Hi all, I have 3 questions: 1- Linear regression assumes that the errors are uncorrelated across observations or with the response variable, is it the same thing? If we assume otherwise (i.e. errors are correlated) then we will end up with biased estimators correct?
2- Linear regression assumes that the variance of the errors is constant around the true curve, whats the advantage in doing that? 3- Whats the difference between dependence and correlation? are independent random variables always uncorrelated? Thanks, albinali . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
