Dear Statalist,

I am grateful for any help on the following question:

I have several datasets, each with the same response variable but with different regressors. Unfortunately, I do not know the correlation between the regressors from the different datasets, but have reason to believe that there is a correlation.

Is there a reasonable way to combine the response variable and all the regressors into a single regression or other coherent analysis without knowing the correlation matrix between the regressors?

Any idea is appreciated. Many thanks in advance.