HOME
BIOST/EPI 536, Autumn 2005
Categorical Data Analysis in Epidemiology
Class Notes
Thursday, September 29:
Introduction
Jama commentary
Lancet article
Mattress table
Tuesday, October 4:
Review: disease occurrence and association; cohort and case-control sampling; confounding and matching.
Thursday, October 6:
Reparameterization. Simplest Logistic models and coefficient interpretation. First Stata output.
Review sheet on logarithms
Tuesday, October 11:
Logistic model; maximum likelihood estimation; confidence intervals and hypothesis tests; Stata.
Tuesday October 18:
Modelling exposure, part 1
Thursday October 20:
Modelling exposure, part 2
Tuesday October 25:
Modelling exposure, part 3
Article about splines
Thursday October 27:
Model selection for modelling exposure
Modelling multiplicative interaction
Tuesday November 1:
Model selection for interaction
Additive models of interaction
Article on effect modification
Tuesday November 8:
R
egression adjustment for grouped and continuous confounders.
Regression model selection for confounders.
Effects of omitting covariates
Thursday November 10:
Marginal and Conditional Models
Greenland, Robins Pearl paper
Tuesday November 15:
Marginal Structural Models
Thursday November 17:
Causal Diagrams and Confounding
Greenland et al. article
Advanced Coding Questions
McKnight et al. paper
Tuesday November 22
Regression Diagnostics
Conditional Logistic Regression
(Notes on these topics were handed out in class on 11/22. Extra copies are available in the Biostat file cabinet.)
Thursday December 1
Exposure Misclassification: Ken Rice
Updated Ken Rice notes
T
uesday December 6
Analysis of two stage designs: Norm Breslow (pp 1-24 will be covered).
Breslow and Chatterjee paper
NWTS Stata data set
NWTexb Stata data set
NWTexc Stata data set
Thursday December 8
Short summary
Conditional Logistic Regression HW
BACK TO TOP
Last Updated:
12/08/05
Contact the instructor at:
bmck@u.washington.edu