Modeling Partially Cross-Classified Multilevel Data

• In Event: Multilevel Modeling Methods and Applications

Abstract

This article proposed a partial cross-classified random effects model (partial CCREM) for the analysis of data that contain both partly nested and partly cross-classified individuals. An empirical example was used to illustrate the proposed model. A simulation study was conducted to compare the proposed model with two other approaches that treat the partially cross-classified data as either fully nested or fully cross-classified. Results showed that the partial CCREM demonstrated desirable statistical properties in terms of bias and variance of the parameter estimates. Both the fully nested model and the fully cross-classified model suffered from biased estimates of variance components and standard errors of fixed effects.

Authors

• Kevin James Cappaert, University of Wisconsin - Milwaukee

• Jaime Leigh Peterson, University of Iowa

• Wen Luo, University of Wisconsin - Milwaukee