Lecture notes 7 irrelevant variables lecture notes Flashcards
1
Q
What is the impact of including irrelevant variables in a regression?
A
Overall, it does not hurt as long as CLRM are satified.
However, it can reduce precision of OLS estimates.
As there is a trade off between sample size and parameters.H
2
Q
How does variance compare of b1 and c1 if c1 has irrelevant variables
A
Both variance estimates are unbiased but c1 with irrelevant is more spread out.
3
Q
What is a specific type of irrelevant variables?
A
Bad controls
Sometimes adding in an extra control can harm the causal interpretation as you are comparing apples to organges.
As this will change the groups and add noise to the regression.