CONTRAST TESTS OF INTERACTION HYPOTHESES

This article argues for the use of contrasts to test a priori interact ion hypotheses in 2-way analysis of variance designs. It focuses on 3 underused types of interaction contrast tests: a ''matching'' pattern for cognate levels of row and column factors; the ''qualitative quadra tic,'' for monotonic profiles of means in the same direction but with opposed concavities; and a ''differential curvilinearity'' test for di fferences in the curvature of two profiles with the same direction of concavity. The circumstances that best capitalize on the potential adv antages of a priori contrast tests are indicated, and an effect size m easure for contrasts is presented. Investigators are urged to examine residuals after accounting for the variation attributable to the chose n contrast for patterns that may provide hints for more textured hypot heses in further research. If a posteriori contrasts are used, their e ffect sizes should be noted.