The Science and Politics of Studying Sex and Gender

Few topics generate as much controversy as the study of sex dif ferences. This is especially true when it comes to examining the possibility that men and women dif fer. As noted in a recent discussion on gender, "public debates about the nature of women and men are frequently in the spotlight, whether in media reports on the latest sex dif fer-ence findings or in highly publicized legal cases involving single-sex educational insti tutions or sexual harassment" (Deaux & LaFrance, 1998). Some worry that findings o sex differences might be used to support certain political agendas, such as excluding women from leadership roles. Some worry that findings of sex di ferences might be used to support the status quo, such as keeping men in power and women out of power . Some argue that findings of sex di ferences merely reflect gender stereotypes, rathe than real dif ferences. Some psychologists ar gue that any discovery of sex dif ferences merely reflects the biases of the scientists, rather than any objective description of real ity. Indeed, some psychologists such as Roy Baumeister have advocated stopping research on sex dif ferences because findings of sex di ferences might conflict wit ideals of egalitarianism (Baumeister , 1988), although he has since changed his views on this (personal communication, May 17, 2006).

Others argue, however, that both scientific psychology and social change wil be impossible without coming to terms with the real sex dif ferences that exist. Feminist psychologist Alice Eagly, for example, argues that sex differences exist, they are consistent across studies, and they should not be ignored merely because they are perceived to conflict with certain political agendas (Eagl , 1995). Indeed, Eagly ar gues that feminists who try to minimize these dif ferences, or pretend that they do not exist, hamper the feminist agenda by presenting a dogma that is out of touch with reality . Still others, such as Janet Hyde, ar gue that sex differences have been exaggerated and that there is so much overlap between the sexes on most personality traits that the differences are minimal (Hyde, 2005; Hyde & Plant, 1995). We will examine these contrasting positions in more detail.

History of the Study of Sex Differences

The study of sex differences has a fascinating history within psychology. Prior to 1973, relatively little attention was paid to sex dif ferences. Indeed, in psychology research, it was common practice to use participants of only one sex, most often males. And even when both men and women were studied, few articles actually analyzed or reported whether the ef fects differed for men and women.

All of this changed in the early 1970s (see Eagly , 1995; Hoyenga & Hoyenga, 1993). In 1974, Elenore Maccoby and Carol Jacklyn published a classic book, The Psychology of Sex Differ ences, in which they reviewed hundreds of studies and drew several key conclusions about how men and women dif fered. They concluded that women were slightly better than men at verbal ability . Men were slightly better than women in mathematical ability (e.g., geometry , algebra) and spatial ability (e.g., ability to visualize what a three-dimensional object would look like if it were rotated in space by 90 degrees). In terms of personality characteristics, they concluded that only one sex difference existed: men were more aggressive than women. With other aspects of personality and social behavior, they concluded that there was not enough evidence to determine whether men and women dif fered. Overall, they concluded that sex differences were few in number and trivial in importance.

The Psychology of Sex Differences set off an avalanche of research on the topic. The book itself was criticized on various grounds. Some ar gued that many more sex differences existed than were portrayed by Maccoby and Jacklyn (Block, 1983). Others challenged the conclusion that men were more aggressive than women (Frodi, Macauley, & Thome, 1977). Furthermore, the methods by which the authors drew their conclusions, although standard practice at that time, were crude by today' s standards—the authors simply summarized the studies by counting how many reached statistical significance and then drew some interpretive conclusions. This method allows for the possibility of considerable subjectivity .

Following the publication of The Psychology of Sex Differ ences, psychology journals changed their reporting practices. They started to require authors to calculate and report sex dif ferences, if members of both sexes were included. Furthermore, protests that many of the findings in psychology were based primarily on studies o men led to calls for the greater inclusion of women as participants. There followed an explosion of research on sex dif ferences. Literally thousands of studies were conducted on the ways in which men and women dif fered. Indeed, by 1992, the federal government had required members of both sexes to be represented in all federally funded research (unless, of course, there was a legitimate reason to limit the research to one sex, such as studies of breast self-exam for breast cancer).

Since Maccoby and Jacklyn' s early work, researchers have developed a more precise quantitative procedure for examining conclusions across studies and thus for determining sex dif ferences, called meta-analysis. Recall that meta-analysis is a statistical method for summarizing the findings of la ge numbers of individual studies. Meta-analysis did not gain popularity until the mid-1980s. Meta-analysis allows researchers to calculate with greater objectivity and precision whether a particular difference—such as a sex dif ference—is consistent across studies. Furthermore, it allows researchers to estimate how lar ge the dif ference actually is—called the ef fect size.

Calculation of Effect Size: How Large Are the Sex Differences?

