Item Parameter Drift 분석을 통한 진로성숙도 척도의 종단적 측정불변성 검증

Abstract

The purpose of this study is to examine longitudinal measurement invariance of a career maturity scale for middle school students through an item parameter drift (IPD) analysis. Data obtained from the middle school panel participating in the Daegu Education Longitudinal Study (DELS) for 3 years were analyzed. Career maturity was measured with polytomous items, and IPD items were detected for the first and the second grade and for the second and the third grade using the likelihood ratio test based on the graded response model. As a result of the IPD analysis, it was found that IPD was detected for 3 items (items 1, 2, and 10) between the first and the second grade and for 4 items (items 1, 2, 6, and 8) between the second and the third grade. Among them, items 1 and 2, which were consecutively flagged as IPD, are self-understanding questions, and the average score of self-understanding items gradually decreased although the step parameters of those items became smaller. As a result, it was found that the IPD items caused a large decrease in the average item scores of career maturity. This suggests that if the IPD items are used without testing longitudinal measurement invariance, the difference may look larger than what it actually is.