188 Part II Psychodynamic Theories
in The American Psychologist. In it, instead of testing the commonly held gender
differences model, she analyzed meta-analyses of gender differences to test the
opposite prediction: that males and females are more alike than they are different.
One nagging problem in the study of gender differences and similarities is
embedded in the very nature of empirical science, and Hyde’s method provided a
way of getting around it. That is, comparing groups relies on statistical tests of
significant difference. When differences do not reach statistical significance, then
such a study is typically not published. This is referred to as the “file drawer
problem.” Consider the fact that there are likely countless studies of gender dif-
ferences in all kinds of cognitive, behavioral, and emotional domains that did not
result in statistically significant differences between males and females, and these
were put in some researchers’ file drawers because the results did not reach the
standard for publication. This means we have more evidence for differences
between the genders not necessarily because they are meaningfully and lastingly
there, but because only the studies that resulted in difference got published.
Meta-analysis is a statistical method of systematically combining data from
many studies (published and unpublished). This increases power over individual
studies, and can improve our capacity to estimate the size of effects. Hyde’s (2005)
analysis did just this, examining 46 meta-analyses, each of which included between
20 and 200 individual gender difference studies, and her examination of effect sizes
supported the gender similarities hypothesis. That is, there were far more arenas
where gender differences were either meaninglessly small in effect size, or nonex-
istent, than the areas where the genders did differ significantly. In some of the
arenas in which we hold the strongest stereotypes about males and females being
different, such as math and verbal ability as well as aggression, self-esteem, and
assertiveness, Hyde found that gender differences were negligible.
It seems that our culture’s obsession with gender difference has remained
strong since Horney’s day, despite her insistence that we ought to stop “bothering
about what is feminine and what is not.” Women are apparently from Venus and
men are supposedly from Mars. Google “gender differences” and you get over 9 million
results! Why do we persist in believing so strongly in gender differences in person-
ality, despite powerful evidence that they are barely there at all? One answer is
that thinking males and females come from different planets appeals to our intu-
itions. The danger here, of course, is that our expectations likely guide our cogni-
tions and our behaviors, creating a self-fulfilling prophecy. We likely make gender
differences come true when we believe that women are “bad at math” or men are
“not emotional.” When we treat each gender based on these expectations, then we
should not be surprised that they behave accordingly. If a girl is not expected to
perform well in math, her parents might not encourage her to study, and she may
begin to lack confidence and eventually not, in fact, get the education that would
make her good at math. If a boy is expected to “suck it up” and not cry when he
feels hurt, he may suppress tears in an effort to be more masculine, and these
efforts may, over time, create an actual inability to cry in adulthood.
Hyde (2005) issued a warning in her article about the costs of overinflated
claims of gender differences: “Arguably they cause harm in numerous realms,
including women’s opportunities in the workplace, couple conflict and communica-
tion, and analyses of self-esteem problems among adolescents. Most important,