A Study on the Cumulative Advantages and Disadvantage of Academic Achievement Gaps Using a Growth Curve Model : Realization and Measurement

The purpose of this paper is to examine aspects of cumulative advantage in Japanese society by analyzing academic achievement from a follow-up survey of elementary and junior high school students. In Japan, analysis with panel data has revealed aspects of academic achievement gaps. However, the findings are inconsistent because (1) the measurement methods of the variables, (2) the grades analyzed, (3) the regions surveyed, and (4) the analysis methods are not the same. Hence, it is not possible to make rigorous comparisons as to whether the achievement gap “widens” or “remains” as grades increase. In addition, scientific research needs to be constantly validated through panel studies. In recent years, the value of re-examination studies, in which another thirdparty researcher verifies the results of a study that has already been published for confirmation, has begun to be reevaluated. Therefore, there is a need to revisit the findings on academic achievement gaps by analyzing new panel data. The data used in this paper are academic achievement panel data obtained from a survey conducted on elementary and junior high schools in a regional core city (population of about 350,000). The analysis targets fourth- through eighth-grade students in the 2009 birth cohort. The academic achievement data is from the Norm Referenced Test (NRT), and the students ʻSocio-economic Status (SES) is the parentsʼ educational background and annual household income obtained from a questionnaire survey for parents. A growth curve model was used as the analytical method. The findings are as follows. (1) The SES variable is significant and positive for the intercept of the academic achievement score, and the academic achievement gap is observed at the 4 th grade. (2) The SES variable is significant and positive to the slope for Japanese and science, and the achievement gap widens for Japanese and science. (3) In math and social studies scores, the SES variables were not significant to their slopes. Hence, the academic achievement gap does not change with age in math and social studies. What these analyses indicate is that changes in the academic achievement gap “vary from subject to subject.”