Family background and school effects on student achievement: A multilevel analysis of the Coleman Data

Background/Context: A main objective of the Equality of Educational Opportunity Survey (EEOS), conducted in 1965, was to document the lack of availability of equal educational opportunities for minority students in public schools. Another equally important objective was to reveal specific inequalities in facilities and resources available to students in predominantly minority or predominantly White schools. Coleman et al. (1966) analyzed the EEOS data and found surprisingly few differences between the characteristics of schools attended by minority and White students. As a result, Coleman et al. concluded that school characteristics are not strongly related to student achievement in the presence of family background and that family inputs are much more valuable predictors of student achievement than school inputs are. Purpose/Objective/ Research Question/Focus of Study: The present study revisited this issue about the importance of schools in promoting student achievement and reanalyzed the EEOS 12th grade data using multilevel models. Our sample included 12th graders in public schools in the U.S. in 1965. We sought to determine the predictive efficacy of school characteristics on student achievement net of the effects of family background. The overarching question motivating this research is: Would Coleman and his colleagues have reached the same conclusions had they had available today's multilevel modeling statistical methods that are more appropriate for determining school effects? Research Design: We used both regression and multilevel models to gauge school effects and compared our findings to those reported by Coleman et al. Our estimates infer strictly correlational, not causal, associations between school characteristics and achievement. Conclusions/Recommendations: We found considerable and significant between-school variance in achievement, which suggests school effects. Similarly, the observed school characteristics used in our models explained a substantial proportion of the between-school variation in achievement. Our results also indicated that schools play meaningful roles in distributing equality or inequality of educational outcomes to females, minorities, and the disadvantaged. These results are in congruence with recent studies that examined school effects from the 1970s to the 1990s using U.S. national probability samples of students (Konslantopoulos, 2006; Konstantopoulos & Hedges, 2008).