[Research Seminar] IFLAME: “Gender differences in reaction to pressure: nature or nuture?” Y. FAWAZ – CEMFI

Speaker: Yarine FAWAZ
CEMFI

Date and Location – Thursday April 20th 2023 from 12:00 to 13:30
in Paris campus (P400) and on Zoom

—————————————–

ABSTRACT

The over-reliance on one single measure of success such as University entrance exams has been criticized mainly because of the pressure it puts on the shoulder of teenagers and because it drives them away from the true purpose of studying, that is, learning. There is yet another reason why high-stakes testing is problematic, as it has been shown to be unfair to certain categories such as females (Azmat et al (2016)) or minority students (Schlosser et al (2019)) when comparing their relative performance in high and low stakes situations.
The main result of this recent literature is that female students perform relatively worse than male students when going from low to high stakes (i.e. the gap between high and low stakes is narrower for males). A less well-known finding of Schlosser et al (2019) is that no such gap is found among the Asian student population, which points to cultural -rather than biological- factors being at play in these gender differences.

We dive into the one country that has probably put the most emphasis on the importance of the final exam, the College Scholastic Aptitude Test (CSAT), and use administrative data on the full universe of students taking the CSAT and the lower-stakes NAEA exam for the 2010-2015 period, to explore whether such gender differences in the performance gap between high and low stakes exist there. Our results are in accordance with the literature, since we do not find any significant gap between males and females, at least in mathematics.

We then exploit the quasi-random assignment of students into either single-sex (SS) or coeducational high schools, and find out that when taking into account specifically how males and females perform differentially in high vs. low stakes depending on whether they come from a SS or coed school, the “usual” gap prevails.

Last, we use longitudinal data from the KYPS survey to look for potential mechanisms that could explain why boys and girls from SS schools do not react to increased stakes in the same way as boys and girls from coed schools.