Hope everyone is doing well. I am running my regression model using pooled 2sls. In the model, two potentially endogenous variables are included along with two instruments. I wanted to examine a relevance test using Shea's R^2 test and non identification test using Cragg-Donald approach. If I plan to use ivregress 2sls in STATA, do I have to make a STATA command like below?
ivregress 2sls DV (endo1 endo2 = IV1 IV2) exogenous variables control variables ?
or, would it be possible to separately examine Shea's R^2 like ivregress 2sls DV (endo1 = IV1) exogenous variables control variables & ivregress 2sls DV (endo2 = IV2) exogenous variables control variables ?
Further, to compute Cragg and Donald F statistics, it seems I am supposed to use ivreg2 to directly get the statistic. Is there any possible way to get through using ivregress 2sls (Pooled 2SLS)? Thank you in advance.
ivregress 2sls DV (endo1 endo2 = IV1 IV2) exogenous variables control variables ?
or, would it be possible to separately examine Shea's R^2 like ivregress 2sls DV (endo1 = IV1) exogenous variables control variables & ivregress 2sls DV (endo2 = IV2) exogenous variables control variables ?
Further, to compute Cragg and Donald F statistics, it seems I am supposed to use ivreg2 to directly get the statistic. Is there any possible way to get through using ivregress 2sls (Pooled 2SLS)? Thank you in advance.
