Model-based academic dismissal policies; a case-study from the Netherlands

Abstract

Many higher education institutions use a policy for academic dismissal. In the Netherlands, the academic dismissal policy is such that students with fewer credits than a certain threshold after their first year, are expelled. This article employs the beta-binomial model to assess whether this method succeeds in filtering those who have potential from those who do not and what the optimal level of the threshold is. The model considers 13,234 students in three consecutive cohorts from around fifty different bachelor's degree programmes at the University of Groningen. We found that demanding 45 out of 60 credits constitutes a fair threshold for this institution. Although a strict dismissal policy has only a minor effect on cohorts, it can have a major effect on specific groups of students. The software employed here is made available.

A tutorial on regression-based norming of psychological tests with GAMLSS.

“…For example, it has been used to model the scales of the nonverbal intelligence test SON-R 2–8 (Tellegen & Laros, 2017). The BB distribution is theoretically motivated if one has a scale composed of binary items where the distribution of the ability of the individuals follow a beta-distribution and the items are of equal difficulty (Wilcox, 1981), or where the difficulty of the items that make up the scale follow a beta-distribution and the ability of the individuals is equal (Albers, Vermue, de Wolff, & Beldhuis, 2018). Under either one of these assumptions, observed scales scores follow the BB distribution.…”

On Natural Variation in Grades in Higher Education, and Its Implications for Assessing Effectiveness of Educational Innovations

“…Another layer of natural variability could occur due to external changes to the system. For example, Albers, Vermue, de Wolff, and Beldhuis () studied the effect of a change in academic dismissal policy on the performance of all first‐year students at the university. Here, the natural variation also included factors such as discontinuation of certain courses, changes in lecturers, and educational innovations in other courses.…”