Application of rasch model in measuring the quality of health and wellness final exam questions

• Main Authors: Norlia Goolamally, Suhaila Abdul Hamid, Ahmad Zulkarnain Ramli, Raijah A. Rahim
• Format: Conference or Workshop Item
• Published: 2019
• Subjects: LC5201 Education extension. Adult education. Continuing education
• Online Access: http://library.oum.edu.my/repository/1168/1/library-document-1168.pdf; http://library.oum.edu.my/repository/1168/
• Institution: Open University Malaysia

The assessment of the Health and Wellness module includes two components, 60% for the continuous examination and 40% for the final examination paper. In this paper we discussed the performance of students in the final examination paper which comprises 40 MCQs (MCQ). The 40 MCQ were constructed based on the Table of Test Specification developed by the subject matter experts at the faculty. The difficulty level of the questions was based on the 5:3:2 proportions. This ratio produces 20 MCQ at lower difficulty level, 12 MCQ at moderately difficult and 8 MCQ at higher difficulty. Results of 764 students were analysed based on their ability and item difficulty of the 40 MCQ using RaschWinsteps 4.01 software. Person Item Map Distribution (PIDM) showed that the item measure is between -3.62 to +4.97 logits and the person measure is between -1.28 to +4.09 logits. This finding indicated that while person ability exceeded item difficulty at the lower logit continuum of the Rasch model, at the upper logit continuum, some items are at higher difficulty level compared to students’ ability. Item difficulty and separation indices statistics for both person and item are also of concern in this study. To ensure and enhance reliability and validity of the questions, it is highly recommended that the subject matter experts review the questions before applying it again in another exam or depositing it in the item bank system.