Evaluation of three multiple-choice assessment methods in a human factors engineering course

This study investigated students objective performance and subjective preference for the conventional, liberal, and confidence marking multiple-choice methods. Eighty-one university students answered multiple-choice quiz questions on human factors engineering using each of the three methods and then provided feedback on each method. Relative to the conventional method, the liberal and confidence marking methods were useful for extracting information about partial knowledge and increased the variance of performance scores. The different multiple-choice methods could cause students to change their response criterion in performing the assessment, whereas the ability to discriminate correct answers from distracters did not vary across the multiple-choice methods. Regarding subjective preferences, the conventional method was the most preferred followed by the liberal method and then the confidence marking method. The findings of this study would be useful for the selection of effective multiple-choice methods for academic assessments. © 2012 Taylor and Francis Group, LLC.