A Comparative Study Of Four-and Three-option Multiple-choice Items In Listening Comprehension Test

This thesis probes into the influence of different numbers of options per item on the item parameters, distractor performance and the test-taking process of multiple-choice items in listening comprehension test.In the study, the four-option multiple-choice listening test was converted to a three-option version by eliminating the least frequently endorsed option in each item. The two tests were all the same except that the three-option test had one distractor removed per item. The two tests were distributed to two groups and were taken by 123 and 126 students respectively. After the tests, 10 students from each group were asked to tell their test-taking process.The major findings of the present study are briefly summarized as follows:When the least endorsed distractor in each item in the four-option test is deleted, the general performance of test-takers, the test reliability and item difficulty do not change significantly. The item discrimination increases in a significant way. The distractor analysis shows that the mean actual equivalent number of options in the three-option test is significantly larger than that in the four-option test. Although the mean numbers of discriminating options per item in both tests are not significantly different, the three-option test demonstrates higher efficiency in providing discriminating distractors. The interview data show that few test-takers resort to blind guessing since the item difficulty is appropriate and that the four-option items are more susceptible to testwiseness since the fourth options provide clues leading to the correct answer. Thus, it can be concluded that tests consisting of three-option items are at least equivalent to or better than tests consisting of four-option items.