A low-dimensional approximation of optimal confidence
PLOS Computational Biology · 2024
DOI: 10.1371/journal.pcbi.1012273 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS Computational Biology, 2024. Verfasst von Pierre Le Denmat und 2 weiteren Autor:innen. Beteiligte Einrichtung: Brain and Cognition, KU Leuven, Leuven, Belgium. Hauptthemen laut MEDLINE: Bayes Theorem, Decision Making. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 72 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: KU Leuven, Francqui Foundation, Research Foundation Flanders, Belgium.
Abstract (Originalsprache)
Human decision making is accompanied by a sense of confidence. According to Bayesian decision theory, confidence reflects the learned probability of making a correct response, given available data (e.g., accumulated stimulus evidence and response time). Although optimal, independently learning these probabilities for all possible data combinations is computationally intractable. Here, we describe a novel model of confidence implementing a low-dimensional approximation of this optimal yet intractable solution. This model allows efficient estimation of confidence, while at the same time accounting for idiosyncrasies, different kinds of biases and deviation from the optimal probability correct. Our model dissociates confidence biases resulting from the estimate of the reliability of evidence by individuals (captured by parameter α), from confidence biases resulting from general stimulus independent under and overconfidence (captured by parameter β). We provide empirical evidence that this model accurately fits both choice data (accuracy, response time) and trial-by-trial confidence ratings simultaneously. Finally, we test and empirically validate two novel predictions of the model, namely that 1) changes in confidence can be independent of performance and 2) selectively manipulating each parameter of our model leads to distinct patterns of confidence judgments. As a tractable and flexible account of the computation of confidence, our model offers a clear framework to interpret and further resolve different forms of confidence biases.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Brain and Cognition, KU Leuven, Leuven, Belgium.
- Department of Experimental Psychology, Ghent University, Ghent Belgium.
ORCID-Profile: Pierre Le Denmat
Förderung
- KU Leuven
- Francqui Foundation
- Research Foundation Flanders, Belgium
- Fondation Francqui - Stichting
- Fonds Wetenschappelijk Onderzoek
Zitieren
APA
Le Denmat, P., Verguts, T., & Desender, K. (2024). A low-dimensional approximation of optimal confidence. PLOS Computational Biology. https://doi.org/10.1371/journal.pcbi.1012273
Vancouver
Le Denmat P, Verguts T, Desender K. A low-dimensional approximation of optimal confidence. PLOS Computational Biology. 2024. doi:10.1371/journal.pcbi.1012273
BibTeX
@article{LeDenmat2024,
author = {Le Denmat, Pierre and Verguts, Tom and Desender, Kobe},
title = {A low-dimensional approximation of optimal confidence},
journal = {PLOS Computational Biology},
year = {2024},
doi = {10.1371/journal.pcbi.1012273}
}
EUVIMED