I have an observational study data set where some people have one measurement and some have 2 or 3. I don't care about the order of the measurements, nor longitudinal trend. I just want to get results without throwing out data - descriptive stats for categorical & continuous variables, and an unadjusted measure of association (binary*binary variable).

Can I use a weighted average for descriptive stats? For measure of association, I can't find recommendations for this situation. I do not want to analyze longitudinal trend. Can someone provide guidance?

  • 1
    $\begingroup$ How do you plan to average categorical variables? $\endgroup$ Dec 10, 2019 at 21:32
  • $\begingroup$ I of course would only average the continuous variables. But is this even an accepted method? $\endgroup$
    – L.S.
    Dec 10, 2019 at 21:53
  • $\begingroup$ 'accepted' probably depends on who you ask, what you are trying to do, etc. Where will these stats go? Who will read them? What are you trying to convey? $\endgroup$ Dec 11, 2019 at 20:18


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.