Beyond Meta-Analyses: Other Possible Sex Differences
What's the Difference Anyway?
Although meta-analyses have helped synthesize huge amounts of research literature on sex differences, there are many important sex differences that have not been summarized by meta-analyses. Certain kinds of sex differences (e.g., sex differences in mental disorders) cannot always be easily tallied with the d statistic, for the behaviors under study are not continuous. For example, people are either clinically depressed or they are not. Although sex differences in the incidence of depression may be real, such differences are better captured by sex ratios (i.e., the ratio of men to women who suffer from depression) than by d statistics. In this final section some additional ways in which males and females may differ are considered. Are there sex differences in mental illness? Do men and women experience emotions somewhat differently? Are the self-concepts of men and women organized differently? Do boys and girls differ in their friendship patterns and styles of play?
Are There Sex Differences in Mental and Behavioral Disorders?
Table 1.5 summarizes evidence on sex differences in various kinds of behavior problems and mental disorders, in both children and adults (Hartung & Widiger, 1998). The information comes from the Diagnostic and Statistical Manual of Mental Disorders, published by the American Psychiatric Association (the most recent version is the DSM-IV, published in 1994, and its revision, published in 2000).
As Table 1.5 shows, boys are much more likely than girls to suffer from mental retardation, reading disorders, stuttering, autism, Tourette's syndrome (a neurological condition characterized by compulsive movements and, sometimes, abusive verbal exclamations), and attention
deficit disorders. In adolescence and adulthood, males are more likely than females to abuse substances, including alcohol, marijuana, amphetamines, opiate drugs, and hallucinogens. Among adult mental disorders that show a tilt in favor of men are many sexual disorders, gender identity disorders, and antisocial, compulsive, schizoid, and narcissistic personality disorders.2 Disorders that show a tilt in favor of women are major and minor depressions, phobias, generalized anxiety disorders, conversion disorders, dissociative disorders ("multiple personalities"), eating disorders, and borderline and histrionic personality disorders.3
Considerable recent research has focused on sex differences in depression, a disorder that is sometimes called the common cold of mental illness because it occurs so frequently. Many studies suggest that clinical depression is at least twice as common in women as in men (Nolen-Hoeksema, 2002). The reasons for women's higher levels of depression continue to be debated and probably include many factors, such as greater sexual abuse and stress, higher tendencies in women than men to ruminate and introspect, higher levels of neuroticism, and sex-linked genetic and hormonal factors (Hankin & Abramson, 2001). San Diego State University psychologist Jean Twenge and University of Michigan psychologist Susan Nolen-Hoeksema (2002) recently conducted a meta-analysis of self-reported depression symptoms in over 61,000 boys and girls aged 8 to 16 years. They found that girls actually had slightly lower depression scores than boys up until age 12, but girls scores then increased and surpassed boys after age 12. This suggests that whatever leads to adult sex differences in depression does not begin to kick in until around puberty.
Whereas depression is disproportionately a female disorder, antisocial behaviors and conduct disorders (e.g., lying, cheating, stealing, fighting, truancy, drug and alcohol abuse, theft and other crimes that result in arrest) are disproportionately male disorders. One particularly well-conducted study repeatedly assessed, from childhood to early adulthood, the antisocial behaviors of most of the children born in a 1-year period in Dunedin, New Zealand (Moffit, Caspi, Rutter, & Silva, 2001). Individuals' levels of antisocial behavior were measured in different ways: via parents' reports, teachers' reports, friends' reports, and self-reports. The results were quite clear. On all measures, males showed more antisocial behaviors than females, and these sex differences grew increasingly large from age 5 to age 21. In the Dunedin study, the male-to-female ratio for the lifetime prevalence (occurrence) of antisocial disorders was about 2.4 to 1.
