if I have a unbalanced panel survey data of four years, I wondered if I still needed to apply panel weights when run "reg", as pool ols regressions on it? Or should I apply cross-sectional weights for "reg" command? There are repeated time values within panel, but the same personal IDs for each year. There might be repeated values for personal IDs. Thank you.


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.