| Jason Meyers
University of Texas, Austin
The impact of the inappropriate modeling of cross-classified data structures
FINAL REPORT:
This dissertation involved a methodological investigation into the functioning of cross-classified random effects modeling (CCREM). CCREM is an extension of hierarchical linear modeling (HLM) that is designed for use with multilevel data that are not purely hierarchical in nature. Two common examples of cross-classified data structures in educational research are students nested within neighborhoods and schools and students nested within middle schools and high schools. Although use of CCREM has increased in various disciplines such as medicine and is mentioned in most multilevel textbooks (for example, Raudenbush & Bryk, 2002; Hox, 2002; Snijders & Boskers, 1999), it is seldom used in educational research.
Two studies were conducted that compared use of CCREM to HLM with cross-classified data. A Monte Carlo Simulation Study was conducted in order to investigate potential factors affecting the need to use CCREM instead of HLM as well as the impact of ignoring the dataÕs cross-classification. As a follow-up study, CCREM and HLM were applied to a large-scale, cross-classified national data set in order to illuminate differences in the results between methods.
Results of both studies indicated that when using HLM instead of CCREM, the fixed effect estimates were unaffected, but the standard error estimates associated with the variables modeled incorrectly were biased. In addition, the estimates of the variance components displayed bias. The observed bias was related to the proportion of the total variance that was between each cross-classified factor, the sample size, and the similarity of the cross-classified factors.
Recent passage of the No Child Left Behind Act (NCLB; Public Law 107-110, 2001), has made understanding the functioning of the techniques available to model cross-classified situations particularly important. NCLB emphasizes change over time (Adequate Yearly Progress) and allows students who attend low-performing schools to leave for a ÒbetterÓ school within the school district. As more and more students begin to transfer between schools and attend schools outside of their neighborhood, the evaluation of institutional effectiveness needs to be enhanced. Because measuring studentsÕ and schoolsÕ educational achievement and progress has taken on added emphasis under the NCLB Act, and because there are likely to be more mobile students under NCLB, the results of this methodological investigation have important policy implications.
Back to Funded Dissertation Grants Page |