealing with the world.
A professional-class Latina lesbian mother may share the same social categories
and hold similar identities to a woman with a demographic background almost
mirroring hers but who identifies as straight. However, because of this difference,
their gender identities, roles, and interactions with the broader social world are
likely to be dissimilar in several crucial ways. Looking specifically at their moth-
ering experiences, they will likely employ different strategies to handle issues of
stigma, access to money and credit, transmission of values to their children, and
interaction with the legal system. In general, the psychological consequences of
everyday interactions will be qualitatively different, despite their shared identity as
a mother. For example, the role of co-mother is a unique aspect of lesbian families,
creating a new parenting model that Wilson (2001) found emphasizes equality in
parental responsibility and caretaking burden in a way that differs from that nego-
tiated in heterosexual couples. Mothering strategies also vary along other social
dimensions such as ethnic backgrounds, economic and social classes, or national
origins, to name a few.
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

534 STEWART ¥ MCDERMOTT

Patricia Hill Collins (1994) centered her analysis of mothering on women iden-
tified with marginalized ethnic groups and observed three themes key to their sense
of motherhood not shared by white women: survival, power, and cultural identity.
Hill Collins asserts that many Native American, African American, Hispanic, and
Asian American mothers need to work to ensure their children’s physical survival,
structure their mothering patterns based on a dominant group’s perception of them
as lacking power as mothers, and emphasize development of their children’s cul-
tural identity in ways that most white women do not because of their greater social
and economic privilege (Collins 1994).
Access provided by Michigan State University Library on 02/02/15. For personal use only.

Psychological Studies that Inform Our Study of Gender

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

from an Intersectional Perspective

Gender often acts as a shared attribute that shapes experiences, identities, emotive
expressions, and thought processes. But the experience of gender can generate
qualitatively different forms of these when gender is experienced in various so-
cial and material contexts. Articles in earlier editions of the Annual Review of
Psychology have explored gender’s influence on self-concept and identity and its
functioning within the broader social web (Banaji & Prentice 1994, Frable 1997).
The studies reviewed revealed many crucial insights into how social structures
affect group and individual psychologies. Generally, these studies investigated the
impact of one social variable at a time. When looking at more than one social
category, researchers kept categories such as gender, class, and ethnicity separated
from one another in terms of how they functioned within an individual or group.
Recently studies have examined gender, class, ethnicity, race, and sexuality not
as separate variables, nor as cumulative, but as intersecting categories that create
unique experiences and psychologies that are qualitatively different rather than a
mechanical combination of their individual effects.
Although nearly any psychological phenomenon might be fruitfully explored
from an intersectional perspective, those strongly influenced by social structures
or taking place in the context of social relations are especially likely prospects. On
both grounds, violence is an excellent candidate. Fine & Weis (1998) examined
the experiences of African American, white, Latina/o, poor and working class
women and men. Strikingly, the kinds of violence as well as the psychological
effects and methods for dealing with violence varied greatly between and among
these groups, depending on several contextual factors. For example, black men
experienced violence primarily outside of their homes and reacted to it by patrolling
borders around their neighborhoods, dealing with it in the streets. Black women
saw the same violence outside and dealt with it by keeping children inside and
working to keep violence out of the home. But inside was not necessarily a safe
space. Domestic violence was an experience shared by African American women,
Latinas, and white women, yet their racial, ethnic, and class identities strongly
influenced both their strategies for coping with it and their beliefs about it. For
example, African American women and Latinas resisted involving the police with
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 535

whom they had an ambivalent relationship; they felt the police did not protect them
appropriately because of their race/ethnicity.
While these two groups shared similar attitudes and experiences with the po-
lice and domestic violence, they dealt with violence at home differently. African
American women held the men who perpetrated the violence more responsible.
The Latinas interviewed expressed an allegiance to marriage and family that kept
them from disrupting the family by leaving violent situations. White working-class
women similarly held on to an ideal of a family based on white middle-class val-
ues, where the husband acts as provider and protector of the family, despite the
incongruity of this model with their lived experiences of family. Within this group,
Access provided by Michigan State University Library on 02/02/15. For personal use only.

women coped with domestic violence in two distinct ways. “Settled” women opted
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

to maintain secrecy about violence, protecting an ideal of a “good” white family.

“Hard-living” women removed themselves from their abusive situation and spoke
out about it, usually at great cost to themselves, especially economically. If Fine
& Weis had treated the women as a homogenous group dealing with domestic
violence (or had treated all ethnic group members as one group ignoring gender),
they would have missed the diversity of attitudes and behaviors that characterized
these women from different ethnic groups.
Press & Cole (1999) similarly took an intersectional approach when they ex-
amined women’s attitudes toward abortion, an area where women are frequently
divided into two camps—prolife or prochoice. The differences within these groups
are thus overlooked because of the emphasis on the differences between them. In
focus groups, Press & Cole spoke with prolife and prochoice women from a variety
of backgrounds about their attitudes toward abortion and media portrayals of the
abortion debate. They found these attitudes were influenced by the women’s social
and material contexts, though not in a uniform manner. Prolife women expressed
similar values and attitudes across classes, while prochoice women’s rationales for
their support of a woman’s right to abortion and value judgments of women who
had abortions varied across both class and race. For example, middle-class women
used a very individualistic perspective to think about abortion decision making,
emphasizing the depth of an individual woman’s feelings as the critical justification
for her decisions. In contrast, working class women justified decisions in terms of
the difficulties that women faced in navigating an authoritative and disempowering
health care system on their own behalf.
Identifying with specific social categories not only shapes attitudes but affects
behavior as well. Priming an ethnic or gender identity, Shih et al. (1999) found
Asian American women’s performance on tasks shifted depending on which iden-
tity, Asian or female, was more salient. When an Asian identity was made salient,
women scored higher on quantitative tests and lower on verbal tests than when
their female identity was made salient. In that scenario, the reverse outcomes were
produced. By considering individuals’ multiple identities, and focusing on how
they can play different roles, the researchers uncovered an important aspect of
stereotypes that might have been missed if the intersection of ethnicity and gender
were ignored.
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

