I am running a few GSEM regressions at the firm-level, wanting to check whether x1, x2, x3, x4 have a positive effect on y, a binary variable.
x1, x2, x3, x4 and y are likely to be affected by the firm's sector (four sector categories, 2000 firms)
Therefore, I believe I have two options:
1. using sector as a control variable,
Code:
gsem ( i.sector x1 x2 x3 x4-> y, logit ), vce(robust)
Code:
gsem ( M[sector] x1 x2 x3 x4-> y, logit ) ( M[sector]-> x1 x2 x3 x4), vce(robust)
Can you please give me a piece of advice on which of the two options better applies to my case, and how to interpret the results in "option 2"?
Thank you very much!
0 Response to Firm-Sector Multilevel generalised structural equation modeling or sector control variable?
Post a Comment