Comparison of Generalized Linear Model and Generalized Linear Mixed Model – An Application to Low Birth Weight Data

Background: Generalized Linear models are mostly fitted to data that are not correlated. However, very often data that are collected from health and epidemiological studies are correlated either as a result of the sampling methods or the randomness associated with the collection of such data. Therefore, fitting generalized linear models to such data that produce only fixed effects could lead to over dispersion in the model estimates. Objectives: The objective of this study is to fit both generalized linear and generalized linear mixed models to a correlated data and compare the results of the two models. Methods: Logistic regression is employed in fitting the generalized linear model since the dependent variable in the study is bivariate whilst the GLIMMIX model in SAS is used to fit the generalized linear mixed model. Results: The generalized linear model produces over dispersion with higher errors among the parameter estimates than the generalized linear mixed model. Conclusion: In dealing with a more correlated data, generalized linear mixed model, which can handle both fixed and random effects, is preferable to generalized linear model.