검사 길이 변화에 따른 국가수준 학업성취도 평가의 안정성 분석1)

Abstract

The 2019 NAEA(National Assessment of Educational Achievement) for high school is planed to implement a new type of item to measure student subject-specific competency to apply the 2015 Revised National Curriculum. Since the new type of items are implemented to measure complex abilities rather than simple knowledge, the issue of reducing the test length was considered to solve higher-level items. Namely, the major change in the 2019 NAEA is the reduction in the number of items caused by the introduction of the new type of items. This study inspected how the test length change of the 2019 NAEA for high school had an effect on the scale stability. In the NAEA, the media attention is focused on the year trend of the proportion of below basic students, so the scale stability of the NAEA which categorizes students into basic and below basic, is main point. According to the results of this simulation study, in order to precisely categorize basic level students from below basic level students with less than 0.05 estimated RMSE at the cut-off score: at least three items should be included for each subject test with the ‘low’ difficulty level (item difficulty parameter b < -1.5) and the ‘moderate’ discrimination level (item discrimination parameter a ≥ 0.5).