Collaborators: James LeBreton and  Jan Dul

We compare Necessary Condition Analysis (NCA) with  Ordinary Least Squares regression analysis (OLS). We discuss the underlying assumptions of NCA and OLS, and show that NCA is insensitive to a common problem of OLS: omitted variable bias, implying that models for testing single necessary conditions can be simple and do not require inclusion of control variables. We conclude that NCA and OLS are complementary logical approaches. NCA can explain why outcomes, that are possible according to OLS, may not be produced because the right level of the necessary condition is not in place. OLS can explain how outcomes can be produced on average when all necessary conditions are in place. We believe that by applying both logics and methods organizational phenomena can be better understood.