Identifying Necessary Conditions in Accounting Research : Logical and Methodological Issues

The study discusses the challenges and opportunities of establishing a logic and method to identify necessary relationships in accounting research. Recent methodological advancements indicate that necessary conditions have not been precisely addressed, neither in hypotheses development, nor in data analysis, implying the presence of ambiguous communication and misleading identification in causal reasoning. A review of accounting literature confirms this lack of attention and guidance in accounting, in particular when formulating, testing, and interpreting necessity-related hypotheses. Therefore, the paper has a threefold purpose. First, we present the logic of necessary conditions to highlight the fundamental difference between sufficiency and necessity. Second, we introduce necessary condition analysis (NCA) to facilitate an appropriate and straightforward testing of necessary conditions. Finally, we illustrate NCA by applying it to a case of designing costing systems and indicate the complementary insights it generated in comparison to the results of a regression analysis