Selvin S. Statistical Tools for Epidemiologic Research

Oxford University Press, 2011. - 510 p. - ISBN: 0199755965, 9780199755967.In this innovative new book, Steve Selvin provides readers with a clear understanding of intermediate biostatistical methods without advanced mathematics or statistical theory (for example, no Bayesian statistics, no causal inference, no linear algebra and only a slight hint of calculus). This text answers the important question: After a typical first-year course in statistical methods, what next?Statistical Tools for Epidemiologic Research thoroughly explains not just how statistical data analysis works, but how the analysis is accomplished. From the basic foundation laid in the introduction, chapters gradually increase in sophistication with particular emphasis on regression techniques (logistic, Poisson, conditional logistic and log-linear) and then beyond to useful techniques that are not typically discussed in an applied context. Intuitive explanations richly supported with numerous examples produce an accessible presentation for readers interested in the analysis of data relevant to epidemiologic or medical research.Two measures of risk: odds ratios and average rates
Tabular data: the 2× k table and summarizing 2 × 2 tables
Two especially useful estimation tools
Linear logistic regression: discrete data
Logistic regression: continuous data
Analysis of count data: Poisson regression model
Analysis of matched case/control data
Spatial data: estimation and analysis
Classification: three examples
Three smoothing techniques
Case study: description and analysis
Longitudinal data analysis
Analysis of multivariate tables
Misclassification: a detailed description of a simple case
Advanced topics