Alternative Methods for Modeling Clustered Ordinal Data

• In Event: Advances in Multilevel Modeling

Abstract

When modeling clustered ordinal data, a multilevel cumulative logit model (MCLM) is typically used. Depending on the research questions and inferences a researcher would like to draw from his/her findings, generalized estimating equations (GEE), a type of population average model, may be used as an alternative to MCLM. Our aim in this paper is to investigate appropriateness of methods for modeling clustered ordinal data under different study conditions with varying degrees of cluster sizes, number of clusters, magnitudes of variance components, number of categories, and distributions of the ordinal outcome using a Monte Carlo simulation study. Recommendations on model selection will be provided, recognizing interpretation differences between different strategies.

Authors

• Meng-Ting Lo, National Yang Ming Chiao Tung University

• Ann A. O'Connell, The Ohio State University