I he Dunedin study is particularly interesting because it showed that sex differences in antisocial behavior were strongly predicted by sex differences in various personality traits (such as aggressiveness, lack of control, lack of social closeness, and alienation). It is likely that sex differences in other kinds of mental illness are also linked to sex differences in personality. For example, neuroticism (one of the Big Five traits) is associated with depression and anxiety disorders (Kendler, 1996). As noted earlier, low agreeableness and low conscientiousness are associated with antisocial behaviors. Thus, observed sex differences in some kinds of mental illness are consistent with observed sex differences in "normal" personality traits. Women's higher levels of anxiety may contribute to their higher rates of depression and anxiety disorders. Men's higher levels of aggressiveness and lower levels of agreeableness and impulse control may contribute to their higher levels of antisocial behavior. And males' greater thing-orientation may be linked to higher male levels of autism (Baron-Cohen, 2003).
There is considerable controversy surrounding the topic of sex differences in mental disorders and behavior problems. Some researchers argue that observed differences may reflect biases in diagnostic practices and criteria more than real sex differences. Others argue that the differences are for real. Despite the possibility of bias, it is likely that many of the sex differences reported in Table 1.5 (such as sex differences in childhood autism, childhood speech and reading disorders, attention deficit disorder, depression, and antisocial personality disorders) are real and large.
Are There Sex Differences in Emotional Experience?
Meta-analyses have shown that women are slightly more self-disclosing than men are (d = 0.18, Dindia & Allen, 1992), That is, women share more personal information about their lives, thoughts, and feelings than men do. However, these differences depend in part on who does the disclosing and who is the target of disclosure. For example, most people (both men and women) are more self-disclosing with a woman than with a man. Self-disclosure also depends on the kinds of emotions expressed. Women express negative feelings, such as sadness and depression, more than men do (Zeman & Garber, 1996), whereas men express anger more than women do (Clark & Reis, 1988; see Brody, 2000, for a review).
Not only do men and women express emotions somewhat differently, they also seem to experience them somewhat differently. James Pennebaker (Pennebaker & Watson, 1988; Roberts & Pennebaker, 1995) found that men infer their emotions more from internal physiological cues (e.g., a man decides he's happy because he experiences increased heart rate and blood pressure), whereas women infer their emotions more from the social setting and context (e.g., a woman decides she's happy because she is in an audience with laughing people). Men prove to be more accurate than women are in estimating internal physiological cues such a heart rate; this may be one reason why they use such cues more to infer their emotions.
A related finding is that men tend to internalize their emotions more than women do. Men may not show their emotions facially as much as women do; however, they may churn more internally. In contrast, women externalize their emotions more than men do. They show their emotions in facial and verbal expressions, and perhaps as a result, they do not have such strong physiological arousal as men do (Buck, Savin, Miller, & Caul, 1972). Leslie Brody (1999) qualified this conclusion for women, however, and argued that, unlike men, women tend generally to express their emotions through many different modalities: facial expressions, verbal expressions, and physiology (see also Brody & Hall, 2002). Research on husband-wife communication provides some support for the conclusion that women, on average, generally are more emotionally expressive than men. Observations of spouses discussing conflicts show that women express their emotions, particularly negative emotions, more than men do, and this can lead to the common marital pattern of pestering wives who want constantly to talk things out, versus stonewalling husbands who clam up, close down, and withdraw both physically and nonverbally from their wives (Levenson, Carstensen, & Gottman, 1994).
Id a similar vein, University of California psychologist Shelley Taylor (Taylor et al., 2000; Taylor, 2002) has argued that, when stressed, men are more likely to show a fight-or-flight response, whereas women are more likely to show a tend-and-befriend response. Men are more likely to respond to threatening situations with aggression or withdrawal, whereas women are more likely to care for others (e.g., their friends and children) and seek out social support.