536 STEWART ¥ MCDERMOTT

Examining race, ethnicity, and gender simultaneously does not always produce
the same kinds of findings. For example, as noted earlier, Gay & Tate (1998)
explored the intersection of ethnicity and race and gender by focusing on the
political attitudes of black women. They found that although many black women
strongly identify both as “women” and as “blacks,” when they are confronted with
political situations such as the Million Man March, O.J. Simpson trial, and Clarence
Thomas nomination, black women’s political attitudes were mostly determined by
their racial identity. Landrine et al. (1992) also found that racial identity affected
attitudes, including attitudes about gender stereotypes. In their study, white and
nonwhite female subjects rated themselves similarly on a measure of personality
Access provided by Michigan State University Library on 02/02/15. For personal use only.

characteristics associated with stereotypic gender roles. But the groups differed in
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

how they defined the terms. Thus, while a white woman and a nonwhite woman
might endorse an item (“passive”) at the same level, they were effectively endorsing
different items (“am laid back/easygoing” versus “don’t say what I really think”).
Mary Waters’ (1996) research on identity development in adolescents also
shows how culture and specifically relations to power from one’s position within a
culture can determine the very definition and value judgment of a stereotype. She
interviewed girls and boys of Caribbean descent, either second generation of immi-
grants or those who immigrated when they were young. Did they identify as black
American, ethnic American or as an immigrant? Class but not gender confounded
their choices. Poor and working-class young people identified more often as black
or immigrant. Middle-class teens identified more frequently as ethnic. While gen-
der did not influence the choice of these identities, the ramifications of these choices
varied greatly along gender lines. Boys’ identities were more rigid than girls’ iden-
tities. Girls were able to move between identities more easily and were less stig-
matized by their peers than their male counterparts who faced stricter enforcement
of their identities from peers. This study shows there are multiple possibilities
for identity development, and not a single one generalizable to all of society or
even the entirety of this particular group. Gender, class, immigrant status, parental
attitudes, and peer relationships all influenced these teens’ identity development.
Most of the research we have discussed has been comparative in nature. It is,
though, possible and sometimes necessary to examine a single, carefully defined
intersection without a “comparison group.” Cortina (2001) developed a measure of
workplace sexual harassment for Latinas based broadly on the Sexual Experiences
Questionnaire (Fitzgerald et al. 1988), which was itself developed primarily to
assess white women’s experiences. Cortina included indications of “sexual racism”
in the Sexual Experience Questionnaire-Latina, and noted differences in Latinas’
interpretations of particular items from the interpretations of Anglo respondents.
Espin (1997, 1999) has also focused some of her research on the experience of
Latina women in the United States. For example, in one study she examined how
a sample of Cuban immigrant lesbians experienced the intersection of ethnicity
and sexuality. She explicitly explored the ways in which, on the one hand, their
ethnic community encouraged them to be closeted, while, on the other hand, the
Anglo lesbian community forced ethnic assimilation. She concluded because of the
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 537

realities of racism and heterosexism that they have to confront, they are forced to
choose for their lives those alternatives that are more tolerable or less costly to them.
Some may choose to live in Miami among Cubans, even if that implies “staying in
the closet.” Others may choose to live in other areas of the country among Anglo
lesbians without feeling fully supported in terms of their Cuban identity (Espin
1997, p. 107). In this study, Espin illuminates the diverse experiences of Latina
lesbians by focusing on the variation within a single intersection of ethnicity,
gender, and sexuality.

How Can Psychologists Use Intersectionality

Access provided by Michigan State University Library on 02/02/15. For personal use only.

to Improve Research?
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

It is clear from this review that some psychologists have made productive use
of the notion of intersectionality to define critical qualitative differences in the
psychologies of certain groups of people in certain social locations. These groups
are sometimes simply identified in terms of their social location with respect to
two or more social statuses (e.g., white and male, African American and female,
female and lesbian, etc.), but often they are defined in terms of their degree of
identification with particular social locations. When we attend to this question of
identification, it becomes clear that individuals’ different identities are more or less
salient (though certainly not present or absent), depending on the context or issue.
Thus, as we have seen, Gay & Tate (1998) found that for African American women,
race is often the most salient identity in the domain of political opinion, while Fine
& Weis (1998) found that gender may be most salient in the domain of domestic
violence. Both of these studies examined the intersection of race and gender;
both also seemed to assume that the women were heterosexual. How do African
American lesbians think about and experience violence? Are they as likely as other
women to experience violence domestically? Are they more likely to experience it
in the public sphere? Thinking intersectionally allows us to ask questions like these,
which are important not only because they force us to examine the experiences
of previously understudied groups. The questions may also inform us about the
assumptions underlying the apparently general and generalizable knowledge we
think we have. If, for example, on average women’s experience of violence is
much more domestic than men’s, this may be crucially predicated on the fact
that on average women live in households with adult men. That fact—and the
importance of that fact to our data and our theories—may only emerge explicitly
when we ask questions from an intersectional perspective.
Often it seems that intersectionality requires us to examine all social locations
simultaneously—and that degree of complexity is daunting and frequently im-
practical. We believe that in fact intersectionality requires us to think about many
different kinds of social locations and identities that might intersect relevantly for
the behavior or experience we are studying. But we can and must make choices
about which locations and identities may be particularly relevant and/or particu-
larly understudied. In some cases we can make a reasonable judgment that a social
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

