Applications of Moderation Analysis for Music Education Research

Abstract

Abstract The purpose of this article is to discuss the potential that moderation hypotheses hold for yielding insights into the complex processes of music teaching and learning. We begin with a discussion of the conceptual relevance of moderation hypotheses to music education research by drawing from a compelling example in the extant literature. This is followed by a brief review and critique of how moderation hypotheses and the concomitant statistical interaction effects have been presented and interpreted in recent research articles in the Journal of Research in Music Education. We offer recommendations to increase the robustness of the theoretical foundations from which moderation hypotheses are generated and the consistency of the ways statistical analyses are reported and interpreted. Next, we provide an illustration of basic statistical models that are useful for testing moderation hypotheses. We then provide an example of moderation analyses in linear regression and ANOVA frameworks as well as advice for reporting the results of these analyses. The examples we present employ real data consisting of 190 collegiate music students’ reports of self-regulated learning processes and motivational beliefs. A summary list of recommendations for readers to consider when conducting their own research is provided.