From theories to models to applications in decision science
Year: 2026
Type: preprint
Abstract: Behavioural decision science aspires to both explain human behaviour and improve real‑world outcomes. Yet despite a wealth of empirical support for decision theories, its impact has often fallen short. We argue that this shortfall stems partly from underspecified theories and overly task‑specific models that limit generalisability. To do so, we revisit the conceptual foundations of modelling in decisionmaking research and propose a framework clarifying the relationship between theories and models. We contrast theory‑driven with emerging data‑driven approaches, such as neural networks, which prioritise predictive accuracy but offer limited explanatory insight and deepen ethical concerns raised by model‑based interventions. Rather than viewing theory‑ and data‑driven approaches as mutually exclusive, we advocate for synergistic integration: data‑driven tools can support theory refinement and the piloting of applications, while the presence of explicit theories preserves interpretability. Greater epistemological clarity can foster progress in both understanding behaviour and designing interventions to change it.
Source: PsyArXiv (OSF Preprints)
Authors: Stefano Palminteri, Maël Lebreton, Bahador Bahrami, Sophie Bavard, Bastien Blain, Anne Collins, Giorgio Coricelli, Benedetto De Martino, Claire Gillan, Catherine A. Hartley, Pierre O. Jacquet, Anna B. Konova, Junseok K Lee, Amy Orben, eva pool, Chiara Varazzani, Valentin Wyart, Ophelia Deroy, Daphna Shohamy
Language: English
Subfield: General Decision Sciences
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7234796429
