F contrasts in two factor balanced fixed effects ANOVAs
Abstract
.In balanced two-factor fixed-effects ANOVAs with the levels of at least one factor not ordered, it can be difficult to construct orthogonal contrasts for the unordered factors that result in meaningful comparisons. One approach is to make pairwise comparisons. Here we show how to do this for both main and interaction effects. The analysis of interaction contrasts allows a data analyst to gain greater insight into what levels of a categorical variable are driving an overall interaction effect. While these comparisons are not orthogonal, their interpretation aligns with an intuitive understanding of what interaction represents.
Fields of Study & Subjects
Statistical Methods and Applications
Statistical Methods and Bayesian Inference
Statistics Education and Methodologies
Publication details
- Journal / source
- Communication in Statistics- Theory and Methods
- Volume
- 55
- Issue
- 14
- Pages
- 4471–4484
- Publication date
- 20 January 2026
- Language
- EN