1.1 Intro to Classical linear Regression Model

Question 1

What are we concerned with in a classical linear regression model?

Answer 1

Modelling a multivariate linear relationship between a dependent variable yt and a number of explanatory variables xti observed across time

Question 2

Why do we typically set x1 =1 for a all t?

Answer 2

So that the regression has a constant term

Question 3

What is ut?

Answer 3

An unobserved stochastic error term

Question 4

What are the assumptions of ut?

Answer 4

E(ut)=0
V(ut) = ó^2
C(ut,us)=0 when s=\t

Question 5

What are the two classical assumptions about the explanatory variables?

Answer 5

-fixed, not random
-linearly independent across k

Question 6

Assumptions of matrix form

Answer 6

• E(u)=0
• v(u) = E(u-E(u))(u-E(u))’) = E(uu’) = ó^2 I
  -x is nonstochastic with full column rank (k<T) so none of our x’s are equal to a linear combination of the other x’s