Why Self-Determination Theory Needs Computational Modelling: The Case of Competence and Optimal Challenge

Computational modelling is a powerful tool to specify psychological theories, render them severely testable, and implement them into digital applications. Yet it has seen little uptake in Self-Determination Theory (SDT). Here, we demonstrate that SDT’s verbal definitions of competence and optimal challenge are underspecified in theoretically and practically relevant ways. Conceptual analyses of key texts identifies four verbal facets of competence definitions that need not co-occur and are inconsistently reflected in common self-report measures and operationalisations. Optimal challenge is insufficiently specified to be severely tested or implemented in practice, and entails a logical and empirical incoherence. Finally, SDT lacks a cognitive model of competence. We outline how computational modelling, inspired by the AI field of computational intrinsic motivation,can help resolve resulting issues.