Consider our population linear regression model
Y=\beta _(0) \beta _(1)x U
with
E(U|
(
x)
|)
=0
. Which assumptions are needed for the OLS estimators to be unbiased? Select all that apply. The data comes from a random sample. The sample outcomes of
x
are not all the same number. The error term has a mean 0 and a constant variance. The conditional expectation is linear
E[Y|
[
x]
|]
=\beta _(0) \beta _(1)x
with finite second-order moments
E[Y^(2)]
and
E[x^(2)]
. The error term is a normally distributed variable with mean 0 and a constant variance.