From: minibela on
Hello,
I’m trying to test my dataset for heteroskedasticity with TESTHET, a function proposed by Oleg Komarov.
I have three questions, maybe someone could answer me:
- if the OLS model has no intercept, the auxiliary model may have one?
- what is the minimal number of observations to apply White’s method?
- if I've got pval = 0.24, that means I have to accept the null hypothesis at the confidence level 0.05, right?
Thanks