Date of Award
12-2014
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Biology
First Advisor
Kenneth G. Ruit
Abstract
Validity and reliability of scores obtained on Multiple-choice questions (MCQs), as well as the benefits of test-enhanced learning, have been of interest to medical educator scholars. Presented in this dissertation are four composite studies on these themes. The following hypotheses were tested:
1. Increased MCQ distractor functioning increases the validity and reliability of obtained scores.
2. Correction of item writing flaws (along with enhancement of tested cognitive level) and replacement or removal of non-functioning distractors equally improves psychometric characteristics of MCQs.
3. Repeated testing via free-response items enhances the retention of knowledge of human anatomy, compared with repeated or once-testing via multiplechoice questions.
Validity and reliability of scores obtained on MCQs was noted to rise as a result of increased MCQ distractor functioning, discrimination amongst high and low performing students was found to be equally improved via removal of flaws and nonfunctioning distractors, and short-term (up to four weeks) retention of human anatomy knowledge was found to be enhanced by repeated testing via free-response questions.
Raising MCQ quality by addressing flaws and low distractor functioning, and using nostakes repeated retrieval practice, is advised for improvement in the assessment and learning practices in pre-clinical medical education.
Recommended Citation
Ali, Syed Haris, "Psychometrics and test-enhanced learning in a patient-centered learning curriculum" (2014). Theses and Dissertations. 1074.
https://commons.und.edu/theses/1074