Proceedings of the Annual Meeting of the Cognitive Science Society

Mindset theory proposes that believing intelligence is improvable fosters academic achievement. Despite its influence, the theory lacks a mechanistic foundation. Here, we examine mindset-driven behaviors from the perspective of the computational problem the mind is solving—optimizing cumulative reward under uncertainty about skill malleability. We formalize this problem as a Markov decision process, where agents balance cultivating for future gains against harvesting immediate rewards. Growth- and fixed-mindsets are represented as the agent's priors over skill malleability. Through simulation, we demonstrate that mindset-driven behaviors, such as persistence and challenge seeking, arise as rational decisions under different prior beliefs, with optimistic priors promoting sustained engagement, belief updating, and higher rewards in favorable environments. Crucially, mindset effects vanish when environmental structures disincentivize long-term investment, leading agents to converge on a harvest-only strategy. Our model offers a unified account of the heterogeneous findings of mindset interventions, highlighting the importance of supportive environments for mindset effects to manifest.