Proposing Difficulty and Discrimination Indices to Validate Teacher-Made Multiple-Choice Test Items: A Remedy to Variations Observed in Scholars’ Propositions
Abstract
The purpose of this research article was to propose difficulty and discrimination indices as a remedy to variations observed in scholars’ indices and check the efficacy of the proposed indices by comparing them with other researchers’ indices. To achieve this, a multiple-choice vocabulary test was constructed and was given to 4th year medical student at AMU. Then, a comparative item analysis was done with the researcher’s proposed indices and with one of other researchers. The analysis revealed that the test had several items that needed to be discarded and replaced by other better items as most of these items had poor discrimination index. This was due to researcher’s choosing the most frequent words from Academic Word List. As the test-takers were familiar with these words, they could easily identify them. In addition, it was found out that there were a few items that needed some revisions. The problems with these items were associated with having many non-functional distractors which attracted none or few of the test-takers. Besides, the comparison showed that the researcher’s indices were better in some respects than that of the other researcher. Eventually, it was recommended that the area needed further studies.
Keywords: multiple-choice, item analysis, difficulty index, discrimination index, distractor efficiency, stem, and distractors
