A psychometric analysis of a newly developed summative, multiple choice question assessment adapted from Canada to a Middle Eastern context

Abstract

Introduction: Accreditation necessitates that assessment methods reflect the standards established by the accrediting body. The process of adapting assessments to a new context can present unique challenges with uncertainty around psychometric defensibility of the adapted exam. Methods: A psychometric analysis of a summative multiple-choice-question (MCQ) assessment, adapted from Canada, for graduating pharmacy students from a Canadian accredited program in Qatar was conducted. Rates of difficult items, item discrimination measured by point biserial correlation (rpb), and non-functioning distractors (NFDs) were calculated to identify deficiencies and challenges with an adapted exam. Challenges encountered throughout the adaption process and recommendations were documented. Results: Overall score of a 90-item, four option, MCQ exam ranged from 46.7% to 78.9% (mean of 61.9%). For difficulty, there were 17 items with less than 30% of students answering correctly, while 29 items had unacceptable or poor discrimination (rpb < 0.1). NFDs occurred in 78 items with 49 containing at least two NFDs. Discussion and conclusions: Difficulty of the exam was deemed acceptable yet discriminator ability requires improvement. The high frequency of questions with NFDs suggests that faculty have difficulty developing plausible distractors for an adapted MCQ exam. This could be due to a lack of training or requirement for inclusion of too many distractor options. While it is feasible to implement an assessment adapted from a different learning environment, measures need to be taken to improve psychometric defensibility. The high number of questions with NFDs indicates that the current method of exam development does not encourage the incorporation of functional distractors. ? 2018 Elsevier Inc.