Using large-scale experiments and machine learning to discover theories of human decision-making

Joshua C Peterson,David D Bourgin,Mayank Agrawal,Daniel Reichman,Thomas L Griffiths,Joshua C. Peterson,David D. Bourgin,Thomas L. Griffiths
DOI: https://doi.org/10.1126/science.abe2629
IF: 56.9
2021-06-11
Science
Abstract:Discovering better theories Theories of human decision-making have proliferated in recent years. However, these theories are often difficult to distinguish from each other and offer limited improvement in accounting for patterns in decision-making over earlier theories. Peterson et al. leverage machine learning to evaluate classical decision theories, increase their predictive power, and generate new theories of decision-making (see the Perspective by Bhatia and He). This method has implications for theory generation in other domains. Science , abe2629, this issue p. 1209 ; see also abi7668, p. 1150
multidisciplinary sciences
What problem does this paper attempt to address?