Dummy Variable Trap

Including a dummy for every category as well as an intercept, which makes the regressors perfectly collinear so OLS cannot be computed.

When to use

Whenever you code a KK-category variable, include only K−1K-1 dummies and treat the omitted one as the Base Category. It is the standard example of Perfect Multicollinearity.

See Lecture 10.

Appears in