1. Consider the model Yi = Xi + ui; i = 1; :::;N where ui satis es the usual hypotheses. (a) Derive by OLS the estimator of the parameter : (b) Show that the estimador obtained in (a) is unbiased. (Hint: Compute the expected value of the estimator.)
my resolution is on the anex file, also please tell me the right way to do it the b) correctly.
thanks and sorry for my english.
October 16th 2012, 08:23 PM
Re: Ols help urgent
One thing is that you can't cancel Sigma (x_i*y_i) / Sigma(x_i^2) the way you did it: you have to leave it like that. Cancelling terms like you do with normal fractions is not the same when you have summations.
You can only cancel things that are common terms outside any summation or multiplication and not inside (and if it's inside you have to show it's independent of the terms in the summation or multiplication and can be brought out).