The most commonly used statistic in meta-analysis is the effect size, or d statistic. The d statistic is used to express a dif ference in standard deviation units (see Chapter 2). A d of 0.50 means that the average dif ference between two groups is half a standard deviation. A d of 1.00 means that the dif ference between the groups is one full standard deviation. A d of 0.25 means that the dif ference between the groups is one-quarter of a standard deviation. An effect size can be calculated for each study of sex differences and then averaged across studies to give a more precise and objective assessment of whether the sexes dif fer and, if so, by how much.

Most meta-analyses have adopted a convention for interpreting ef fect sizes (Cohen, 1977):

d Score Meaning

0.20 or -0.20 Small difference

0.50 or -0.50 Medium difference

0.80 or -0.80 Large difference

When comparing men to women, assume that positive d scores, such as .20 or .50, indicate that men score higher than women. Negative d scores, such as -.20 or -.50, indicate that women score higher than men. For example, a d score of -0.85 means that women score much higher on a particular trait.

To get a feel for various ef fect sizes, let' s examine a few findings outside th realm of personality. Which sex can throw a ball farther , men or women? Although there are great individual differences within each sex, it is clear that men can, on average, throw farther than women. The d is approximately 2.00 (Ashmore, 1990). This means that the sexes differ, on average, by two full standard deviations, which is considered quite large. Which sex has a higher grade-point average in college? The d for grade-point average is -.04, which is very close to zero. This means that men and women are essentially the same in their grade-point average.

Which sex scores higher in verbal ability? It turns out that women are slightly better than men, but the d is only - 0.11. Are men better at math? The d here also turns out to be quite small, only 0.15. These findings are in line with a vast literature tha now documents that men and women are essentially the same (or do not dif fer by much) on most measures of cognitive ability (Hyde, 2005). About the only well-documented exception to this conclusion pertains to spatial rotation ability , such as the spatial ability involved in throwing a spear (or football) so that it correctly anticipates the trajectory of a moving object, such as an animal or a receiver . The d for this sort of spatial ability is 0.73, which comes close to the standard for "large" (Ashmore, 1990).

It is important to keep in mind that even lar ge effect sizes for average sex dif ferences do not necessarily have implications for any particular individual. Even with ad of 2.00 for throwing distance, some women can throw much farther than the average man and some men cannot throw as far as the average woman. This overlap in the distributions of the sexes must be kept in mind when evaluating ef fect sizes (see Figure 16.1).

When it comes to who can throw a ball farther, the effect size for the difference between men and women is 2.00, in favor of the men. While this is a large difference in average ability, there will nevertheless be some women who can throw farther than most men because the distributions still overlap.

When it comes to who can throw a ball farther, the effect size for the difference between men and women is 2.00, in favor of the men. While this is a large difference in average ability, there will nevertheless be some women who can throw farther than most men because the distributions still overlap.

Figure 16.1

Overlap between the sexes in context of a mean difference. Even when one sex greatly exceeds the other in a particular ability, there is a large area of overlap. Women whose throwing ability falls in the shaded area exceed the throwing ability of the average man.

Minimalists and Maximalists

A central focus of the debate on sex dif ferences follows from a consideration of ef fect sizes—on whether sex dif ferences are small and relatively inconsequential or substantial and important. Those who describe sex dif ferences as small and inconsequential, those who take the minimalist position, of fer two ar guments. The first i that, empirically, most findings of sex di ferences show small magnitudes of ef fect (Deaux, 1984; Hyde, 2005; Hyde & Plant, 1995). Minimalists tend to emphasize that the distributions of men and women on any given personality variable show tremendous overlap, which reflect their small magnitude of e fect (review Figure 16.1). A second argument advanced by minimalists is that whatever dif ferences exist do not have much practical importance for behavior in everyday life. If the sex dif ferences are small and don' t have consequences for people' s lives, then perhaps we should concentrate on other psychological issues that are more important.

In contrast, those who take the maximalist position tend to ar gue that the magnitude of sex dif ferences is comparable to the magnitude of many other ef fects in psychology and should not be trivialized (Eagly , 1995). Some sex dif ferences tend to be small in magnitude, others are lar ge in magnitude, and many are in the moderate range, according to this view . Furthermore, Eagly (1995) notes that even small sex differences can have large practical importance. A small sex difference in the proclivity to help other people, for example, could result in a lar ge sex difference in the number of lives each sex aids over the long run in times of distress. As you read through this chapter , you should keep in mind the range of positions psychologists have taken on sex differences, from the minimalist stance to the maximalist stance.

Figure 16.1

Overlap between the sexes in context of a mean difference. Even when one sex greatly exceeds the other in a particular ability, there is a large area of overlap. Women whose throwing ability falls in the shaded area exceed the throwing ability of the average man.

Eliminating Stress and Anxiety From Your Life

Eliminating Stress and Anxiety From Your Life

It seems like you hear it all the time from nearly every one you know I'm SO stressed out!? Pressures abound in this world today. Those pressures cause stress and anxiety, and often we are ill-equipped to deal with those stressors that trigger anxiety and other feelings that can make us sick. Literally, sick.

Get My Free Ebook


Post a comment