In one study, for example, a group of French high school students was asked to rate the truth of stereotypes about gender difference in talent in math and the arts before rating their own abilities in these domains. So, for these students, gender stereotypes were very salient as they rated their own ability. Next, they were asked to report their scores in math and the arts on a very important national standardized test taken about two years earlier. Unlike students in a control condition, those in the stereotype-salient group altered the memory of their own objective achievements to fit the well-known stereotype. The girls remembered doing better than they really had in the arts, while the boys inflated their marks in math. They gave themselves, on average, almost an extra 3 percent on their real score while the girls subtracted the same amount from their actual math score. This might not seem like a large effect, but it’s not impossible to imagine two young people considering different occupational paths when, with gender in mind, a boy sees himself as an A student while an equally successful girl thinks she’s only a B.
— Cordelia Fine, Delusions of Gender: How Our Minds, Society, and Neurosexism Create Difference