Performance samples on academic tasks : improving prediction of academic performance

Abstract

This thesis is about the development and validation of a performance-based test, labeled as Performance Samples on academic tasks in Education and Child Studies (PSEd). PSEd is designed to identify students who are most able to perform the academic tasks involved in an Education and Child Studies bridging program. Many Dutch universities set up bridging programs that aim to prepare students with non-university degrees in the Netherlands for Master’s programs at the university level. Some universities set up admission procedures to the bridging programs, primarily because students vary in acquired competencies, and because of the limited resources accessible to the bridging programs. The development and validation of PSEd is part of establishing an admission procedure for the Education and Child Studies bridging program at Leiden University. One implication of using performance-based tests such as PSEd for student admission purposes is that these tests can have significant incremental value beyond conventional academic predictors. In addition, performance-based tests provide a baseline of students’ abilities which can be employed to assess students’ learning growth and to evaluate the effectiveness of a curriculum. Validity and reliability of PSEd scores, as well as item properties of the test were addressed in the current thesis.