Mode Effects in Correcting Students’ Errors: A Comparison of Computer-Based and Paper-Pencil Tests

Abstract

Computer-based testing (CBT) is considered to have several advantages compared with paper-pencil-based tests (PPT). It allows the embedding of different formats (e.g. audio and video files), quick and (semi-)automatic scoring and therefore opportunities for adoptions, measuring of additional information, such as response times, inclusion of a broader variety of test subjects as well as the avoidance of errors in data transmission (e.g. ambiguous and illegible information) and analysis. Overall, it is generally accepted that CBT saves resources such as time, materials and personnel in comparison to PPT situations. Against this background we favour the CBT approach for data collection. Nevertheless, CBT also entails several disadvantages that need to be considered, such as hardware or software problems (e.g. freezing, crashing, display errors, or when the same content is displayed differently). In addition, influencing factors conditioned not only by the technical aspects of a CBT situation but also due to individual characteristics of the participants, are discussed in the literature. Main concerns in using CBT address questions of comparability and equivalence of paper-pencil testing (PPT) and computer based testing (CBT), which has provoked a long history of research with regard to so-called mode effects. To answer the question whether the tests used to evaluate the training programme for prospective teachers can be used in CBT as well as in PPT, we conducted a pilot study and analysed whether mode effects exist. Results indicate that there is no systematic influence of the testing mode on test persons’ performance.