Question

Q. # 3

What are the consequences of the following model specification errors?

1. Omitting an important variable
2. Including an irrelevant variable
3. Measurement errors in the dependent variable
4. Measurement errors in the explanatory variables

Note: This is an econometrics question so please answer considering the subject.

Answer 1

Q3)

(a) Omitting an important variable will create biased estimates if the omitted variable is correlated with an independent variable. If not, all the independent variables will have unbiased estimates

(b) Including an irrelevant variable does not affect the biasedness of the estimates but does increase their variance and standard errors. Thus, increasing the probability of Type I error

(c)  Measurement errors in the dependent variable will not lead t biased estimates. It will just increase the unexplained part of the regression and decrease R2

(d) Measurement errors in the explanatory variables will lead to the estimate of that variable becoming biased with the direction depending on the direction of measurement error.