Regression Methods in Biostatistics: Linear, Logistic, Survival, and Repeated Measures ModelsTheprimarybiostatisticaltoolsinmodernmedicalresearcharesingle-outcome, multiple-predictor methods: multiple linear regression for continuous o- comes, logistic regression for binary outcomes, and the Cox proportional h- ardsmodelfortime-to-eventoutcomes. Morerecently,generalizedlinearm- els and regression methods for repeated outcomes have come into widespread use in the medical research literature. Applying these methods and interpr- ing the results requires some introduction. However, introductory statistics courses have no time to spend on such topics and hence they are often r- egated to a third or fourth course in a sequence. Books tend to have either very brief coverage or to be treatments of a single topic and more theoretical than the typical researcher wants or needs. Our goal in writing this book was to provide an accessible introduction to multipredictor methods, emphasizing their proper use and interpretation. We feel strongly that this can only be accomplished by illustrating the te- niques using a variety of real datasets. We have incorporated as little theory as feasible. Further, we have tried to keep the book relatively short and to the point. Our hope in doing so is that the important issues and similarities between the methods, rather than their di?erences, will come through. We hope this book will be attractive to medical researchers needing familiarity with these methods and to students studying statistics who would like to see them applied to real data. |
Contents
1 | |
Basic Statistical Methods | 29 |
Linear Regression 69 | 68 |
Predictor Selection | 133 |
Logistic Regression | 157 |
Survival Analysis 211 | 210 |
Repeated Measures Analysis | 253 |
Generalized Linear Models | 291 |
Complex Surveys 305 | 304 |
Summary | 317 |
323 | |
333 | |
Other editions - View all
Common terms and phrases
addition adjusted alternative analysis approach assessing association assumed assumption average baseline binary calculate categorical causal effect Chapter coefficient compared compute Conf confidence intervals confounding consider continuous correlation covariates Cox model defined difference discussed distribution effect equal errors estimate example exercise factors function gives glucose groups hazard ratio hypothesis important increase indicator individuals interaction interest interpretation less levels likelihood linear model logistic model logistic regression mean measures mediation methods mg/dL multiple normal Note observations obtained odds ratio outcome parameter patients pattern plot points population potential prediction predictor primary Prob probability Problem procedure proportional random reference regression model relationship represent residuals risk sample scale Sect selection shown shows similar simple specific standard Stata statistically statistically significant Table transformation treatment values variable variance weight women
Popular passages
Page 327 - Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/Progestin Replacement Study (HERS) Research Group.