I have a queston on how to interpret a Hausman-test.
- Dependent variable (y) is suffering from an accident or injury on a scale 0-10 (pl05)
- Independent variable (x) is work stress (yes/no 1/0) (pw604)
- Panel data for years 2010 and 2016
FE coefficient=0.084 p-value=0.00 (same as Pooled OLS)
RE coefficient=0.620 p-value= 0.875
Hausman-test results (also in attachment):
Chi2=4.83
Prob>Chi2= 0.1847
In this case, Ho would be rejected, which would mean both models are the same, but they are clearly different based on the estimated coefficients and p-values.
My questions are:
- Why does the test reject Ho when RE and FE seem clearly different?
- Does rejecting the null Hypothesis mean that RE is preferred over FE?
0 Response to Hausman-test interpretation for RE vs FE
Post a Comment