Abstract

Admission into institutions of higher learning in Malawi is very competitive considering the limited capacity of higher education institutions. Admission procedures vary by institution; some offer entrance examinations, while others do not. Nevertheless, college entrance examinations provide a standardized and objective measure of the academic level of students from diverse educational backgrounds and boost a high predictive validity of student's success in college. In Malawi, the institutions which administer entrance examinations use paper and pencil mode of delivery which seem to have limited reliability among other challenges. On the other hand, computer-adaptive testing (CAT) provide a changed approach and has practical advantages that could be leveraged to overcome the challenges of PBT. However, previous studies have reached different conclusions when comparing the scores from CBT to PBT. Hence, this research assessed the reliability of (CAT) as an alternative to paper-based testing (PBT). The study took a quantitative approach with a quasi-experimental design. Data was collected from a sample of the 2022 DCE and NCE scripts of the paper-based test and the Live CAT administration. The item parameters were examined to determine the paper's quality, then, theta estimates and standard error of measurement were compared. Finally, Pearson moment correlation was used to assess the linear relationship of CAT to PBT estimated scores. The results indicate that Paper-based entrance exams are of moderate quality, CAT estimate scores more precisely than PBT, and with a correlation statistic of 0.717, the scores from CAT and PBT have a positive relationship. The findings suggest that CAT can be relied upon as an alternative to PBT in entrance examinations. Further research must be conducted on the consequences of using a misfit model in computer adaptive tests, understanding the perception of students with the transition from paper based tests to computer adaptive tests and study on minimum items for calibrating an item bank. Keywords: Computer adaptive testing, College admission criteria, Item response theory, Test delivery method.

More details

School	: School of Education
Issued Date	: 2024

Download full document