Big data analyses of educational achievement

Many important questions in education require analysing large and complex datasets including measures of educational achievement, cognitive skills, environmental factors, and sometimes even genetic factors. We have access to several such datasets: The Eden cohort (N=400 to 1400), the Elfe cohort (N=9000), the DEPP 2007 panel (N=35000), the DEPP 2011 panel, the DEPP maternelle panel (N=35000), as well as more recent data from the national evaluations in many grades (N=800 000 per grade). Other international datasets may also be available (e.g., PISA, PIRLS, TIMMS, TEDS, BiB, Millenium cohort…).

Projects may bear on the following questions, among others:

• What are the early predictors of reading and math ability? to what extent are they mediated by early cognitive skills? to what extent do they differ between boys and girls?
• What are the characteristics of schools, teachers, pedagogical practices and manuals that produce superior achievement?
• Bring your own question!

Student profile: data science, statistics, econometrics, psychology, cognitive science. Any other profile, but with strong analytical skills.