ABSTRACT
Moderation and mediation provide a means for accounting behavioural researchers to gain additional insight and understanding when investigating increasingly complex accounting relationships. The common goal of both moderation and mediation is to understand the effects of an additional variable on the relationship between the independent and dependent variable. However, the underlying theory, assumptions, and statistical techniques used to investigate mediation and moderation are very different. An understanding of these differences is critical for conducting high-quality and relevant accounting behavioural research. This chapter provides a general high-level introduction to moderation, mediation, and moderated mediation techniques. We focus on issues relevant to future accounting behavioural researchers who, while familiar with statistics, may not be knowledgeable about the challenges of conducting research using these techniques. Our discussion is divided into three broad sections: moderation, mediation, and moderated mediation. For each, we begin by defining the key terms and concepts. Next, we introduce and briefly discuss the different statistical techniques and approaches for testing and evaluation, highlighting both strengths and weaknesses. Because our presentation is broad, references are provided for researchers wishing to gain more in-depth information. Throughout we also call attention to accounting behavioural research that use the approaches discussed. Finally, we offer some brief concluding thoughts.
