Replication (i.e., empirical studies that attempt to gauge the reproducibility of prior empirical findings) is a necessary, but not sufficient, element of legitimate scientific discovery (Popper 1959). Occasionally, reproducibility and replication are at the forefront of scientific inquiry (and popular culture—see e.g., Huizenga [1994] for a description of the role of replication in discrediting the infamous Fleischmann and Pons cold fusion experiments); however, many fields, including accounting, have largely neglected the necessary role that replication plays in forming a legitimate body of scientific knowledge.
Although recent events have increased the salience of possible intentional bias in published accounting research,1 published replications are likely more important because of the myriad unintentional factors that can influence the inferences generated by our work. For example, a theoretical analysis by Ioannidis (2005) suggests that common incentives and pressures present in the research production function (e.g., publication is necessary for promotion, tenure,...
