Statistics for Epidemiology
Written by one of the top biostatisticians in the field, Statistics for Epidemiology fills a substantial niche by explaining the ideas behind the analysis of epidemiological data without requiring a high level of mathematics and without resorting a ‘cookbook’ format. It covers the basic material for analyzing data that arise from simple epidemiological studies, including case-control and matched studies. Methodologically, it describes stratification techniques for handling confounding and interaction and the logistic regression model. The author uses a case-study approach, following a few simple examples through several method chapters rather than introducing a different example at each stage.
Univ. of California, Berkeley. Provides techniques for analyzing risk factors and disease data with step-by-step extensions to include the use of binary regression. Features simple case studies illustrating elementary analyses to more complex regression modeling. Expanded-outline format.
Outstanding book, June 5, 2006
This book is a hybrid, part epidemiology and part statistics. It is a resource for those that want to DO epidemiologic studies and ensure that they are performing and interpreting the statistics correctly. Jewell does all the little things right: he delivers the message in plain English, he explains thoroughly the foundations of the various epidemiologic measures of association, and he points out the pitfalls and potential misapplications of the presented statistical tools. The chapters on confounding and interaction are the clearest and the best that I have read. I endorse this book whole-heartedly