Asking Students to Confirm Their Gender Before a Test Leads to Lower Scores for Girls, Higher for Boys
Consider the Following
There’s a lot of research out there on the placebo effect where it applies to harmful stereotypes. Findings agree that if you remind a group of students that their demographic typically doesn’t do as well as another in a particular subject, and then test them on that subject, the added pressure to defy expectations causes test takers to underperform.
But it turns out that in some cases you don’t even have to remind students that they’re expected to underperform based on their race or gender. You just have to remind them that they belong to that race or gender.
The Glass Hammer had a talk with Joshua Aronson, a professor of Applied Psychology at NYU, and turned up some pretty interesting stuff. In a study by Educational Testing Services, the people who design the GRE and Praxis tests, a group of male and female AP students were asked to confirm their gender at the end of taking their AP exam, instead of at the beginning.
Females who received the gender inquiry before the test scored an average AP Formula Score of 12.5, while males scored an average of 16.5. In the groups that received the gender inquiry after the test, females scored an average of 15, while males scored an average of 14. Not only did stereotype threat significantly harm girls’ scores, but boys benefited from being reminded of their gender before taking the test.
According to Aronson, the same sort of test has shown a deviation of twenty or thirty SAT points in the tests of female students, and can apply to any test that asks its takers to confirm their racial and ethnic background before taking it, as well. So, basically, any test you need to get into an American college. One relatively simple solution that Aronson proposes is that teachers should emphasize tests as tools of learning in progress, not evaluations of learning in total. Still, easier said than done in the days of No Child Left Behind.
Have a tip we should know? [email protected]