538 STEWART ¥ MCDERMOTT

location is irrelevant, because it seems logically to be so, and we have empirical

evidence that groups based on that location do not differ on the relevant dimension.
In many other cases, we have no empirical evidence about the particular location
within psychology. In those cases we can benefit from our neighbor disciplines.
Anthropologists, sociologists, and historians have often provided rich evidence
about the very social locations about which we do not know enough in psychology
even to make good first guesses.
We hope that as researchers adopt a perspective informed by intersectional-
ity they will rely more on multiple, especially qualitative, methods. Criticism of
mono-method bias in psychological research has a long history. Adopting intersec-
Access provided by Michigan State University Library on 02/02/15. For personal use only.

tionality as a research perspective may lead us to more inclusive methodological

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

choices. It may be noteworthy that although research programs implying intersec-

tionality as a perspective have sometimes relied solely on traditional methods in
psychology (e.g., experiments and surveys), many have employed mixed methods,
or more qualitative approaches. In fact, in research identifying different experi-
ences of high school as a joint function of race and gender, Stewart (2003) makes
a specific case for ethnographically informed methods when exploring understud-
ied topics in psychology. Bettie (2000) used the same method to examine high
school experiences at the intersections of gender, ethnicity, and social class. It is
partly for these reasons that Rabinowitz & Martin (2001) argued that “method-
ological pluralism is the strategy of choice in addressing the complexity of gender
arrangements in our society” (p. 51).

Gender’s Diverse Meanings in Psychology

Within mainstream psychology research, gender does not have one meaning. It
sometimes signifies sex differences, sometimes intrasex variability, and sometimes
gendered social roles and institutions. Locating gender in both individuals and
social structures—however persuasive—implies a need for concepts that bridge
those levels of analysis.
The notion of identity, both as articulated in Erikson’s (1950/1963) compre-
hensive theory and in Tajfel’s (1981) social identity theory, provides a helpful
conceptual tool for understanding both how social experiences are inscribed on
individual personalities, and how individuals define themselves in relation to social
groups, strata, and institutions. Thus we have tried to show that the development of
theories of identity in personality and social psychology in fact can serve to enrich
and deepen our understanding of gender. In particular, identity theorists have high-
lighted the importance of social context and point to the need for psychologists
to give greater consideration to this influential factor. Similarly, feminist theorists
have recognized the significance of social context in individuals’ construction and
use of gendered identities. Consequently, they have developed some additional
conceptual resources that are compatible with the identity theories and with the
development of understanding of gender in psychology. These include the use of
gender as an analytic tool (rather than a social category), the analysis of gender
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 539

as a system of power relations embedded in other systems of power relations, and

analysis of the multiplicity and instability of identity.
All of these tools can be employed to examine the different psychologies con-
structed at the intersections of different power relations (gender, race, class, sex-
uality, and so forth). The challenge for psychology is to incorporate all of these
conceptual tools into our research designs and practices. Our goal in this review
was to demonstrate that many creative efforts along these lines have already yielded
fruit, and to plant the seeds for an even richer harvest.
Access provided by Michigan State University Library on 02/02/15. For personal use only.

ACKNOWLEDGMENTS
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

We are grateful to Jennifer Churchwell, Laura Citrin, Elizabeth R. Cole, Ngaire

Donaghue, Andrea Dottolo, Breanne Fahs, Carla Grayson, Jana Haritatos, Kathi
Miner-Rubino, Susan Nolen-Hoeksema, Perry Silverschanz, Allison G. Smith,
Cynthia Torges, David G. Winter, and Carolyn Zahn-Waxler for helpful sugges-
tions, discussions, and comments on drafts of this manuscript.

