This is the second post about my quick impressions after reading three critical books about the science of sex differences, and it has to do with the preliminary questions researchers ask, how they frame the research and perhaps even the question how one becomes a researcher of biological sex differences. (The first post can be found here.)
Publication and Study Biases
Note first, that almost all of us are viewed as either female or male by the society. None of the researchers of sex differences among humans are non-sexed aliens from outer space (well, I don't think so though one never knows for sure) and every one of them has already made his or her non-scientific impressions about gendered behavior and what might cause it.
If your impressions make you believe that gendered behavior is mostly caused by biological sex differences, you are much more likely to enter a field studying those differences than if your impressions make you believe that gendered behavior is mostly caused by either environmental and cultural factors or some complicated soup of all possible factors.
Because of this, I believe that the overall field of biological sex differences may have an inner bias in what questions it analyzes and how it looks at the evidence: It will begin with observed gender differences in a particular society and then move from that to the attempt to find corresponding biological differences as an explanation for those. Environmental/cultural factors will be mentioned, but astonishingly seldom actually controlled for. This may over-attribute gender differences to biological sex differences.
I am not arguing that these biases were overt. They are most likely unacknowledged, except in a few famous cases of clear misogyny being the motivating factor (coughBaron-Cohencough).
But what it probably DOES mean is that the research will search for certain types of differences, ignore other types of differences and most certainly ignore the similarities which are found. It also means that alternative explanations will be down-played.
The same thing would also happen if there was a science of sex similarities. But it doesn't exist, so we don't have to worry about that one! Whenever results fail to show a sex difference they are either "put into a file drawer" (meaning that they are not published), never to see sunlight again, or they are posted in a rewritten form where something else is emphasized as the major finding.
The file drawer bias in research findings is not unique for the field of sex differences. It applies to most research where findings of "no support for a particular theory I like" are not published as often as they should be.
But its impact is especially severe in this field, because the process of excluding certain kinds of findings from public awareness is two-fold here: First, findings of "no difference" are under-reported in the academic publications, and, second, the popularizers grab almost always only those findings which reinforce the story of biological/evolutionary sex differences.
To see how this research bias works in quite subtle forms, consider the vast literature into women and girls with congenital adrenal hyperplasia (CAH). Jordan-Young defines it as follows (pp 30-31 in Brainstorm):
CAH women and girls are a major source of data for those who want to analyze biological sex differences and also one of the pieces of evidence which is usually offered as support for innate causes of masculine vs. feminine behavior. Jordan-Young discusses these studies in great detail and I recommend reading her book on those.
This syndrome, a genetic disorder that causes overproduction of androgens from the adrenal glands, is the most common cause of genital ambiguity. Androgens are elevated throughout fetal development, which is an especially unusual situation for female fetuses. Because of the hypothesis that high androgen levels may masculinize the brain as well as the genitals, people with this disorder -- especially girls and women -- have been much studied by scientists interested in brain organization.
What I want to discuss here is the way these studies have tried to control for all other possible effects than CAH. This is by comparing CAH women/girls to their female siblings or to some larger population of women/girls without CAH but with the same age, ethnicity, social class and so on. A neat way of controlling for any environmental factors, right?
Not quite, and the reason is that having CAH means corrective surgery, continuous medical supervision (every three months during childhood and adolescence) and hormonal treatments to induce higher adult height, for example. What is not controlled for in the vast majority of CAH studies are these very facts, because the control group consists of individuals who are not suffering from a chronic condition. Neither do these studies really control for the stigma of CAH and the way it affects the woman's sexual and reproductive opportunities.
To give an idea of one study which did try to take these into account (Brainstorm p. 229):
The Search For Sex Differences And Their Meaning*
While some aspects of interest are indisputably masculinized in girls and women with CAH, perhaps an "organizing" effect of prenatal androgens is not the best explanation. Note especially that few studies have attempted to evaluate the effect of illness itself, or the medical intervention that chronic illness entails. As an exception, Froukje Slijper (1984) compared girls with CAH to girls with diabetes as well as to healthy controls and found that both groups of girls with chronic illness scored in the more masculine range than controls on the gender scale.
How does one go about searching for biological differences between the sexes in fields such as cognition? The obvious answer (problematic though it is, as will be seen) is that one starts with observed differences, then excludes non-biological explanations and finally tries to find some other difference, preferably genetic or prenatal, which can be linked to those observed behavioral differences.
Sounds good, right? Or at least familiar. But what if we reversed the search? What if we could begin with some biological sex differences and then see what they produce in the behavior of the sexes? Some of that search might even find that such differences produce not gender differences in behavior but gender similarities!
Cordelia Fine's Delusions of Gender (pp. 142-143):
Of course humans are not prairies voles. But neither are humans rats, and I keep reading how rat behavior is relevant for understanding the limitations of the human female all the time.
One very striking example of the principle that brain difference can yield behavioral similarity, discussed by De Vries, comes from the prairie vole. In this species, males and females contribute equally to parenting (excepting, of course, nursing). In female prairie voles, parenting behavior is primed by the hormonal changes of pregnancy. But this leaves a mystery. How do father voles, which experience none of these hormonal changes, come to show paternal behavior? The answer turns out to lie in part of a region of the brain called the lateral septum, which is involved in the triggering of parental behavior. This part of the brain is very different in males and females, being much more richly endowed with receptors of the hormone vasopressin in the male, yet this striking sex difference in the brain enables male and female prairie voles to behave the same.
It's therefore important to remember that we may be biased if we start only with observed gender differences in behavior. It's quite possible to have biological sex differences which create similarities in behavior by compensating for some other biological difference between the sexes.
Not that all observed gender differences get the same attention from the researchers or the politicizers, by the way. You may well be aware of the clear difference in the average score between men/boys and women/girls in the skill of three-dimensional mental rotation. That one has been extensively studied. It's political uses are equally many.
But this is not the only test of spatial ability in which we find gender differences. Lise Eliot in Pink Brain, Blue Brain (p. 122) discusses the object-location test. This consists of showing test subjects a picture with many randomly scattered objects, and then showing them a different picture with some objects moved to different positions:
How fascinating that the gendering of objects matters in this test! That suggests to me that the three-dimensional mental rotation tests should be done with a weird furry creature or a six-legged and asymmetric Barbie doll, instead of something that looks like building blocks, especially given that those are now sold in boys' aisles in toy stores.
The task is to circle the items still in their original place and cross out the items that had been moved.
A recent summary of more than three-dozen such studies shows that women have a small-to-moderate advantage in object-location memory, but the difference depends in part on the type of objects subjects are asked to remember. Women do better with most objects except for stereotypically masculine ones, like a necktie, golf ball, trophy, suit coat and aftershave (men are better at remembering these objects' locations.)
It's important to note that one can practice these kinds of spacial abilities and raise one's scores that way. This means that the scores do not measure a purely innate characteristic.
A Final Reminder
It is important to remember our biases and partial blindness when discussing this field. Jordan-Young, Brainstorm, p. 256:
------
Historians and philosophers of science are giving increased attention to the way that gaps in knowledge, as well as knowledge itself, are actively produced and maintained. The study of this phenomenon, what Tuana calls "the epistemology of ignorance" and Proctor (2008) calls "agnotology" reveals that specific ideologies, cultural schema, and political interests systematically block certain forms of information and cause people to "forget" or fail to incorporate certain facts into the overall thinking on the subject.
*Added later: I forgot to note here that much recent research suggests a very complicated interplay between what used to be called nature vs. nurture, and that in some ways we may have been asking the wrong questions even more generally.