New rules, new tools: Predicting academic achievement in college admissions

Admission procedures for higher education aim to assess the fit between students and programs through selection or matching procedures. Just as in personnel selection, decisions are commonly made based on intuition and often lack scientific support. So, the central question in this thesis was: How should we select students?

I showed that:
(1) Using questionnaires to measure personality and motivation does not work in selection, because of socially-desirable responding by the applicants. I conclude that motivation and personality are very difficult to measure in selection contexts and that no solution is available yet.
(2) The best method to select students is to give them tasks that are highly similar to what they would have to do in the study program they apply to, also referred to as curriculum sampling. Curriculum samples predicted academic achievement well, and provide applicants with insight into their fit to the program. In addition, applicants appreciated the use of curriculum samples, while the use of lottery and high school grades was not appreciated.
(3) However, even if valid methods are used, the effects of selection on study performance are small when most applicants are suitable and few are rejected, as is the case for most Dutch higher education programs.

A major challenge in establishing effective and fair selection decisions in education and in organizations is to overcome the reliance on intuitions and gut feelings; Scientific knowledge based on selection psychology should be utilized in practice much more.