Why I Do Not Like Sex

As a variable, I mean.I, as a research psychologist, do not like sex as a variable in our studies.In fact, I do not like education, race, ethnicity, income, marital status, and all other variables of that ilk.The problem is that sex (and the other variables) is not a psychological construct.When sex is used as a variable in psychological studies, we are never (well, next to never) interested in what sex really is, a biological variable.We are not interested in the nature of the genitalia, in the gonads, and so on.Not even in the distribution of body hair or the nature of the upper body profile.We use sex as a proxy or surrogate for a latent, but unspecified, variable.We use sex because it is easily observed or measured and, for all practical purposes, really can be assumed to be measured without error.But sex, the biological, observed variable, is often, maybe even usually, only weakly correlated with the latent variable.That is, biological sex is often a weak proxy for the variable we are really interested in.What is this latent variable?Well, it is actually one of a whole set of potential latent variables that are amorphous and overlapping.But, essentially, when psychologists use sex in their studies, it is because they believe that it captures the variance in a variable that they almost never make explicit but that represents in some way the residue of the array of life experiences that fall differentially to the lot of men or women.For example, the observation that females are far more likely than males to develop anorexia/bulimia is not usually premised on any notions about the actual effects of biological sex, e.g., such as might result from sex-related hormones.Rather, the underlying variable is couched in terms of ideas about such experiences as pressures toward certain body images, male domination, etc., etc. (Not so very long ago, Anne Quindlen's column-admittedly she is not a psychologist-blamed eating disorders on Barbie dolls).When it is observed that females seem to find it more difficult to quit smoking than do males, that difference, again, is usually not referred to biological sex differences, e.g., in addictive potential.The difference is usually taken to derive from differences in the social meaning of smoking, in susceptibility to influence, in differential support for quitting, and so on.So? Females differ considerably in the degree to which they are exposed to all the life experiences that contribute to that feminine residue.Some females are much more exposed than others, and the experiences rub off more on some females than others.Therefore, female (sex) is only an inexact measure of the latent variable.In fact, *males* differ in the degree to which they are exposed to the very same experiences.Some males are