The use of limited dependent variable techniques in strategy research: assessment and critique

Strategy researchers are increasingly turning their attention from examining the implications of strategic choices on firm performance to examining the factors that determine strategic choice at the firm level. This shift of research orientation has meant that researchers are increasingly faced with a limited dependent variable (LDV) that takes a limited number of usually discrete values. In such cases researchers typically use discrete LDV methods such as Logit or Probit and, in fact, the use of such methods has increased significantly in recent years. Despite their growing popularity, there appear to be widespread problems in the application, reporting, and interpretation of LDV methods and their results within the literature. We examined the use of LDV methods in 50 papers published since 2002 in two top-tier journals that are primary outlets for empirical strategy research (Strategic Management Journal and Academy of Management Journal). One particularly troublesome issue is the finding that researchers fail to correctly analyze moderating hypotheses, a situation that likely stems from a lack of familiarity with the nonlinear nature of LDV models. Based on our review of the literature, this paper provides an assessment of the use of the most common LDV methods, highlights problems and inconsistencies regarding their use and interpretation, and provides guidelines and suggestions for researchers seeking to use LDV statistical techniques.