Analysis of rater-(method-)effects in empirical educational research

The research proposal aims at a systematic and comprehensive analysis of rater (or method) effects in the empirical education research. In this project, method effects are defined as latent residual variables capturing teachers' unique perspective on students' non-cognitive characteristics corrected for students' self-evaluations. Hence, latent method effects carry systematic influences of teacher inaccuracy that needs to examined thoroughly and should not be misinterpreted as a false evaluation. Using adequate latent variable modeling approaches for the analysis of multilevel MTMM designs with nested structurally different methods, it will be examined what predictor variables can explain latent method (teacher) effect at the individual and the class level. This project will fill two key research gaps. First, it will provide important information with regard to the determinants of teacher inaccuracy when judging students' non-cognitive characteristics. So far, little is known about the potential causes of method (rater) effects in empirical educational research. Secondly, an important methodological contribution toward the adequate analysis of latent factors methods in CFA MTMM models is done in this project.