Item Analysis for a Better Quality Test

Abstract

This study is a small-scale study of item analysis of a teacher’s own-made summative test. It examines the quality of multiple-choice items in terms of the difficulty level, the discriminating power, and the effectiveness of distractors. The study employed a qualitative approach which also used a simple quantitative analysis to analyze the quality of the test items through the document analysis of the teacher’s English summative test and the students’ answer sheets.  The result shows that the summative test has more easy items than difficult items with the ratio of 19:25:6 while they should be 1:2:1 for easy, medium, and difficult.  In terms of the Discriminating Power, there are 3, 13, and 16 for excellent, Good, and satisfactory level, but there are 17 and 2 for poor and bad levels of Discriminating Power.  There are 43 (21.5%) of all distractors which are dysfunctional which, in turns, makes the items too easy which also makes the items fail to discriminate the upper-group students from the lower ones. Therefore, the 43 dysfunctional distractors should be revised to alter the difficulty level and improve the discriminating power.  This research is expected to serve as a reflective means for teachers to examine their own-made test to ensure the quality of their test items.