ANALYSIS AND EVALUATION OF MULTIPLE-CHOICE TEST ITEMS AND TEST DESIGN: A STUDY ON APPLICATION OF ITEM RESPONSE THEORY

Abstract

This article presents the findings on applying the item response theory for a 2-parameter model in analyzing and evaluating question items and designing multiple-choice tests. Based on the results of data analysis by R (package ltm) of non-majored students’ performance in English 1 exam papers used in Dong Thap University from 2017 to 2021, the study identified the satisfactory items which could meet the exam requirements and unsatisfactory question ones for further adjustment or improvement. Among the unsatisfactory items, some abnormal characteristics, seriously violating the tests’ limitations in terms of difficulty and discriminate level, must have been definitely removed from the test papers.  In addition, the study findings also show that the test items had a relatively low level of measuring students’ competence (below 0.0 according to the competence scale). Finally, the study introduced the way of applying the information curve tool to design test items to help accurately measure the students’ competence based on the characteristic parameters of items.