The Annual Review of Psychology is online at http://psych.annualreviews.org

LITERATURE CITED
Anker R. 1997. Theories of occupational segre- RC Annis, pp. 41–52. Amsterdam: Swets &
gation by sex: an overview. Int. Labour Rev. Zeitlinger
136(3):315–40 Bettie J. 2000. Women without class: chicas,
Apfelbaum E. 1999 (1979). Relations of dom- cholas, trash, and the presence/absence of
ination and movements for liberation: an class identity. Signs 26:1–35
analysis of power between groups. In Social Block J. 1984. Sex Role Identity and Ego De-
Psychology of Intergroup Relations, ed. WG velopment. San Francisco: Jossey-Bass
Austin, S Worchel, pp. 188–204. Belmont, Bohan J. 2002. Sex differences and/in the self:
MA: Brooks/Cole classic themes, feminist variations, postmod-
Aronson J, Quinn DM, Spencer SJ. 1998. ern challenges. Psychol. Women Q. 26:74–88
Stereotype threat and the academic under- Bordo S. 1993. Feminism, postmodernism, and
performance of minorities and women. See gender skepticism. In Unbearable Weight,
Swim & Stangor 1998, pp. 83–103 pp. 215–43. Berkeley: Univ. Calif. Press
Banaji MR, Prentice DA. 1994. The self in so- Brewer MB, Gardner W. 1996. Who is this
cial contexts. Annu. Rev. Psychol. 45:297– “we”? Levels of collective identity and self
332 representations. J. Personal. Soc. Psychol.
Benjamin J. 1988. The Bonds of Love: Psycho- 71:83–93
analysis, Feminism, and the Problem of Dom- Brody L. 1999. Gender, Emotion and the Fam-
ination. New York: Pantheon ily. Cambridge, MA: Harvard Univ. Press
Berry JW. 1988. Acculturation and psycholog- Bronfenbrenner U. 1986. Ecology of the family
ical adaptation: conceptual overview. In Eth- as context for human development: research
nic Psychology: Research and Practice with perspectives. Dev. Psychol. 22(6):723–42
Immigrants, Refugees, Native Peoples, Eth- Brubaker R, Cooper F. 2000. Beyond identity.
nic Groups and Sojourners, ed. JW Berry, Theory Soc. 29:1–47
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

540 STEWART ¥ MCDERMOTT

Buss D. 1995. Psychological sex differences: Cross WE. 1995. The psychology of nigres-
origins through natural selection. Am. Psy- cence: revising the Cross model. In Hand-
chol. 50:164–68 book of Multicultural Counseling, ed. JG
Cass VC. 1979. Homosexuality identity forma- Ponterotto, JM Casas, LA Suzuki, CM
tion: a theoretical model. J. Homosex. 4:219– Alexander, pp. 93–122. Thousand Oaks, CA:
35 Sage
Chodorow N. 1978. The Reproduction of Moth- Deaux K, Ethier KA. 1998. Negotiating social
ering: Psychoanalysis and the Sociology of identity. See Swim & Stangor 1998, pp. 301–
Gender. Berkeley: Univ. Calif. Press 23
Chodorow N. 1994. Femininities, Masculini- Deaux K, Stewart AJ. 2001. Framing gendered
ties, Sexualities: Freud and Beyond. Lexing- identities. See Unger 2001, pp. 83–97
Access provided by Michigan State University Library on 02/02/15. For personal use only.

ton: Univ. Kentucky Press Diamond L, Savin-Williams RC. 2000. Ex-

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Cole E, Stewart AJ. 2001. Invidious compar- plaining diversity in the development of
isons: imagining a psychology of race and same-sex sexuality among young women. J.
gender beyond differences. Polit. Psychol. Soc. Issues 56(2):297–314
22:293–308 Downing NE, Roush KL. 1985. From passive
Cole E, Zucker AN, Duncan LE. 2001. Chang- acceptance to active commitment: a model
ing society, changing women (and men). See of feminist identity development for women.
Unger 2001, pp. 410–24 Couns. Psychol. 13:695–709
Collins PH. 1994. Shifting the center: race, Downing Hansen N. 2002. Reflections on fem-
class and feminist theorizing about mother- inist identity development: implications for
hood. In Representations of Motherhood, ed. theory, measurement, and research. Couns.
D Bassin, M Honey, MM Kaplan, pp. 56–74. Psychol. 30(1):87–95
New Haven, CT: Yale Eagly A. 1994. On comparing women and men.
Cortina LM. 2001. Assessing sexual harass- Fem. Psychol. 4:513–22
ment among Latinas: development of an in- Eccles J, Jacobs JE. 1986. Social forces
strument. Cult. Divers. Ethn. Minor. Psychol. shape math attitudes and performance. Signs
7(2):164–81 11:367–80
Cortina LM, Lonsway KA, Magley VJ, Free- Eccles J, Jacobs JE, Harold RD. 1990. Gen-
man LV, Collinsworth LL, et al. 2002. What’s der role stereotypes, expectancy effects, and
gender got to do with it? Incivility in the fed- parents’ socialization of gender differences.
eral courts. Law Soc. Inq. 27:235–70 J. Soc. Issues 46(2):183–201
Crenshaw K. 1997 (1991). Beyond racism and Erikson EH. 1963 (1950). Childhood and Soci-
misogyny: black feminism and 2 Live Crew. ety. New York: Norton. 2nd ed.
In Women Transforming Politics: An Alter- Erikson EH. 1968. Identity: Youth and Crisis.
native Reader, ed. CJ Cohen, KB Jones, JC New York: Norton
Tronto, pp. 549–68. New York: NY Univ. Erikson EH. 1974. Once more the inner space:
Press letter to former student. In Women and Anal-
Crosby F. 1984. The denial of personal discrim- ysis: Dialogues on Psychoanalytic Views of
ination. Am. Behav. Sci. 27:371–86 Femininity, ed. J Strouse, pp. 365–87. New
Crosby F, Cordova DI, Jaskar K. 1993. On the York: Dell
failure to see oneself as disadvantaged: Cog- Espin OM. 1997. Latina Realities: Essays on
nitive and emotional components. In Group Healing, Migration, and Sexuality. Boulder,
Motivation: Social Psychological Perspec- CO: Westview
tives, ed. MA Hogg, D Abrams, pp. 87–104. Espin OM. 1999. Women Crossing Boundaries:
New York: Harvester Wheatsheaf A Psychology of Immigration and Transfor-
Cross WE. 1991. Shades of black. Philadelphia: mations of Sexuality. New York: Routledge
Temple Univ. Press Essed P. 1990. Everyday Racism: Reports from
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 541

