Equivalence of Computer-Based and Paper-Based Testing for the topic of Geometry

Abstract

This study investigated whether paper-based testing (PBT) and computer-based testing (CBT) can be considered equivalent testing modes for geometry in the VMBO-KB level final exams. Parallel tests were designed and administered in both delivery modes to 129 third-year VMBO-KB level students using a research design that controlled for order and version effects. Equivalence was measured by two components: similarity in rank order between scores on the test modes and similar characteristics of the score distributions. The disattenuated correlation between the test scores in both conditions was found to be substantially high. However, for the PBT, the mean was higher, the variance lower and the distribution was more negatively skewed. Moreover, an investigation of item-level differences led to the conclusion that practicing questions decreased the test mode effect, while questions that required more mathematical thinking skills increased the test mode effect. Nevertheless, since the rank order was similar and rescaling eliminated any differences in characteristics of the score distributions, PBT and CBT can be considered equivalent test modes for VMBO-KB final exams.