| Kazuaki Uekawa
University of Chicago
Children's achievement and social backgrounds: An international comparison
FINAL REPORT:
This study compares 42 national education systems in terms of social inequality of schooling. Here, social inequality refers to the state in which the quality of student experience in school (e.g., academic achievement and instructional experience) is partially determined by characteristics of students' social ascription (e.g., race, gender, and social class) rather than by their merit and effort. Although social inequality of schooling is a common phenomenon in industrialized nations, some nations may be more egalitarian or less egalitarian than are others in educational opportunities and in educational outcomes. Explored is a theory as to why this international variation is possible. I propose that nations are different in terms of the processes that suppress the influence of social ascription, such as social class, race, and gender. Four of such processes are "discursive mediation," "compositional mediation," "organizational buffer mediation" and "pedagogical mediation" and each describes how the disadvantages associated with social ascription are minimized in schools. Then, I explore why some nations may activate more mediation processes than do others, thus, making the nations different in their ability to promote equity of schooling. The particular focus is on institutional characteristics of national education systems and on the question of "how do nations' curricula, examinations, and selection influence the workings of the mediation processes?" I test these theories by conducting empirical analyses on three kinds of student experiences. The first analysis is about the growth of diversity in student achievement, the second is about teacher instructions, and the third is about student achievement. The goal of these exercises is to answer the larger question, "how and why is the influence of social class on student experiences different across nations?" The data is TIMSS (Third International Mathematics and Science Study, collected in 1995 from 42 nations) and main methods are multilevel statistical model and Rasch model.
Back to Funded Dissertation Grants Page |