Development of Grade Six Mathematics Item Bank by Applying Item Response Theory
The purpose of the study was to construct a Mathematics Item Bank for Grade Six students by using item response theory (IRT). The participants were selected by using stratified random sampling technique. Descriptive research design and quantitative survey method were used in this study. Lesson 1 to 5 of Grade Six Mathematics Textbook (1) and Lesson 1 to 6 of Grade Six Mathematics Textbook (2) were selected as the content area for item bank. In the preliminary test, after item analysis with 2-PLM of IRT, out of 219 multiple-choice items, 170 items were remained for three field tests (Form A, Form B and Form C) and each form contained about 56 items. These field tests were administered to 1513 Grade Six students from the selected schools in Yangon Region in which 505 students were administered for Form A, 504 students for Form B and 504 students for Form C. Then, 164 items of field test (56 items in Form A, 55 items in Form B and 55 items in Form C) had the discrimination range from 0 to +2 and the difficulty range from −3 to +3. Therefore, all items were stored in item bank. To support advantage of item bank, a new sample test was developed by using items from item bank. It was found that this test was more appropriate for examinees whose ability (θ) range is −1.65 to +1.95 and provided the highest information for the examinees with ability level −0.05.