Stereotype threat

When I took my GCSE science exams, sometimes the questions were set in the context of an experiment.

John and Sarah are investigating the effect of temperature on the rate of photosynthesis.

or

A physicist was investigating this or that phenomenon.

I used to notice how the individuals who featured in these questions were scrupulously gender-balanced.

An engineer has been consulted about the design of a bridge. She has the following concerns…

That I noticed this at all says something, I suppose, about stereotypes and about women and career choices. I never thought a huge amount of it. I certainly didn’t think that the gender of these characters, mentioned in passing, could affect exam performance.

I was reminded of this phenomenon, when I was listening to the All in The Mind podcast this week. (All In The Mind is a BBC Radio 4 “programme exploring the limits and potential of the human mind”). From about 16 min 30 s into the programme, Claudia Hammond interviews Jessica Good, a PhD student in the psychology departement at Rutgers University, about her paper The effects of gender stereotypic and counter-stereotypic textbook images on science performance.

Good explains that science textbooks that contain photographs are more likely to portray scientists as male than as female. Where women are featured, they are more likely to be portrayed in lower status roles – the male doctor and the female nurse; the male scientist and the female assistant. Sound familiar?

Good presented high school students with some pages of a chemistry text book to study. The students then took a comprehension test evaluating their learning and understanding. Unbeknown to the students, they had been given one of three versions of the text. In one version, the scientists in the photographs accompanying the text were male, in another the scientists were female, and in the third photographs of both men and women were used. The hypothesis, confirmed by the study, was that female students would perform better when the photographs accompanying the text they had read were of female scientists. Male students performed better when the photographs were of males. When the photographs were of male and female scientists, there was no difference between the performance of male and of female students in the comprehension test.

Good relates this result to a phenomenon termed Stereotype Threat – the idea that in a situation where a relevant negative stereotype about your group exists, you worry about confirming that negative stereotype, and this makes your performance worse. She points out that students did not always remember the gender of the scientists in the photographs, but that their performance was nonetheless affected.

Good does not claim that making science textbooks more gender balanced would eliminate the gender gap in performance. However, as her study found that the gender of the scientists in the images does have some effect on student performance, she suggests that using mixed-gender images might contributed positively to reducing this gap.

I wonder whether hope that science text books produced in recent years are more likely to portray scientists as women, compared to books produced longer ago.

This study made me wonder, whether the gender balance in exam questions is deliberately chosen with this phenomenon in mind? Would it be feasible to conduct this study on a country-wide scale? Print three versions of GCSE exam papers, with questions featuring either John and James, John and Sarah, or Sarah and Sue, and see whether this had an effect on student performance.

Maybe not, but anyway, I agree with Good’s conclusion.

Although eliminating gender bias in textbooks will most likely not eradicate the gender gap in science interest and achievement, it will begin to chip away at an ever-crumbling foundation.

¹ Good JJ, Woodzicka JA, & Wingfield LC (2010). The effects of gender stereotypic and counter-stereotypic textbook images on science performance. The Journal of social psychology, 150 (2), 132-47 PMID: 20397590