Comparison of Competing Approaches to Analyzing Cross-Classified Data

• In Event: Innovative Approaches to Modeling Educational Data

Abstract

Cross-classified random effects modeling (CCREM) is a common approach to analyzing cross-classified data in education. However, when the focus of a study is on the regression coefficients at level-1 rather than on the random effects, ordinary least squares regression with cluster robust variance estimators (OLS-CRVE) or fixed effects regression with CRVE (FE-CRVE) could also be appropriate approaches. These alternatives have the advantage of working under weaker assumptions than CCREM. We conducted a Monte Carlo Simulation study to compare the performance of CCREM, OLS-CRVE, and FE-CRVE. Similar to CCREM, OLS-CRVE and FE-CRVE provided negligible bias on level-1 regression coefficients, but FE-CRVE performed better in precision and Type-I error rates than OLS-CRVE. We suggest FE-CRVE as a good alternative to CCREM.

Authors

• Young Ri Lee, The University of Texas at Austin

• James Eric Pustejovsky, University of Wisconsin - Madison