I understand that postive autocorrellation = Type 1 errors… What about negative autocorrellation? Is it Type 2 error?
Yes, exact opposite of positive serial correlation. Positive SC = low standard errors, high t-stats Negative SC = high std errors, low t-stats As far as t-tests go: Positive SC is like heteroskedasticity & Negative Serial Correlation is like multicollinearity
Perfect. Thanks for the help!