Re: Incosistency of OLS estimators

On 23 фев, 22:46, borislav.vange...@xxxxxxxxxxxxxx wrote:
Hi, Borislav.

I would appreciate if you could tell me how u is related to your
model. If cov(e,x) does not equal zero than your estimates will be
biased. If you still have some problem regarding this model I could
help you.

Kirill.



I have a really stupid question. We are arguing with a friend if an
estimator is incosistent.

The model is the simplest :
y = b0 + b1*x + e ( e is an error term)

assuming the E[u|x] =a a != 0, the problem is if the model is
necessarily inconsistent or is this only the case when cov (u,x) !=0

Thanks in advance
Borislav


.