Theory Testing and Process Evidence in Accounting Experiments

This paper outlines strengths and limitations of three approaches to providing process evidence in experimental accounting research: moderation, mediation, and multiple experiments and methods. We argue that the research question determines the importance of process evidence in experimental accounting research. When process is central to the research question, process evidence is essential. In other cases, process can play a secondary role by strengthening the internal and external validity of inferences or suggesting possible approaches to improving accounting decisions. In either case, we recommend a design-based approach and make recommendations for collecting process evidence in ways that minimize the threats to validity that are common to the method employed