Can someone verify the degrees of freedom when looking up critical values for all the tests?
n = observations k = slopes
T-test
df = n - (k+1) [note: is 1 due to the intercept?]
F-test
numerator df = k denominator df = n - (k+1) [note: is 1 due to the intercept?]
Chi squared (heteroskedasticity)
df = k
Durbin-Watson (serial correlation)
k is number of slopes? n is observations?
I’m not missing anything else, am I?