Generally lucid and narrative writing style, rigorous presentation and attention to detail: these are the ways in which Paul Allison “spoils” his readers. The work under review is no exception. It represents a much awaited second edition of a well-known book. As I haven’t had a chance to read the first edition the review will describe only on the second edition without trying to compare it to the first.
The table of contents is straightforward, covering the main topics in survival analysis with a practical emphasis on using SAS as a calculation engine. Chapters 1 and 2 provide a very thoughtful introduction to the subject. Starting from scratch, they provide a very intuitive introduction emphasizing the types of censoring and the meaning of the hazard function. Readers who have not studied survival analysis will find these two chapters extremely useful and, most importantly, will obtain a good visual understanding of where and how the journey in survival analysis begins. The next five chapters deal with different methods, namely PROC LIFETEST, PROC LIFEREG, PROC PHREG, competing risks, PROC LOGISTIC, heterogeneity, repeated event and other methods.
The presentation strategy is constant for all chapters: the author uses quite a bit of space to provide an intuitive introduction (including even some historical flashbacks) before delving into the method itself. The main focus of the book is on methods rather than theory, so the theory is presented on a “need-to-know” basis, which does not, however, undermine the quality of the presentation. The author provides relevant references for further theoretical exposition. First time students of survival analysis might feel a bit perplexed with the lack of theoretical presentation; the additional references provided will be of help.
The amount of detail presented with respect to the computational methods (both the code and SAS output are shown) used in SAS, along with the explanation and description of the results, is rarely seen even in applied books. What contributes to such a lucid presentation style? The author spends pages and pages to properly present and describe the data used, the method applied, the computational aspects and, last but not least, the results. The way that this is done is what makes a difference. While reading one feels that one is receiving personal and clear communication from the author. Furthermore, conclusions in each chapter provide a nice flashback, just in case a reader missed something along the way.
Chapter 9 provides general guidance of what methods to use for certain types of problem. It is definitely good advice for novice and not so experienced users of survival analysis techniques. I find this to be of great importance to students who don’t have much experience; questions like this often flow through their minds.
This book should be on the bookshelf every practitioner who uses survival analysis methods, especially those who need a jump start. With this book one can really take a seat and go for a smooth ride.
Ita Cirovic Donev holds a Masters degree in statistics from Rice University. Her main research areas are in mathematical finance; more precisely, statistical methods for credit and market risk. Apart from the academic work she does statistical consulting work for financial institutions in the area of risk management.