Using Capacity to Model School Effectiveness

Abstract

This study examines concepts and conditions that show promise for creating more accurate indicators of school performance than those currently in use and dominated by student outcome measures. Data were collected from randomly sampled students and their parents from the 5th, 8th, 9th, and 11th grades in 71 urban-setting schools, and from the teachers and principals of these schools. The study hypothesized that identified school-level capacities (organizational capacity, instructional capacity, learning capacity, and home capacity) are latent variables comprised of social indicators predictive of school outcomes (student achievement on standardized tests and student psychological health measures). Using confirmatory factor analysis and structural equation modeling through Amos 7.0 in SPSS, this study investigated the formation of these capacities and their effects on one another and student outcomes. Confirmatory factor analysis provided empirical results that support the conceptualization and measure of each hypothesized capacity as consisting of the proposed observed variables. The structural equation model demonstrated the relationships that exist among capacities as well as their effect on achievement and student psychological health. Organizational capacity relates positively to home capacity, instructional capacity, and learning capacity. Home capacity relates positively to learning capacity and instructional capacity. Learning capacity is predictive of greater math achievement and student psychological health. Instructional capacity is predictive of greater math achievement. These findings provide evidence for how school performance can be more accurately and comprehensively understood, assessed, and acted on.