Skip to main content

OLS assumptions

OLS assumptions are the conditions data and errors must satisfy for ordinary least squares to deliver unbiased, efficient estimates.

Also known asGauss-Markov assumptions

ByHoang TruongUpdated

FrameworkOrdinary least squares (OLS)

See it move

Loading infographic...

The infographic is a formula card stating the five Gauss-Markov conditions under which ordinary least squares is the Best Linear Unbiased Estimator (BLUE): linearity, random sampling, no perfect collinearity, zero conditional mean E(u|X) = 0, and constant variance Var(u|X) = σ². The most dangerous violation is E(u|X) ≠ 0, which biases the slope estimates; violating Var(u|X) = σ² — heteroskedasticity — distorts standard errors only, leaving the slopes unbiased.

Where it fits
SubjectData Analysis & StatisticsCoreTopicRegression Diagnostics & ProblemsCore
OLS assumptions (Gauss-Markov conditions)