Dear Statalisters,
I am implementing a 2SRI IV estimation following Terza (2017) (see here https://journals.sagepub.com/doi/abs...urnalCode=stja) but I am unable to figure out one thing. In the second step of the first stage, the IVs are not strongly correlated with the treatment intensity variable (in Terza's example, number of cigarettes smoked). Also the F-statistic is below the conventional cut off of 10. My question is, which F-statistic (first step or second step) should one report and why? Doesnt the second step F-stat also need to be above the cut offs?
I look forward to your insights.
Emmanuel
Related Posts with Understanding and interpretation of second step results of the first stage 2SRI in the two step estimation.
Subscribe to:
Post Comments (Atom)
0 Response to Understanding and interpretation of second step results of the first stage 2SRI in the two step estimation.
Post a Comment