Reminders/Things due this week
- Practice Exercise 1.5 Due Tuesday
- Practice Exercise 1.6 Due Thursday (It is okay to submit by Sunday)
- Don’t forget about Problem Set 1.1 (Due Feb 28)!! Email Yiqing([email protected]) if you have any questions about the problem set.
Learning objectives
- Refresh knowledge of odds, odds ratio (OR), and log transformation
- Understand how logistic regression works
- What is a logit function?
- What is a generic equation of logistic regression models?
- When should we use logistic regression instead of linear regression?
- How about assumptions in logistic regression?
- Understand the basic syntax of PROC GENMOD
- Be able to implement a logistic regression model in SAS for a given scenario
- Be able to interpret the SAS outputs of a logistic regression model
- Interpret untransformed estimates (raw estimates)
- Perform transformation and interpret transformed estimates
- Be able to perform model comparions/selection
- Use Likelihood Ratio Test to compare nested models (recap Maximum Likelihood Estimation)
- Use information criteria (AIC/BIC) to compare non-nested models
Readings & videos
- Reading:
- Allison ch 2&3 (available on Sakai)
- (Optional) GHV ch13
- Below are some YouTube videos about logs(logarithms), natural log, odds, and log(odds) that I have found very helpful. If you feel unsure about these concepts, I think you would benefit from watching these videos before Tuesday’s class (or even before you start to read the assigned chapters).
Tuesday Session
Lecture: Introduction to logistic regression