Applied Logistic Regression
David Hosmer
University of Massachusetts
October 22 -23
Abstract:
This two-day short course will present an introduction to using the logistic regression model. Topics to be covered will include model formulation, parameter estimation, estimation and interpretation of odds-ratios and probabilities, model building strategies, assessment of goodness-of-fit, and presentation and interpretation of results. The course will consider the logistic regression model for binary, multinomial, and ordinal scaled outcomes.
The course will be taught by Professor David W. Hosmer of the Biostatistics Department of the University of Massachusetts. Prof. Hosmer has over 10 years experience teaching similar short courses to statisticians, epidemiologists, physicians and other subject matter scientists.
The course will be based upon selected chapters and sections in Professor Hosmer's recent text, Applied Logistic Regression. Co-authored by Professor Stanley Lemeshow of Ohio State University, the second edition of this widely referenced text was published in 2000 by John Wiley & Sons. Topics to be covered from this edition, with sections and page numbers noted, appear below:
1 Introduction to the Logistic Regression Model
1.1 Introduction, 1
1.2 Fitting the Logistic Regression Model, 7
1.3 Testing for the Significance of the Coefficients, 11
1.4 Confidence Interval Estimation, 17
1.5 Other Methods of Estimation, 21
2 Multiple Logistic Regression
2.1 Introduction, 31
2.2 The Multiple Logistic Regression Model, 31
2.3 Fitting the Multiple Logistic Regression Model, 33
2.4 Testing for the Significance of the Model, 36
2.5 Confidence Interval Estimation, 40
2.6 Other Methods of Estimation, 43
3 Interpretation of the Fitted Logistic Regression Model
3.1 Introduction, 47
3.2 Dichotomous Independent Variable, 48
3.3 Polychotomous Independent Variable, 56
3.4 Continuous Independent Variable, 63
3.5 The Multivariable Model, 64
3.6 Interaction and Confounding, 70
3.7 Estimation of Odds Ratios in the Presence of Interaction, 74
4 Model-Building Strategies and Methods for Logistic Regression
4.1 Introduction, 91
4.2 Variable Selection, 92
4.5 Numerical Problems, 135
5 Assessing the Fit of the Model
5.1 Introduction, 143
5.2 Summary Measures of Goodness-of-Fit, 144
5.2.1 Pearson Chi-Square Statistic and Deviance, 145
5.2.2 The Hosmer-Lemeshow Tests, 147
5.5 Interpretation and Presentation of Results from a Fitted Logistic Regression Model, 188
8 Special Topics
8.1 The Multinomial Logistic Regression Model, 260
8.1.1 Introduction to the Model and Estimation of the Parameters, 260
8.1.2 Interpreting and Assessing the Significance of the Estimated Coefficients, 264
8.1.3 Model-Building Strategies for Multinomial Logistic Regression, 273
8.2 Ordinal Logistic Regression Models, 288
8.2.1 Introduction to the Models, Methods for Fitting and Interpretation of Model Parameters, 288
8.2.2 Model Building Strategies for Ordinal Logistic Regression Models,
305
The short course is a free service offered to conference registrants. No additional fees are required beyond conference registration.