Hausmann Test: the rank of the differenced variance matrix (2) does not equal the number of coefficients being tested (3)

[Jasmin Music]

Hi,

I have run a Hausmann test, to check whether I should use fixed-effects or random-effects method. Normally, according to the p-Value of 0,000 being lower than the significance level, I should use fixed-effects method.
But I do not understand the NOTES from Stata, can someone tell me, what does it mean? And especially if I can still go with the fixed-effects method?

Best regards

Jasmin

[Carlo Lazzaro]

Jasmin:
you may have singletons somewhere in your variables.

[Jasmin Music]

- How can I discover if I have singletons variables? Because a mistyped variable is not the case in my model, I checked it now.

"Singleton variables is a warning, not an error", can I still go for a fixed-effects method, however?

[Carlo Lazzaro]

Jasmin:
what you report is basically Ok (see also https://www.statalist.org/forums/for...a-hausman-test)
The issue with -hausman- is that it works well asymptotically.
You may want to give the community-contributed module -xtoverid- a shot.