Omitted Variable Bias
Also known as · OVB
Bias in when a relevant variable correlated with the included is left out: it folds into the error, violating A2, and the bias is . There is no bias if or if the omitted variable's true coefficient is zero.
When to use
Use the sign table (true effect times covariance) to say whether the estimate is biased up or down. Worked cases are the computer-price Hedonic Pricing example and the birthweight regression in Assignment 4, where the cigarette coefficient moved from with controls to without.
See Lecture 9 and Lecture 13 and Lecture 11.