Although the testing of interaction effects via moderated regression is commonplace in management research, many misunderstandings and gaps in knowledge persist, and knowledge of recent methodological developments is slow to catch on. This workshop is aimed at researchers, reviewers, and PhD students alike, with a basic understanding of moderation, but who would like to advance their knowledge beyond this point. It will use a series of examples to refresh the basic principles of testing two-way and three-way interactions, focusing on how (and when) these should be interpreted based on various post-hoc probing and plotting techniques. In particular, it will cover different approaches to describing interaction effects, offering guidance as to when and how post-hoc probing techniques such as simple slope testing should (or shouldn’t) be used, and will include recommendations based on recent publications. It will also cover the testing, plotting and probing of some more complex types of interactions, in particular those involving curvilinear effects, and those where the underlying regression model is nonlinear (e.g. logistic or Poisson regression). Participants will have the opportunity to request topics and ask specific questions on moderation by sending an email to j.f.dawson@sheffield.ac.uk at least a week before the workshop. The final section of the workshop will discuss these questions, and others generated in the workshop itself. Participants will be provided with resources (including data files and syntax for R, SPSS, and Stata and screencasts) to replicate the methods demonstrated in the workshop.
