Unbiasedness

The mean of the estimator's sampling distribution equals the true parameter, E[θ^]=θE[\hat\theta]=\theta. For OLS, assumptions A1 and A2 deliver it.

When to use

Use it when asked which assumptions make OLS unbiased, or to distinguish bias (centring) from variance (spread).

See Lec 4.

Appears in

No references yet.