Skip to content
Q147Section III (Statistics)147 of 180

Estimability in the general linear model — the row-space characterisation

Consider the linear model Y‾n−1=Xθ‾p−1+ξ\underline Y_{n-1}=X\underline \theta_{p-1}+\xi with E(ξ)=0‾E(\xi)=\underline0, Cov⁡(ξ)=σ2In\operatorname{Cov}(\xi)=\sigma^2I_n. Then which of the following statements is always true?

Options

Pick an option to check your answer.