A Critical Note on the Application of Analysis of Variance Techniques in Ergonomics Research

Abstract

A survey was made of the statistical techniques used in 108 papers appearing in Ergonomics between 1969 and 1971. It was found that analysis of variance techniques had been misapplied in many cases, sometimes with potentially serious consequences. The major criticisms concern the use of inappropriate error terms, the inadequate consideration of some of the assumptions involved in repeated measures and other designs, and the omission of a posteriori tests.