Modeling educational effectiveness and equity at the school level

点击翻译按钮获取中文摘要

School effectiveness research deals with associations between school input variables (e.g. student composition, school resources, and school structure), processes within school (e.g. school climate, management style of the school principal, and instructional quality), and school output measures. Output predominantly focuses on student achievement, but also on motivational or affective variables. There are different ways to operationalize school effectiveness in empirical research. Within the proposed project, we distinguish between quality and equity. Quality is operationalized by the school specific gain in achievement adjusted for student background characteristics ("value added"). Equity is assessed by estimating school specific relations between student background variables (e.g. gender, SES, migration background) and educational outcomes.Using data from the LAU and KESS studies, the proposed project aims at answering three groups of research questions: 1. How consistent are different operationalizations of school effectiveness?, 2. What input variables are associated with school effectiveness?, and 3. how stable is school effectiveness across time and what are predictors for change in school effectiveness? Analyses for questions from groups 1 and 2 can conducted with data from LAU and KESS separately, and results can be cross-validated with data from the respective other study. To answer questions from group 3, the specific structure of the MILES data is utilized by treating both studies as two "repeated measurements" on school level. For both, quality and equity are operationalized using longitudinal data on student level. This allows to examine stability and change in school effectiveness across time and to predict school effectiveness at the second occasion (KESS) with school level predictors assessed at the earlier time point (LAU). In addition to the research questions related to school effectiveness, the proposed project will also pursue methodological questions related to the required analysis techniques. Specifically, factors determining the reliability of model based school effectiveness variables will be examined, latent and observed variable models will be compared, and the generation of plausible values for school effectiveness variables will be explored.