BIOSTATISTICAL ANALYSIS OF EPIDEMIOLOGIC DATA II: POISSON AND CONDITIONAL LOGISTIC REGRESSION ANALYSIS
June 22-26, 2015
8:30 a.m. – 12:00 p.m.
Course Number: 140.677.11
Presents Poisson regression techniques, the primary statistical tool for the evaluation and interpretation of data consisting of counts as well as rates and probabilities. Poisson regresssion methods are contrasted to the more traditional analyses of mortality and incidence rates. A case/control matched pairs design is examined. A discussion of methods for counted responses. Describes Poisson models for application to morbidity and mortality outcomes. Discusses matched pairs data and analysis. Presents conditional logistic models for matched data. Presents randomization tests and bootstrap estimation techniques.
Learning Objectives: Upon successful completion of this course, students will be able to
- Measure count data as occur in epidemiologic research
- Utilize Poisson regression methods for counted responses
- Perform appropriate analyses of matched pairs data
- Employ conditional logistic models for matched data
- Recognize the merits of randomization tests and bootstrap estimation techniques.
Grading Options: Letter Grade or Pass/Fail
Course Materials: Provided in class