Estimability­Consistency and Regularity of Factorial Effects in Disconnected Designs

Abstract

This paper studies the estimability of an l factor design, d_x, which is derived from a p factor factorial design, d, by deleting the last (p -l) factors of d. It is shown that under certain regularity conditions a given contrast belonging to a factorial effect in d_x; is estimable in d_x; if and only if its corresponding contrast in d is estimable in d.