We compared the psychometrics of quiz questions randomly selected from a test bank with the psychometrics of quiz questions the instructor had selected from the bank for quality and modified (if necessary). On multiple psychometric indices, the instructor selected/modified questions were superior to questions randomly selected from the test bank. Most notably, when compared with instructor written/modified questions, randomly selected bank questions were nearly 6.5 times more likely to contain a distractor that drew more responses than the correct answer. Details and implications are discussed.
This is an original work
This work has not been previously published
IRB approval verification
Richman, H., & Hrezo, M. (2017). The Trouble with Test Banks. Perspectives In Learning, 16 (1). Retrieved from http://csuepress.columbusstate.edu/pil/vol16/iss1/2