44

Do oral examinations expose candidates to sexually biased assessment ? Using the example of the oral exam for admission (final selection) to the prestigious social science university Sciences Po Paris, we explore several hypotheses for why female candidates do consistently less well on this exam than their male counterparts, while doing better on the written entrance exam that determines the candidate shortlist. We show that boys get higher marks when they are the only member of their gender in a pool of oral exam candidates evaluated by the same jury or an exclusively male jury. We also show that while direct discrimination does exist—a different mark for an equivalent assessment—it only marginally explains grading differences, which are due primarily to the specific qualities or abilities tested by oral exams (speaking with ease, self-confidence, presentation abilities), skills unequally distributed by gender.