---

Generalized linear mixed models are commonly employed to tackle a wide variety of data analysis problems in a multitude of settings. In these data settings understanding the statistical results can be a bigger issue than generating the numbers from the analysis in the first place. Communicating the results of an analysis can be a challenge as at times there is not a clear picture of what is going on and one may see different results between a simple aggregate analysis and the results of a regression analysis. The goal of this monograph is to bridge this gap and develop tools for understanding this sort of data analysis. This document addresses 3 aims: to review statistical methods for the analysis of count data in hierarchical settings; to examine graphical methods for presenting findings and evaluating confounding; and to illustrate these methods with actual data.

---

Alex Kiss received his Ph.D. in Biostatistics from Columbia University. He has taught courses at both Columbia University and the University of Toronto where he currently holds an appointment. His publications include over 100 peer reviewed journal articles in the fields of statistics, medicine and law.

---