Pigeon, Joseph G.; McAllister, Paul R. (1989, ASQC and the American Statistical Association) Villanova University, Villanova, PA; Rohm and Haas Company, Philadelphia, PA
Orthogonal main-effect plans (OMEP's) are resolution III factorial designs that permit the uncorrelated estimation of all main effects under the assumption of no interactions among the factors. In these OMEP situations, it is sometimes possible to include partial replication of the design allowing for an unbiased estimate of the error variance, without sacrificing orthogonality of main effects. An assessment of the no-interactions assumption may then be made by comparing the available sample of confounded interactions to the pure error. Furthermore, in some situations these goals may be met without exceeding the minimum number of runs for an OMEP. Two examples of these partially replicated orthogonal main-effect plans are given.
Fractional factorial design,Interactions,Main effects,Replication