SEMINAR 7 February 2012: Veronica Frisancho Robles, Pennsylvania State University

Learning Gains among Repeat Takers of the Turkish College Entrance Exam: Disentangling Selection and Learning Effects

Abstract

We use a unique dataset on a sample of about 115,000 students who took the centralized college entrance exam (¨OSS) required for college placement in Turkey in 2002 to estimate the learning (in terms of score improvement) predicted to occur upon retaking the exam. We assume that students know their own ability, that exams have a random element, and that the system is in steady state. We show how to use information on retakers to estimate learning across attempts while controlling for selection into retaking in terms of unobserved ability and observable characteristics. We identify important learning gains among repeat takers, especially for students from less advantaged backgrounds. If the less privileged learn more on retrying than the more privileged as our results suggest, the Turkish approach, by simply allowing students to take the entrance exam as many times as they want, may reduce the effects of background inequalities on college admission. The option of retaking levels the playing field for less advantaged students and enables them to get better prepared before they go to college.

Keywords: learning gains, higher education, factor analysis, selection

JEL Codes: C13, C38, I23