Women of Two Cultures. Claremont, CA: ual, and class identities. Annu. Rev. Psychol.
Hunter House 48:139–62
Ethier KA, Deaux K. 1994. Negotiating social Franz C, White KM. 1985. Individuation
identity when contexts change: maintaining and attachment in personality development:
identification and responding to threat. J. Per- extending Erikson’s theory. J. Personal.
sonal. Soc. Psychol. 67(2):243–51 53:224–56
Favreau O. 1993. Do the n’s justify the means? Fredrickson BL, Roberts T. 1997. Objectifica-
Null hypothesis testing applied to sex and tion theory: toward understanding women’s
other differences. Can. Psychol. 34:64–78 lived experiences and mental health risks.
Favreau O. 1997. Sex and gender comparisons: Psychol. Women Q. 21:173–206
Does null hypothesis testing create a false Freud S. 1961 (1930). Civilization and its dis-
Access provided by Michigan State University Library on 02/02/15. For personal use only.

dichotomy? Fem. Psychol. 7:63–81 contents. In The Standard Edition of the

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Feingold A. 1994. Gender differences in per- Complete Psychological Works of Sigmund

sonality: a meta-analysis. Psychol. Bull. Freud, ed. J Strachey, 21:57–145. London:
116:429–56 Hogarth
Fine M, Weis L. 1998. The Unknown City: Lives Fromm E. 1947. Man for Himself. New York:
of Poor and Working-Class Young Adults. Holt, Rinehart & Winston
Boston, MA: Beacon Gay C, Tate K. 1998. Doubly bound: the impact
Fiske ST. 1993. Controlling other people: the of gender and race on the politics of black
impact of power on stereotyping. Am. Psy- women. Polit. Psychol. 19:169–84
chol. 48:621–28 Gergen M. 2001. Feminist Reconstructions in
Fiske ST, Stevens LE. 1993. What’s so special Psychology: Narrative, Gender and Perfor-
about sex? Gender stereotyping and discrim- mance. Thousand Oaks, CA: Sage
ination. In Gender Issues in Contemporary Goldberg LR, Sweeney D, Merenda PF, Hughes
Society, ed. S Oskamp, M Costanzo, pp. 173– JE. 1998. Demographic variables and person-
96. Thousand Oaks, CA: Sage ality: the effects of gender, age, education,
Fitzgerald LF, Shullman SL, Bailey N, Richards and ethnic/racial status on self-descriptions
M, Swecker J, et al. 1988. The incidence of personality attributes. Personal. Individ.
and dimensions of sexual harassment in Differ. 24:393–403
academia and the workplace. J. Vocat. Be- Goodwin SA, Fiske ST. 2001. Power and gen-
hav. 32:152–75 der: the double-edged sword of ambivalence.
Fivush R, Buckner JP. 2000. Gender, sadness See Unger 2001, pp. 358–66
and depression: the development of emo- Gurin P. 1985. Women’s gender consciousness.
tional focus through gendered discourse. In Public Opin. Q. 49:143–63
Gender and Emotion: Social Psychological Gurin P. 1987. The political implications of
Perspectives, ed. AH Fisher, pp. 232–53. women’s statuses. In Spouse, Parent, Worker:
New York: Cambridge Univ. Press On Gender and Multiple Roles, ed. FJ
Flax J. 1990. Thinking Fragments: Psychoanal- Crosby, pp. 167–98. New Haven, CT: Yale
ysis, Feminism, and Postmodernism in the Univ. Press
Contemporary West. Berkeley: Univ. Calif. Gurin P, Markus H. 1989. Cognitive conse-
Press quences of gender identity. In The Social
Frable DES. 1990. Marginal and mindful: de- Identity of Women, ed. S Skevington, D
viants in social interactions. J. Personal. Soc. Baker, pp. 152–72. Newbury Park, CA: Sage
Psychol. 59(1):140–49 Gurin P, Miller H, Gurin G. 1980. Stratum iden-
Frable DES. 1993. Dimensions of marginality: tification and consciousness. Soc. Psychol.
Distinctions among those who are different. Q. 43:30–47
Personal. Soc. Psychol. Bull. 19(4):370–80 Gutek G, Done RS. 2001. Sexual harassment.
Frable DES. 1997. Gender, racial, ethnic, sex- See Unger 2001, pp. 367–87
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

542 STEWART ¥ MCDERMOTT

Harden J. 2001. ‘Mother Russia’ at work. Eur. Countries. Cambridge, MA: Harvard Univ.
J. Women’s Stud. 8(2):181–99 Press
Hare-Mustin R, Marecek J, eds. 1990. Making Kanter RM. 1977. Men and Women of the Cor-
a Difference: Psychology and the Construc- poration. New York: Basic Books
tion of Gender. New Haven, CT: Yale Univ. King DK. 1988. Multiple jeopardy, multiple
Press consciousness: the context of a black fem-
Helson R, Stewart AJ, Ostrove J. 1995. Identity inist ideology. Signs 14:42–72
in three cohorts of women. J. Personal. Soc. Klein M. 1965. Contributions to Psycho-
Psychol. 69:544–57 Analysis, 1921–1945. New York: McGraw-
Henley NM, Meng K, O’Brien D, McCarthy Hill
WJ, Sockloskie RJ. 1998. Developing a scale Kluckhohn C, Murray HA, Schneider DM.
Access provided by Michigan State University Library on 02/02/15. For personal use only.

