Abstract
The available procedures for the computation and use of expected mean squares (EMS) in analysis of variance are reviewed. The discussion is tutorial in nature and includes algorithms for EMS computation, the selection of appropriate error terms in constructing F tests for the significance of factor effects, the estimation of variance components, synthesized mean squares and associated degrees of freedom and approximate F tests. Emphasis is on practical applications. Examples of both balanced and unbalanced designs are presented.
Additional information
Notes on contributors
Ronald D. Snee
Dr. Snee, a Member of ASQC, is a Senior Statistician in the Engineering Service Division of E.I. du Pont de Nemours & Company in Wilmington, Delaware. This paper was presented at the Sixteenth Annual Fall Technical Conference of the Chemical Division of the American Society for Quality Control, co-sponsored by the Section on Physical and Engineering Sciences of the American Statistical Association, October 19 and 20, 1972, in Knoxville, Tennessee. He received the 1972 Shewell Prize for this paper.