Evaluating Theories Efficiently: The Nuts and Bolts of Contrast Analysis

Abstract

Contrast analysis is a way to evaluate theories efficiently and relatively easily, but despite its efficiency, informativeness, and simplicity, contrast analysis is rarely found in the research literature. We present and illustrate step-by-step outlines of contrast procedures for 4 basic kinds of analyses, and we highlight the ways in which they provide focused information that complements more conventional analytic techniques such as omnibus analysis of variance-based F tests and count-based chi-squares. In addition, we provide examples of contrast analysis using 2 widely used statistical packages. We believe that the social sciences will profit if researchers, teachers, students, reviewers, and editors have a deeper understanding and appreciation of the advantages, informativeness, and ease of contrast analysis.