Mitigating the Negative Effects of Causal Models: Encouraging a Hypothesis Testing Mindset and Managers’ Quantitative Knowledge

Abstract

The causal model of a firm may change as its competitive environment changes. We use an experiment to examine how providing an initially accurate causal model that subsequently becomes inaccurate affects managerial learning after the accuracy of the model changed. We predict and find a negative effect of providing a causal model on learning. However, we predict and find a positive interaction effect on learning from encouraging a hypothesis-testing mindset (versus not doing so) and providing a causal model, such that encouraging a hypothesis-testing mindset mitigates the negative effect of providing a causal model. Similarly, we predict and find a positive interaction effect on learning from having more quantitative knowledge and providing a causal model, with more quantitative knowledge mitigating some of the negative effect of providing a causal model. Lastly, we find that encouraging a hypothesis testing mindset and having more quantitative knowledge are substitutes in terms of mitigating the negative effect of a causal model, in that there is significant positive interaction effect of quantitative knowledge and causal model only for participants who have not been encouraged to adopt a hypothesis-testing mindset whereas there is no significant interaction effect for participants who have been encouraged with a hypothesis-testing mindset. Our results help companies understand the potential negative implications of providing a causal model which potentially changes over time, and the possible mechanisms to mitigate those negative effects.