0

I am doing some regression analysis on a stratified sample where the stratification variable is the dependent variable. Do I need to adjust the regression or the results to account for the disproportionate subpopulations at all?

In other words, I have a sample containing groups A and B where B is very rare in the population but relatively common in the sample. Are the results of a regression on "Group" valid as they are, or do they need to be adjusted to reduce the effect of B members?

aabceh
  • 1
  • Do these links answer your question? https://stats.stackexchange.com/questions/69561/case-control-study-and-logistic-regression and https://stats.stackexchange.com/questions/67903/does-down-sampling-change-logistic-regression-coefficients/ – jdcrossval Oct 08 '23 at 18:34
  • This sounds like a case-control study: https://en.wikipedia.org/wiki/Case%E2%80%93control_study . That link is a place to start to find out what is and is not possible with that design. – Maarten Buis Aug 01 '23 at 08:11

0 Answers0