Abstract
This study investigates how distractors function in multiple-choice listening tests. Conventionally, each distractor is evaluated by its attractiveness. In other words, distractors that can plausibly be chosen by test takers are considered to perform well on the test. On the other hand, distractors that are chosen by few test takers are recognized as performing poorly. Considering that test takers have to choose only one among several options, it can be assumed that some unselected distractors may have in fact performed adequately. Therefore, it is prudent to independently analyze the attractiveness of distractors. A total of 75 Japanese university students evaluated their confidence in selecting both correct and incorrect answers. The results indicated that (a) the least chosen distractors were not always the least attractive, (b) less proficient listeners were more likely to be allured by distractors, and (c) more proficient listeners were more likely to answer with higher confidence. The researcher explains the process of eliminating distractors and reevaluates the unselected distractors.
