I want to analyze some data that asked the same set of survey items, but across dramatically different samples (all demographics and most other variables are statistically significantly different between the samples). The research questions were somewhat consistent and related to the specific survey items.

If I use covariates/confounding variables in my regression analysis with the merged data, do I still need to use 'source' (sample ID) in the models?


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.