Proof of Unbiasedness for Ordinary Least Squares (OLS) Regression Coefficients
Students often confuse unbiasedness with efficiency. Unbiasedness only means the 'average' result is correct, not that the result is 'close' to the truth in a single trial. An estimator can be unbiased but still have such massive variance that individual estimates are practically useless.
Institutional Reference: Applied Statistics
View Full Proof →