Taylor's hypothesis received support from a recently published metaanalysis of sex differences in coping behaviors (Tamres, Janicki, & Helgeson, 2002). People can deal with stress in a variety of ways: by seeking information, by planning concrete actions, by problem-solving, by seeking emotional support from others, by distraction (drinking alcohol and watching TV), and so on. In general, Tamres, Janicki, and Helgeson found that women report using virtually all kinds of coping strategies more than men do, perhaps because they are higher than men on neuroticism (i.e., women worry more and perceive more life events as stressful). The strongest and most consistent sex difference, however, was that women sought out emotional social support more than men did (d = 0.41). When these researchers collected additional data to look at relative coping styles (i.e., regardless of a person's absolute level of coping behaviors, which styles does a person use relatively more, and which relatively less), they found that women are more likely than men to prefer social support as a way to cope with stress, whereas men are more likely than women to prefer problem-solving and avoidance as ways to cope with stress.
Are There Sex Differences in the Self-Concept?
Psychologists Susan Cross and Laura Madson (1997) proposed that the self-concepts of men and women are organized somewhat differently. Men have a more independent view of themselves. They view themselves more in terms of their individual achievements, traits, values, and abilities, the ways in which they are unique and separate from other people. In contrast, women have a more interdependent and connected sense of self. They view themselves more in terms of their relations with others and in terms of social roles and obligations.
In recent years this distinction between the independent and the interconnected self has been studied by cross-cultural psychologists, who argue that the independent view of self is more common among people who live in individualistic countries such as the United States (Markus & Kitayama, 1991). People from such cultures often view themselves in terms of their autonomous principles, traits, values, and abilities. For example, an American woman might describe herself as "honest," "intelligent," "interested in cultural activities and the arts," and "good at statistics." In contrast, the interconnected view of self is more common in traditional, collectivist cultures, which are frequently found in Asia, Africa, and Latin America. People from collectivist cultures view themselves more in terms of their social roles and relations to others. A traditional Japanese woman might describe herself as a "wife," "mother," "good daughter to her aging parents," and "loyal employee."
Cross and Madson (1997) argued that men have self-concepts that are more typical of people from individualistic cultures, whereas women have self-concepts that are more typical of people from collectivist cultures. In a series of cross-cultural studies conducted in the mainland United States, Hawaii, Australia, Japan, and Korea, however, Yoshihisa Kasbima and his colleagues (1995) found evidence that cultural differences and sex differences in self-concept are somewhat different. Collectivist and individualistic cultures differ most in the degree to which people view themselves as independent of others or dependent on others, whereas women and men differ most on the degree to which they view themselves as emotionally related to other people.
Robert Josephs, Hazel Markus, and Roman Tafarodi (1992) reported several experiments that demonstrated differences in men's and women's self-concepts. In these studies, college men's self-esteem proved to be more linked to their accomplishments, and women's to their personal relationships. Furthermore, men's self-esteem was more threatened when they were challenged about their achievements and abilities, whereas women's self-esteem was more threatened when they were challenged about their nurturance and responsiveness to others.
Elaborating on men's and women's differing conceptions of social relationships, Roy Baumeister and Kristin Somraer (1997) argued that both women and men view themselves in relation to other people. However, women conceive of themselves more in terms of warm, one-on-one, intimate relations (e.g., daughter, spouse, best friend), whereas men conceive of themselves more in terms of social groups and hierarchical relationships (e.g., boss, member of sports team, American). Baumeister and Sommer (1997) put the matter succinctly, "... female sociality is dyadic, whereas male sociality is tribal" (p. 39). Gabriel and Gardner (1999) conducted studies that supported this proposed difference in men and women's conceptions of relatedness (see Gabriel & Gardner, 2004, for a review and update). One recent study found that even among 6- to 8-month-old infants, boys show their group orientation by looking more at groups of figures, whereas girls show their one-on-one orientation by looking more at single figures (Benenson, Duggan, & Markovits, 2004).
Are There Sex Differences in Children's Play and Friendship Patterns?
