Abstract
A well-designed experiment interrupted before its intended completion may prove inefficient and leave important questions unanswered. This article addresses the construction of “interruptible” designs for those cases in which a first-order model is assumed and factors at only two levels are considered. The maximum | X′X | criterion is used in developing interruptible designs for two different design strategies that attempt to limit the information lost if an experiment is prematurely ended. The resulting designs for four to nine factors are tabulated and selection of an interruptible design is discussed.