Women are crap at mental rotation

June 12, 2011 - Women in Computing

In psychology, there’s a classic study of mental rotation. In this, you get an example shape and some other shapes, and you have to work out which are the same as the example. The first rotation study was done by Roger Shepard and Jacqueline Metzler back in 1971. The technique is used to show various things – it appears that the time we take to match a shape is proportional to the angle we’d have to rotate it through. Which suggests that we are, indeed, rotating the image in our head. So, we may well have images in our heads (rather than “features”, or other more abstract representations). This might seem obvious to you, but for those of us who work in vision, determining the way in which our brains choose to represent the visual world is actually a very interesting and far from settled research question. So mental rotation is important, and one of the most repeated experiments in psychology.

In some of the many many variants of mental rotation study that have been run since Shepard’s original, researchers have found startlingly consistent gender differences. Women are a bit slower than men. This has been held up as an example (or even proof) of women’s poorer visual-spatial ability, and used to explain our “difficulty” with mathematical, scientific, visual disciplines. It’s even been linked to our lamentable inability to park a car. But I’m reading a book that describes more recent research into stereotype threat showing we can reverse this effect in certain situations. It’s making me think again about the whole women in science question.

If you cast the mental rotation task as a test of cognitive or perceptual ability (as researchers normally do), men do better. If you describe it as correlated with visual, spatial, or engineering ability, men do better again. But if you say that good performance on mental rotation tests is correlated with ability in interior design, flower arranging and needlework, men’s advantage starts to slip. Remember, we’re talking about exactly the same test here, just with a different preamble. Apparently, in another experiment, an Italian researcher [1] repeated the mental rotation stuff with the following preamble: “Apparently, women perform better than men in this test, probably for genetic reasons” and the opposite statement. You guessed it – the women did better when they were told they were expected to, and the men did better when they were told men were expected to perform better.

The problem here is that the general expectation – the stereotype – is that women are worse at some tasks and men are worse at others. The effect can be exacerbated, exaggerated, and emphasised by circumstances (women don’t do as well at tests if they’re in an obvious minority, for example). This is what is meant by stereotype threat: the threat of conforming to your stereotype causes all sorts of additional cognitive load, which ironically makes you more likely to conform to your stereotype. People have even done experiments where they’ve showed participants TV ads featuring dizzy women (or neutral ads) before tests, and shown that simply watching a TV ad with an airhead in it can make women perform less well in a test environment [2]. This has such far-reaching implications for stuff like testing, teaching, role models, and… pretty much all of STEM education and policy. It’s going to take me a while to mentally digest it all.

The book that’s introduced me to all this is Delusions of Gender, and it’s turning out to be a great buy. I’m only on chapter five, but it’s an eminently readable overview of current gender research and I’m finding it frightening and enlightening. I’d seriously recommend it to anyone interested in gender issues – particularly those of use who are working in computing and other disciplines where there’s a serious gender gap. It’s surprising to think how much this stuff might be affecting us, even when we’re as open and as serious as we can be about trying to negate the effects of gender on our performance and experiences. And that goes for the blokes, too.

[1] Moè, A. (2009). Are males always better than females at mental rotation? Learning and individual differences 19(1) 21-27 http://dx.doi.org/10.1016/j.lindif.2008.02.002

[2] Davies, P. G., Spencer, S. J., & Steele, C. M. (2002). Consuming images: How television commercials that elicit stereotype threat can restrain women academically and professionally. Personality and Social Psychology Bulletin, 28(12), 1615-1628 http://dx.doi.org/10.1177/014616702237644


  1. Fenny says:

    One of the things I often wonder when I read the “men do X faster than women” type articles is whether women are generally more cautious and like to check their work and men want to just get it done. I’m sure that’s a stereotype, but it may be true in some instances.

  2. Jo Hoy says:

    you just got her a sale. I love 1-click ordering to Kindle :-)

Leave a Reply

Your email address will not be published. Required fields are marked *