Chapter 5: SLR Assumptions Flashcards
1
Q
SLR Assumptions:
A
- Linear in parameters
- Random sampling
- There is sample variation in x
- Zero conditional mean
- -> Unbiasedness of Beta0hat and Beta1hat
- Homoskedasticity
2
Q
SLR1 - Linear in parameters:
A
y=Beta0^1+Beta1^1*x+u
Violation:
y=Beta0^2+Beta1^2*x=u
3
Q
SLR2 - Random sampling:
A
((xi;yi):i=1,2,…,n), so yi=Beta0+Beta1*xi=ui
Violation:
Cov(ui;uj) is unequal 0 for all i unequal j
4
Q
SRL3 - There is sample variation in x:
A
Summation of (xi-xbar)^2 >0
Violation: Summation of (xi-xbar)^2=0
5
Q
SRL4 - Zero conditional mean:
A
E(ui|xi)=0
Violation:
E(ui|xi) unequal 0.
6
Q
SRL5 - Homoskedasticity:
A
The conditional variance of u is a constant.
Var(ui|xi)=Sigma^2 (=const.)
-> Var(ui)=Sigma^2 (=const.)
-> Var(yi|xi)=Sigma^2 (=const.)
Violation:
Var(ui|xi)=Sigmai^2