The Beneficial Learning Effects of Combining a Hypothesis-Testing Mindset with a Causal Model

Abstract

Firms often use causal models to align decision-making with strategic objectives. However, firms often operate in changing environments such that an accurate causal model can become inaccurate. Prior research has not examined the consequences a change in the accuracy of causal models may have for managerial learning. Using an experiment, we predict and find that providing an accurate causal model positively affects managerial learning, and this positive effect is not reduced by encouraging a hypothesis-testing mindset (HTM). However, when the model subsequently becomes inaccurate, we predict and observe that providing a causal model alone negatively affects managerial learning, although this effect is partially mitigated by additionally encouraging a HTM. Our results can inform designers of control systems about the potential implications of providing a causal model when its accuracy changes over time and demonstrate how simple encouragement of a HTM moderates the effects of providing a causal model.