Does Upbringing Really Affect Gender Differences In Spatial Ability?

Link to the abstract
Today’s post discusses the paper “Nurture Affects Gender Differences in Spatial Abilities” by Moshe Hoffman and Uri Gnez. This paper is a rebuttal to the most widely known and empirically supported gender difference in cognitive ability to date – spatial ability. I was delighted to find this paper, because my paper deals with gender differences in cognition – I have a love-hate relationship with my research, because my feminist leanings make me deeply interested in seeing evidence that these differences are due to socialization rather than innate ability. These things are, of course, deeply and intricately intertwined. First, however, let’s return to the article at hand.
Hoffman and Gnitz hypothesize that matrilineal or patrilineal social structures contribute to the observed sex differences by providing resources and education to one sex. And this is exactly what they found – that

As we can see here, in patrilineal societies, males outperformed females by finishing the spatial puzzle in 42.31s (on average), while females took 57.17s (on average). In comparison, there were no gender differences present in matrilineal societies. Accounting for a decent sized portion of the difference was the factor of education–in the matrilineal societies, men and women had about the same level of education, whereas there was a larger difference in patrilineal societies (males had around 5 more years of education than women).
Now, this might invalidate my whole area of research. . . except for this:

The puzzle they used was a simple 4-piece puzzle.
Now, we already know that education has been shown to eliminate a large portion of the gender differences in spatial ability. However, research has also shown that gender differences are particularly pronounced as spatial tasks become more difficult. These findings do demonstrate that socialization does have a large effect on spatial ability, but they have not been shown to explain the entire phenomenon.
I will discuss possible reasons for this in a later post, but it seems that as spatial tasks become more difficult, females are more likely than males to use an analytical and top-down approach to problem solving. In contrast, men seem to use a bottom-up, more automatic approach to solving these problems – so they seem to be much faster at it. the FMRI study supports these findings; men seem to have more automatic visual areas activated when solving spatial problems, while women do not. Of course, not all men and women behave cognitively in this way, and there seems to be a 65/35 bias towards respective cognitive gender biases – so about a third of men think in atypically male ways, and women do as well. So far, I’ve concluded that these “innate” differences are very subtle at a young age and are amplified by social influences. Taken as a whole, these observed gender differences are very small and are over-hyped because the media likes to make mountains out of them.