I'm new to working with matched case-control data. I have a study where case individuals are matched to controls in a 1:3 way. This means that for every case (diseased) I have three matched controls (unrelated individuals). My question is how can perform a t-test concerning a variable for such a cohort between the case and their matched controls. I know about paired t-tests, but how would it work in a case when it isn't a 1:1 match?


  • $\begingroup$ What exactly does 1:3 way mean? Do you have 3 control measurements and 1 non-control measurement for each case? Are these measurements dependent? Are the control measurements dependent? $\endgroup$ – user2974951 Feb 5 at 9:56
  • $\begingroup$ So the individuals in each case are different and independent? So where is the matching? $\endgroup$ – user2974951 Feb 6 at 8:46
  • $\begingroup$ They are matched on demographic variables, age and sex. $\endgroup$ – raygozag Feb 16 at 5:01
  • $\begingroup$ So matching in this case simply means that the individuals that are "matched" have roughly similar values when it comes to demographic variables (age, sex, ...)? But they are all independent, that is these are different individuals that share nothing with each other else except for being "similar" in demographics. $\endgroup$ – user2974951 Feb 16 at 9:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.