my question is about OLS standard erros in multicollinearity l understand that with heteroskedacity and postive serial collelation, the standard errors are underestimated, because in truth, they were bigger. so t-test ist bigger than it should be. with multicollineartity it says that OLS standard errors are usually inflated. can someone conecptionally explain me why this is the case? and anther thing popped into my head, i remember reading that in a linear regression, where there is only one independent variable, R2 = r2. is that true?
i can answer the second part…yes all you have to do is square the R to get r^2 in a simple linear…
square r to get R2