to measure the diversity of feminist attitudes. 1953. Personality in Nature, Culture and So-
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Psychol. Women. Q. 22(3):317–48 ciety. New York: Knopf

Henley NM, Spalding LR, Kosta A. 2000. De- Knox M, Funk J, Elliott R, Bush EG. 2000.
velopment of the short form of the Femi- Gender differences in adolescents’ possible
nist Perspectives Scale. Psychol. Women Q. selves. Youth Soc. 31(3):287–309
24(3):254–56 LaFromboise T, Hardin LKC, Gerton J. 1993.
Hong Y, Morris MW, Chiu C, Chiu V. 2000. Psychological impact of biculturalism: evi-
Multicultural minds: a dynamic construc- dence and theory. Psychol. Bull. 114(3):395–
tivist approach to culture and cognition. Am. 412
Psychol. 55(7):709–20 Landrine H, Klonoff EA, Brown-Collins A.
Honigman JJ. 1967. Personality in Culture. 1992. Cultural diversity and methodology in
New York: Harper & Row feminist psychology: critique, proposal, em-
Horney K. 1939. New Ways in Psychoanalysis. pirical example. Psychol. Women Q. 16:145–
New York: Norton 64
House JS. 1981. Social structure and personal- Levy J, Heller W. 1992. Gender differences
ity. In Social Psychology: Sociological Per- in human neuropsychological function. In
spectives, ed. M Rosenberg, RH Turner, Sex Differentiation: Handbook of Behavioral
pp. 525–61. New York: Basic Books Neurobiology, ed. A Gerall, H Moltz, IL
Howard J. 2000. Social psychology of identi- Ward, pp. 245–74. New York: Plenum
ties. Annu. Rev. Sociol. 26:367–93 Linville PW. 1985. Self-complexity and affec-
Hurtado A. 1996. The Color of Privilege: Three tive extremity: Don’t put all your eggs in one
Blasphemies on Race and Feminism. Ann Ar- cognitive basket. Soc. Cogn. 3:94–120
bor: Univ. Mich. Press Linville PW. 1987. Self-complexity as a cog-
Hyde JS. 2002. Feminist identity development: nitive buffer against stress-related illness
the current state of theory, research, and prac- and depression. J. Personal. Soc. Psychol.
tice. Couns. Psychol. 30(1):105–10 52:663–76
Inkeles A, Levinson DJ. 1954. National char- Maccoby E. 1998. The Two Sexes: Growing Up
acter: the study of modal personality and so- Apart, Coming Together. Cambridge, MA:
ciocultural systems. In Handbook of Social Belknap
Psychology, ed. G Lindzey, 2:977–1020. Maccoby E, Jacklin C. 1974. The Psychology
Reading, MA: Addison-Wesley of Sex Differences. Stanford, CA: Stanford
Inkeles A, Levinson DJ. 1969. National char- Univ. Press
acter. In Handbook of Social Psychology, ed. MacKinnon C. 1987. Feminism Unmodified:
G Lindzey, E Aronson, 4:418–506. Reading, Discourses on Life and Law. Cambridge,
MA: Addison-Wesley. Rev. ed. MA: Harvard Univ. Press
Inkeles A, Smith DH. 1974. Becoming Mod- Markus H, Nurius P. 1986. Possible selves. Am.
ern: Individual Change in Six Developing Psychol. 41:954–69
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

GENDER IN PSYCHOLOGY 543

Martin JR. 1994. Methodological essentialism, Rosenberg S, Gara MA. 1985. The multiplicity
false difference, and other dangerous traps. of personal identity. In Review of Personality
Signs 19(3):630–57 and Social Psychology, ed. P Shaver, 6:87–
McClelland DC. 1967. The Achieving Society. 115. Beverly Hills, CA: Sage
New York: Free Press Ryff C. 1987. The place of personality and so-
Moradi B, Mezydlo Subich L. 2002. Femi- cial structure research in social psychology.
nist identity development measures: compar- J. Personal. Soc. Psychol. 53(6):1192–202
ing the psychometrics of three instruments. Sackett PR, DuBois CLZ, Noe AW. 1991. To-
Couns. Psychol. 30(1):66–86 kenism in performance evaluation: the effects
Nolen-Hoeksema S, Larson J, Grayson C. 1999. of work group representation on male-female
Explaining the gender difference in depres- and white-black differences in performance
Access provided by Michigan State University Library on 02/02/15. For personal use only.

