Your first recommendation is In a Different Voice by Carol Gilligan. Tell us a little about this book.

This was written about 30 years ago, by a psychologist, and it’s a classic text. It’s a book about why the way women make decisions, for instance, may be different from men. By doing studies only on male subjects and then saying that is the norm always makes women look less developed. That was the conventional viewpoint, that women somehow got stuck at an earlier stage of development. Instead of which you should be saying they are different, and compare women with women and men with men. Also, if you test drugs on men, the results may be very different from if you test them on women. It’s a very important and basic idea that men and women may be different, for whatever reason, and if you try to compare results from studies you may be making the wrong comparison. For me, when I read the book, I realised that I perhaps present my arguments in a different way – which may not be because I’m bad at presenting arguments, it may be something more fundamental and I shouldn’t therefore see it as a failing in me. But I still have to learn how to use it to convince other people.

What differences between men and women do you feel are evident, besides physical strength?

I don’t think there is any single identifiable thing. You could say men tend to be more self-confident, but that’s almost certainly nurture not nature – it’s not a characteristic difference in the way that strength is. But even talking about strength, a weak man could lose to a strong woman, so I don’t think that’s a useful way of looking at it. I think that as an individual you should try to learn what your strengths and weaknesses are, then try to do something about your weaknesses and play to your strengths. It’s about working out who you are – if you’re trying to succeed, that’s what matters.

Let’s move onto your second choice, Whistling Vivaldi, and the concept of “stereotype threat”.

This book is again written by a psychologist, a black academic from the United States who studied why bright black students did less well than would be expected when they went to university. He demonstrated that a lot of people underperform because they expect to underperform – because of “stereotype threat”, a phrase he coined. That view of things has been tested out in a lot of different situations, including girls doing maths tests. And if you take away the threat, you can see a noticeable difference in the way the underperforming group does. So if you tell girls doing a maths test, “this is a maths test, we’re testing your maths ability,” they will do significantly worse than if you say, “this is a test to look at how you set about problem solving,” which takes away the girls-are-bad-at-maths message.

Steele also demonstrates something I find completely incomprehensible, that very small interventions – which he calls “self-affirmations” – really improve performance. There have been studies published recently about self-affirmation before girls do physics tests, which show that the girls’ performance goes up. I think this is something that, in terms of how we approach teaching, is very important – how we subtly portray things to enable people to perform to the best of their ability.

So how does the stereotype of women not being good at physics and maths affect the choices they make?

If women are systematically underperforming in tests, even though they may enjoy the subject, then they are not going to choose to do it at a higher level.