Go to training host page. You will leave this site.
This issue of FOCUS offered by the UNC Gillings School of Global Public Health will explore logistic regression and accounting for matched data. Logistic regression is an efficient way to control for many potential confounders at one time. Matching, if done correctly in the study design stage of the investigation, reduces confounding before the analysis even begins.
- Define and describe confounders
- Discuss three ways to reduce the effects of confounders on your data: restriction, stratification, and logistic regression
- Learn how to identify and interpret effect measure modifiers
- Learn how to calculate matched odds ratios for case-control studies