sive symptoms. J. Personal. Soc. Psychol. ratings. J. Appl. Psychol. 76:263–67

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

77:1061–72 Scott J. 1988. Gender and the Politics of His-

Parsons T. 1955. Family, Socialization and In- tory. New York: Columbia Univ. Press
teraction Process. New York: Free Press Sekaquaptewa D, Thompson M. 2002. The dif-
Parsons T. 1967. Sociological Theory and Mod- ferential effects of solo status on members of
ern Society. New York: Free Press high- and low-status groups. Personal. Soc.
Phinney JS. 1989. Stages of ethnic identity in Psychol. Bull. 28(5):694–707
minority group adolescents. J. Early Adolesc. Sellers RM, Smith MM, Shelton JN, Rowley
9:34–49 SAJ, Chavous TM. 1998. Multidimensional
Pimlott-Kubiak S, Cortina L. 2004. Gender, vic- model of racial identity: a reconceptualiza-
timization and outcomes: reconceptualizing tion of African American racial identity. Per-
risk. J. Clin. Consult. Psychol. In press sonal. Soc. Rev. 2(1):18–39
Powlishta KK, Sen MG, Serbin LA, Poulin- Shih M, Pittinsky TL, Ambady N. 1999. Stereo-
Dubois D, Eichstedt JA. 2001. From infancy type susceptibility: identity salience and
through middle childhood: the role of cogni- shifts in quantitative performance. Psychol.
tive and social factors in becoming gendered. Sci. 10:80–83
See Unger 2001, pp. 116–32 Shweder RA. 1991. Thinking through Cultures:
Press AL, Cole ER. 1999. Speaking of Abor- Expeditions in Cultural Psychology. Cam-
tion: Television and Authority in the Lives of bridge, MA: Harvard Univ. Press
Women. Chicago, IL: Univ. Chicago Press Smith AG, Stewart AJ, Winter DG. 2004. Close
Rabinowitz VC, Martin D. 2001. Choices and encounters with the Midwest: forming iden-
consequences: methodological issues in the tity in a bicultural context. Polit. Psychol. In
study of gender. See Unger 2001, pp. 29– press
52 Spangler E, Gordon MA, Pipkin PM. 1978. To-
Riger S. 1991. Gender dilemmas in sexual ha- ken women: an empirical test of Kanter’s hy-
rassment policies and procedures. Am. Psy- pothesis. Am. J. Sociol. 84:160–70
chol. 46:497–505 Spelman EV. 1988. Inessential Woman: Prob-
Riger S. 1992. Epistemological debates, fem- lems of Exclusion in Feminist Thought.
inist voices: science, social values, and Boston, MA: Beacon
the story of women. Am. Psychol. 47:730– Spencer SJ, Steele CM, Quinn DM. 1999.
40 Stereotype threat and women’s math perfor-
Riger S. 2000. Transforming Psychology: Gen- mance. J. Exp. Soc. Psychol. 35(1):4–28
der in Theory and Practice. New York: Ox- Stewart AJ. 1998. Doing personality research:
ford How can feminist theories help? In The Gen-
Roccas S, Brewer M. 2002. Social identity com- der and Psychology Reader, ed. BM Clinchy,
plexity. Personal. Soc. Psychol. Rev. 6(2):88– JK Norem, pp. 54–68. New York: NY Univ.
106 Press
 18 Nov 2003 15:31 AR AR207-PS55-18.tex AR207-PS55-18.sgm LaTeX2e(2002/01/18) P1: GCE

544 STEWART ¥ MCDERMOTT

Stewart AJ. 2003. Gender, race, and gen- Unger RK. 2001. Handbook of the Psychology
eration in a Midwest high school: using of Women and Gender. New York: Wiley
ethnographically-informed methods in psy- Veroff J. 1983. Contextual determinants of per-
chology. Psychol. Women Q. 27:1–11 sonality. Personal. Soc. Psychol. Bull. 9:331–
Stewart AJ, Healy JM. 1989. Linking individual 43
development and social change. Am. Psychol. Waters MC. 1996. The intersection of gen-
44:30–42 der, race and ethnicity in identity develop-
Strouse J, ed. 1974. Women and Analysis: Dia- ment of Caribbean American teens. In Urban
logues on Psychoanalytic Views of Feminin- Girls: Resisting Stereotypes, Creating Iden-
ity. New York: Dell tities, ed. B Leadbeater, N Way, pp. 65–81.
Sullivan HS. 1953. The Interpersonal Theory of New York: New York Univ. Press
Access provided by Michigan State University Library on 02/02/15. For personal use only.

Psychiatry. New York: Norton White JW, Donat PLN, Bondurant B. 2001.
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Swim JK, Stangor C, eds. 1998. Prejudice: The A developmental examination of violence
Target’s Perspective. San Diego, CA: Aca- against girls and women. See Unger 2001,
demic pp. 343–57
Tajfel H. 1957. Value and the perceptual judg- Whiting B. 1963. Six Cultures. New York: Wi-
ment of magnitude. Psychol. Rev. 64:192– ley
204 Wilson CM. 2001. The creation of mother-
Tajfel H. 1981. Human Groups and Social Cat- hood: exploring the experiences of lesbian
egories. London: Cambridge Univ. Press co-mothers. J. Fem. Fam. Therapy 12:21–44
Taylor SE, Fiske ST, Etcoff NL, Ruderman Woolfolk RL, Novalany J, Gara MA, Allen
AJ. 1978. Categorical and contextual bases LA, Polino M. 1995. Self-complexity, self-
of person memory and stereotyping. J. Per- evaluation, and depression: an examination
sonal. Soc. Psychol. 36:778–93 of form and content within the self-schema.
Twenge J. 1997. Changes in masculine and fem- J. Personal. Soc. Psychol. 68:1108–20
inine traits over time: a meta-analysis. Sex Yoder J, Sinnett J. 1985. Is it all in the numbers?
Roles 36:305–25 A case study of tokenism. Psychol. Women Q.
Twenge J. 2001. Changes in women’s as- 9:413–18
sertiveness in response to status and roles: a Zahn-Waxler C. 2000. The development of em-
cross-temporal meta-analysis, 1931–1993. J. pathy, guilt, and internalization of distress:
Personal. Soc. Psychol. 81:133–45 implications for gender differences in inter-
Unger RK. 1989. Sex, gender and epistemol- nalizing and externalizing problems. In Anx-
ogy. In Gender and Thought: Psychologi- iety, Depression and Emotion, ed. RJ David-
cal Perspectives, ed. M Crawford, M Gentry, son, pp. 222–65. New York: Oxford Univ.
pp. 17–35. New York: Springer-Verlag Press
P1: FDS
December 9, 2003 14:49 Annual Reviews AR207-FM