Most of this chapter has focused on sex differences in adults behaviors. To paraphrase the words of poet William Wordsworth, however, if "the boy is the father of the man" and "the girl is the mother of the woman," then it is important to consider children's behaviors as well. Much recent research has documented sex differences in children's play patterns and social interaction (Maccoby, 1998; see also Chapter 5). For example, boy toddlers get into trouble more and have more difficulty controlling their impulses than girl toddlers do. Older boys play in groups more than girls do, and boys' groups tend to be larger and more independent of adult supervision than girls' groups are (Benenson, Apostoleris, & Parnass, 1998). In contrast, girls play in same-sex dyads (one-on-one pairs) more than boys do.
Boys' group-oriented social life centers on dominance, hierarchy, and competition more than girls' one-on-one social life. (Think of boys' playing stickball and cowboy-and-Indian fights and girls' playing house.) Boys test one another's strength and toughness more than girls do. They also test and break adults' rules more than girls do. Boys tend to have somewhat higher activity levels than girls, and they engage in more rough and-tumble play, which sometimes degenerates into physical fighting. In contrast, girls' aggression often takes a more verbal form than boys'. Girls sully one another's reputations and ostracize outcasts from their groups when they want to be hostile; boys confront, shove, and punch. Girls get what they want through negotiation and verba! influence, boys through physical dominance, challenge, and combat.
Boys fantasy lives center more around enacting heroic figures (superheroes, sports figures, cops, and warriors), whereas girls' fantasy lives center more around enacting reciprocal social roles (mother-child, teacher-student, doctor-patient), often with other girls (see Maccoby, 1998, for a review). On average, boys and girls play with different kinds of toys (Power, 2000). Boys play more with mechanical toys (trucks, cars, erector sets), and girls play more with dolls and domestic toys (tea sets, dollhouses). Boys enjoy toys that allow them to role-play aggression (guns, swords, tanks). In their creative activities, boys play more with blocks and construction toys, girls more with art and music materials; in their pretend activities, boys play more with carpentry materials, war and fantasy toys, and transportation toys, whereas girls play more with domestic toys, dolls, and telephones. In activities that involve large movements, boys engage more in rough-and-tumble play, riding bikes, skateboards, and scooters, and playing ball and other team sports, whereas girls engage more in dance and social activities such as jump rope (Power, 2000).
In many cultures, girls not only play more with domestic toys than boys do but also actually tend more to other children and infants and perform more housework than boys do (Edwards, 2002). When interacting with same-sex peers, girls are warmer and more affiliative than boys, and they strive more for mutual participation in shared activities (Strough & Berg, 2000). Girls collaborate more and make their requests in kinder, gentler, and more nonverbally friendly ways; boys are more likely to command, demand, threaten, and boast (Leaper, 1991). Girls compete more contingently, depending on the situation (e.g., when resources are scare) whereas boys tend to compete regardless of the situation (Roy & Benenson, 2002). Perhaps all these childhood sex differences set the stage for the different interaction and communication styles of adult men and women, with men more verbally assertive and competitive and women more verbally collaborative and accommodating (Tannen, 1990).
All of the childhood sex differences just cataloged likely contribute to what is probably the most dramatic and consequential of all sex differences observed in children; sex segregation of friendships and playmates. Starting at around age 3 years, children interact more with members of their own sex than with children of the other sex, and as childhood progresses, children increasingly play and socialize more and more exclusively with members of their own sex (Maccoby, 1998). Sex segregation is a very strong phenomenon. Indeed, if plotted in the form of frequency distributions, boys' and girls' amount of interaction with boys (or with girls) would form two largely non-overlapping distributions. Boys "hang out" mostly with other boys—typically in groups—and girls hang out mostly with other girls— often in pairs or iri small friendship clusters. Childhood sex segregation does not dwindle until puberty approaches and children begin to experience the romantic and sexual attractions that will entice most of them back into frequent interaction with the opposite sex.