Event History Modeling: A Guide for Social Scientists
by Janet M. Box-Steffensmeier and Bradford S. Jones
This book covers event history modeling in depth. It was suggested by George Ball during his job talk at IU. I needed the insights provided by this book for better understanding of my call center project. In that project, we are looking at the determinants of how long a customer will stay on hold once they hear an announcement about the expected wait. We model the patience of the customer as certain covariates vary. This book encompasses the field of survival analysis, which is named for the medical literature that studies how certain covariates affect survival rates.
The book is about 200 pages, and I highly recommend it for someone that is new to survival analysis or event history modeling. It covers the Cox Proportional Hazards model in depth, along with various parametric models. It covers issues of model selection, time-varying covariates, mis-specified models that go against assumptions, heterogeneity, and multiple events. Of most use to me was Chapter 8, which discusses diagnostic methods for determining if your model is performing properly. I will apply these methods to my project to ensure it is working as expected (and to appease referees).