Practical randomly selected question exam design to address replicated and sequential questions in online examinations

Abstract

Examinations form part of the assessment processes that constitute the basis for benchmarking individual educational progress, and must consequently fulfill credibility, reliability, and transparency standards in order to promote learning outcomes and ensure academic integrity. A randomly selected question examination (RSQE) is considered to be an effective solution to mitigate sharing of questions between students by addressing replicated inter-examination questions that compromise examination integrity and sequential intra- examination questions that compromise examination comprehensivity. In this study, a Monte Carlo approach was used to design six examination schemes for the purpose of generating and evaluating 600 RSQEs in order to investigate the effects of RSQE design on replicated inter-examination and sequential and intra-examination questions. Results revealed that the number of randomly selected questions from the pool and the number of sub-pools inversely affected the replication and sequencing of the examination questions. Thus, by designing the RSQE in many sub-pools, in equivalence to the number of examination questions and selecting only one question from each sub-pool, and updating the sub-pools after each examination, the passing of information can be prevented, ensuring the integrity of the examinations.