Dear Statalist,
Is it okay to go ahead with ESR (Endogenous Switching Regression) when untreated observations are less than Treated observations? Or will such a sample size affect the ESR results particularly the MTE (marginal treatment effect) on TU or avg treated on Untreated? Data may be assumed as one year survey data.
Please try to give your suggestions.
Thank you.
Is it okay to go ahead with ESR (Endogenous Switching Regression) when untreated observations are less than Treated observations? Or will such a sample size affect the ESR results particularly the MTE (marginal treatment effect) on TU or avg treated on Untreated? Data may be assumed as one year survey data.
Please try to give your suggestions.
Thank you.
