ABSTRACT
Behavioral scientists have tried for more than fifty years to document sex differences at the psychological level (for example, differences in personality attributes and cognitive abilities). Close analysis of the data used to support the general hypothesis of the existence of sex differences shows that there are few, if any, real differences at this level. Even aggression, traditionally considered male, or "masculine'" behavior is not as exclusive to men as researchers and the public once thought it was. Yet researchers continue to try not only to document sex differences but to explain them. Thus some rather elaborate theories have been devised to explain a phenomenon that is nonexistent. Sociobiologists have carried this practice to an extreme, by stressing comparisons of males and females across species and ignoring those features of humans that differentiate them from other species—consciousness, a sense of self, and a sense of humor. Hyde's chapter is important because it challenges this tradition on its own terms—that is, she uses the style, format, and methodology of traditional science to show that sex differences in certain cognitive abilities are so small they are meaningless. ("Small" here means that a comparison of both women and men with respect to the ability in question shows almost similar distributions. See Hyde's Figures 1 and 2.)
208To fully understand and appreciate this article, it is helpful to have some basic training in statistics. But those who do not should keep in mind the meaning of the term statistical significance. It does not necessarily mean that the difference between two groups (in this case, males and females) is a large one nor does it mean the difference is substantive. Even though evidence of statistical significance sometimes determines whether a study will be published, the determination of statistical significance depends to a great extent on the size of the sample: The larger the sample, the more likely a small or even miniscule difference between two groups will appear statistically significant. Further, the size, or magnitude, of a sex diffluence—that is, how much variation on a given attribute in a given population is due to sex (what Hyde calls effect size)—is an altogether different and separate statistical question than that of whether there is any difference at all. Hyde refers to the latter question as "hypothesis-testing." The point is that because of the nature of the methodology employed, sex differences in cognitive abilities (and in other individual attributes) have been greatly exaggerated—some might even say constructed—by the research process itself.
This note and Hyde's chapter refer to sex differences at the individual level, not to those at the social-structural, or interactional, level. There are sex differences in social patterns like the gender gap in earnings, sex segregation in occupations, and even some sex differences reported in this book—for example, Fish man's analysis of intimate conversation. These seem to be real differences in the sense that (1) their magnitude is greater than that for psychological differences of the kind Hyde is discussing and (2) they are partly explained by reference to ideological factors such as a culture's perceived differences about women and men. The discrepancy between evidence for sex differences at the individual level and those at the interactional and social levels itself needs to be explained.
