Hi,

I'm trying to do a Hausman test for checking endogeneity. I have cross sectional data with a matched pair design.

I'm using clogit for regression.

In understand that hausman test can be used as a post estimation test to check for endogeneity in STATA.

I have used the following commands, however hausman test requires two models to be specified. One before allccats and one after allcats.


a) clogit FFVsCF l_DirectorsTotalCompensation RoE RoA MVBV AvgAge BoardSize IndDirPRCT FeDirPRCT ExecDirPRCT CDual MLEV ABig4 NoBM, group(MatchID)


b) estimates store allcats

c)

d) hausman . allcats, alleqs constant


My question is which model to specify in the second round (refer to point (c) above).

Can anyone please guide me.

Best regards,
Juhee