chapter  30
14 Pages

Specific Usages of Regression Models

So the original idea is to look from the case-control status onto the covariate values, which is exactly the opposite what we do in regression modelling. Nevertheless, it can be shown that it is possible, allowed, and useful to analyse data from casecontrol studies using logistic regression with the case-control status as outcome, that is, with Y = 1 for the cases and Y = 0 for the controls.