Annual Review of Psychology

Volume 55, 2004

CONTENTS
Frontispiece—Walter Mischel xvi
PREFATORY
Access provided by Michigan State University Library on 02/02/15. For personal use only.
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Toward an Integrative Science of the Person, Walter Mischel 1

LEARNING AND MEMORY PLASTICITY
On Building a Bridge Between Brain and Behavior, Jeffrey D. Schall 23
The Neurobiology of Consolidations, Or, How Stable is the Engram?,
Yadin Dudai 51
BRAIN IMAGING/COGNITIVE NEUROSCIENCE
Understanding Other Minds: Linking Developmental Psychology and
Functional Neuroimaging, R. Saxe, S. Carey, and N. Kanwisher 87
SLEEP
Hypocretin (Orexin): Role in Normal Behavior and Neuropathology,
Jerome M. Siegel 125
SPEECH PERCEPTION
Speech Perception, Randy L. Diehl, Andrew J. Lotto, and Lori L. Holt 149
DEPTH, SPACE, AND MOTION
Visual Mechanisms of Motion Analysis and Motion Perception,
Andrew M. Derrington, Harriet A. Allen, and Louise S. Delicato 181
ATTENTION AND PERFORMANCE
Cumulative Progress in Formal Theories of Attention, Gordon D. Logan 207
MEMORY
The Psychology and Neuroscience of Forgetting, John T. Wixted 235
FORM PERCEPTION AND OBJECT RECOGNITION
Object Perception as Bayesian Inference, Daniel Kersten,
Pascal Mamassian, and Alan Yuille 271
ADULTHOOD AND AGING
Development in Midlife, Margie E. Lachman 305

vii
P1: FDS
December 9, 2003 14:49 Annual Reviews AR207-FM

viii CONTENTS

DEVELOPMENT IN SOCIETAL CONTEXT

The Intergenerational Transfer of Psychosocial Risk: Mediators of
Vulnerability and Resilience, Lisa A. Serbin and Jennifer Karp 333
DEVELOPMENT IN THE FAMILY
Development in the Family, Ross D. Parke 365
SCHIZOPHRENIA AND RELATED DISORDERS
Schizophrenia: Etiology and Course, Elaine Walker, Lisa Kestler,
Annie Bollini, and Karen M. Hochman 401
Access provided by Michigan State University Library on 02/02/15. For personal use only.

SUBSTANCE ABUSE DISORDERS

Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Clinical Implications of Reinforcement as a Determinant of Substance

Use Disorders, Stephen T. Higgins, Sarah H. Heil,
and Jennifer Plebani Lussier 431
Motivational Influences on Cigarette Smoking, Timothy B. Baker,
Thomas H. Brandon, and Laurie Chassin 463
INFERENCE, PERSON PERCEPTION, ATTRIBUTION
Self-Knowledge: Its Limits, Value, and Potential for Improvement,
Timothy D. Wilson and Elizabeth W. Dunn 493
GENDER
Gender in Psychology, Abigail J. Stewart and Christa McDermott 519
MASS MEDIA
Mediated Politics and Citizenship in the Twenty-First Century,
Doris Graber 545
NONVERBAL AND VERBAL COMMUNICATION
The Internet and Social Life, John A. Bargh and
Katelyn Y.A. McKenna 573
SOCIAL INFLUENCE
Social Influence: Compliance and Conformity, Robert B. Cialdini
and Noah J. Goldstein 591
SMALL GROUPS
Group Performance and Decision Making, Norbert L. Kerr
and R. Scott Tindale 623
PERSONALITY PROCESSES
Creativity, Mark A. Runco 657
P1: FDS
December 9, 2003 14:49 Annual Reviews AR207-FM

CONTENTS ix

PSYCHOLOGY AND CULTURE

Psychology and Culture, Darrin R. Lehman, Chi-yue Chiu,
and Mark Schaller 689
TEACHING OF SUBJECT MATTER
Teaching of Subject Matter, Richard E. Mayer 715
PERSONALITY AND COPING STYLES
Coping: Pitfalls and Promise, Susan Folkman and Judith Tedlie Moskowitz 745
Access provided by Michigan State University Library on 02/02/15. For personal use only.

SURVEY METHODOLOGY
Annu. Rev. Psychol. 2004.55:519-544. Downloaded from www.annualreviews.org

Survey Research and Societal Change, Roger Tourangeau 775

Human Research and Data Collection via the Internet,
Michael H. Birnbaum 803

INDEXES
Author Index 833
Subject Index 877
Cumulative Index of Contributing Authors, Volumes 45–55 921
Cumulative Index of Chapter Titles, Volumes 45–55 926

ERRATA
An online log of corrections to Annual Review of Psychology chapters
may be found at http://psych.annualreviews.org/errata.shtml