Unix

Survival Analysis Using the SAS System: A Practical Guide

Format: Paperback
Author: Paul D. Allison
ReleaseDate: 13 December, 1995
Publisher: SAS Publishing
Rating:

Learn By Doing
If you are working largely on your own, and you learn best by doing, you cannot--as far as I know--do better than Allison's book. If you have data that fit the general category "time to event," and are not suitably analyzed by repeated measures ANOVA, you are probably looking at doing a survival analysis (also known by several other names). Of course it all but locks you into using SAS for analyses, but his explanations of proportional hazards and other models are the best I've found among a dozen textbooks and stats package manuals (some of which made sense only after reading Allison). What makes this book so good is that it will have you running your analyses in just hours. The examples are superb take-off points. I was not a SAS user before reading the book and therefore took a little extra time to figure out dataset manipulations and such in SAS, but that was minor effort compared to the rewards of having Allison's clearly written book as a guide. The price of this book represents only a fraction of its value. .

Extraordinarily Clear and Useful
This book makes it easy to grasp complex ideas, provides comprehensible examples, gives sample SAS code so that implementing the methods is as straightforward as possible. I've used a number of this author's books and they all share in common lucidity, utility, and rigor. Plus, it is clear that the author is a subtle and first-rate methodologist, who innovates in this area as well as teaches it.

Best how-to book on survival analysis using SAS. Very useful
I found it invaluable in conducting my research. This book is well-written, well-organized, and very practical. My only recommendation for the author for his next edition is to include a chapter on dealing with correlated event times, like time-to-promotion and time-to-quiting in his policemen example (pg 249).