How to predict university performance: a case study from a prestigious Turkish university?

Turkish education system is based on a centralized selection scheme starting from secondary education level. For the post-secondary studies, students are sorted according to scores they obtained from a national competitive exam. By statute, Galatasaray University, which is a French speaking university founded in 1992 by Turkey and France, enrolls one half of its students amongst the very best candidates. The other half comes from French speaking high schools (relatively top-ranked). These students pass a specific competitive exam in French, and must be classified amongst the first 25000 at the national competition. But their national ranking remains lower than the other group. The first batch has to learn French before starting undergraduate studies, whereas French speaking students are entitled to enter directly the first year. Within the public university system where admission is strictly based on national exam scores, differentiated admission scheme of Galatasaray University offers a unique case to test the respective performance of these two groups of students. Using a special data, we estimate the impact of high school background (public vs. private, types of high school) and national exam score on the university performance of students admitted for the years 1994-2011. We do not have information on family background but public-private distinction can capture some of income effect which is missing in our data. We also use additional controls for selection into graduation (time to complete) and departments. Regional variation is controlled with the location of high school. If we assume that the initial academic level and final grades are correlated, we can measure the trade-off in terms of total academic output linked to the recruitment of French-speaking students through a less-demanding specific competition versus creaming process of the national competition. Finally, we highlight the validity of national exam to sort students according to their abilities.