In other breaking news, women like pink.

Why doesn't anyone seem to think through their assumptions when doing research on gender differences? Proceedings of the Royal Society just did a study on women, men, and shopping. Like most studies that deal with gender, the science seems shoddy and the reporting (in this case, The Economist) is worse.

Here's the study:

[Yale Psychologist Joshua New] recruited 41 women and 45 men and led each of them individually on a merry dance around the chosen market. In the course of this peregrination, each participant visited six of the 90 food stalls in the market. At each of those stalls, participants were given a piece of food to eat. They were asked their preference for the taste of the food, how often they ate that food in normal life, how attractive they found the stall and how often they had made purchases from that stall in the past. After visiting all six stalls, they were taken to the centre of the market and asked to point toward those stalls, one at a time, using an arrow on a dial. In addition, they were asked to rate their own sense of direction.

And here's what they found:

On average, women were 9° more accurate than men at pointing to each stall—a significant deviation if you have to walk some distance to get to a place. This was not because those women had more experience of visiting the market than the men had. Nor did the women rate themselves as having a better sense of direction—indeed the men rated their own navigating skills more highly....Among both the British and the Chinese, women preferred reddish hues such as pink to greenish-blue ones. Among men it was the other way round.

Now, this is all very interesting, but then you get to the conclusions:

Dr New suggests that these results show women are better than men at the particular task of relocating sources of food. That contrasts with the idea that men are better at navigation in general. In other words, women's minds are specialised for their ancestral task of gathering the sort of food that cannot run away....Moreover, though anatomical sex is binary, mental “gender” is more pliable. To see how masculine or feminine the brains of their participants were, Dr Hurlbert and Dr Ling used what is known as the Bem Sex Role Inventory, which asks about personality traits more often associated with one sex than the other. This showed that the more feminine a brain was, regardless of the body it inhabited, the more it liked red and pink.
All this suggests a biological, rather than a cultural, explanation for colour preference. And Dr Hurlbert and Dr Ling have produced one. They suggest that their result may be connected with the fact that the colour of many fruits is at the red end of the spectrum. An evolved preference for red, pink and allied shades—particularly in contrast with green—could thus bring advantage to those who gather such things. And if they can also remember which tree (or stall) to go and visit next time, then so much the better.

Lets start with the test Hurlbert and Ling used. Their test points to which individuals identify more with culturally constructed gender traits, right? Because all their test associates gender with is certain common traits that appear in a given gender. Is it really a surprise, then, that they found that individuals who identify with female-associated traits, identify with a specific female-associated trait, pink? The Bem Sex Role Inventory is descriptive, not prescriptive. Here's wikipedia:

In 1971, [Dr. Bem] created the Bem Sex Role Inventory to measure how well you fit into your traditional gender role by characterizing your personality as masculine, feminine, androgynous, or undifferentiated. She believed that through gender-schematic processing, a person spontaneously sorts attributes and behaviors into masculine and feminine categories. Therefore, an individual processes information and regulate their behavior based on whatever definitions of femininity and masculinity their culture provides.

So it's not really surprising that a test designed to measure traditional gender roles produced traditionally-gendered results.

But that's not even the beginning of the problems with the "biological proof" conclusion. Let's also ignore for a second that the article makes the error of assuming that anatomical sex is binary. This experiment does nothing to control for the decades of social conditioning these women had to a) be good at shopping and b) like pink. Women (and, probably, "effeminate" men) simply have more practice at shopping than men (and, probably, "masculine" women) have. And they have been dressed in pink and red since birth, where little boys get blue and green. Why, then, is this all assumed to be based on a history of women gathering? My guess is that New et al went looking for a prehistoric answer and found one, but that doesn't make their control problems less grave (and having someone "rate themselves" as a control for their sense of direction is almost laughable).

Why do sloppy experiments with specious conclusions get press? Hearing about biological roots to our gender assumptions is interesting—it certainly caught my eye. Our daily lives are constantly inflected with gendered expectations, and it's hardly surprising we want to know those expectations' source. But it would be awfully nice to get that information from better hands.