Free Version
Difficult

Omitted Variable Bias Direction 1

EMETRC-0KKLMC

Consider the general population model in which $y$ depends upon two $x$ variables:

$$y = \beta_{0} + \beta_{1} x_{1} + \beta_{2} x_{2} + u$$

If we instead estimate the simple regression:

$$\tilde{y} = \tilde{\beta_{0}} + \tilde{\beta_{1}} x_{1}$$

...which factors solely determine the direction of bias in the estimated parameter $\tilde{\beta_{1}}$?

Select ALL that apply.

A

The sign of the true $\beta_{2}$

B

The sign of the true $\beta_{1}$

C

The magnitude of $\hat\alpha_{1}$

D

The sign of the correlation between $x_{1}$ and $x_{2}$

E

The sign of the correlation between $u$ and $x_{2}$