A new model of decision processing in instrumental learning tasks
Abstract
Learning and decision making are interactive processes, yet cognitive modelling of error-driven learning and decision making have largely evolved separately. Recently, evidence accumulation models (EAMs) of decision making and reinforcement learning (RL) models of error-driven learning have been combined into joint RL-EAMs that can in principle address these interactions. However, we show that the most commonly used combination, based on the diffusion decision model (DDM) for binary choice, consistently fails to capture crucial aspects of response times observed during reinforcement learning. We propose a new RL-EAM based on an advantage racing diffusion (ARD) framework for choices among two or more options that not only addresses this problem but captures stimulus difficulty, speed-accuracy trade-off, and stimulus-response-mapping reversal effects. The RL-ARD avoids fundamental limitations imposed by the DDM on addressing effects of absolute values of choices, as well as extensions beyond binary choice, and provides a computationally tractable basis for wider applications.
Data availability
All data analysed in this study are available from https://osf.io/ygrve/
-
A new model of decision processing in instrumental learning tasksOpen Science Framework, ygrve.
Article and author information
Author details
Funding
Nederlandse Organisatie voor Wetenschappelijk Onderzoek (016.vici.185.052)
- Birte U Forstmann
Australian Research Council (DP150100272)
- Andrew Heathcote
Australian Research Council (DP160101891)
- Andrew Heathcote
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Ethics
Human subjects: Informed consent was obtained in all experiments prior to the experiment onset. The local ethics board of the University of Amsterdam, Department of Psychology approved the study, with reference numbers 2018-BC-9620 (experiment 1), 2019-BC-10672 (experiment 2), 2019-BC-10250 (experiment 3), and 2020-BC-12788 (experiment 4).
Copyright
© 2021, Miletić et al.
This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 3,621
- views
-
- 434
- downloads
-
- 38
- citations
Views, downloads and citations are aggregated across all versions of this paper published by eLife.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Ecology
- Neuroscience
Prey must balance predator avoidance with feeding, a central dilemma in prey refuge theory. Additionally, prey must assess predatory imminence—how close threats are in space and time. Predatory imminence theory classifies defensive behaviors into three defense modes: pre-encounter, post-encounter, and circa-strike, corresponding to increasing levels of threat—–suspecting, detecting, and contacting a predator. Although predatory risk often varies in spatial distribution and imminence, how these factors intersect to influence defensive behaviors is poorly understood. Integrating these factors into a naturalistic environment enables comprehensive analysis of multiple defense modes in consistent conditions. Here, we combine prey refuge and predatory imminence theories to develop a model system of nematode defensive behaviors, with Caenorhabditis elegans as prey and Pristionchus pacificus as predator. In a foraging environment comprised of a food-rich, high-risk patch and a food-poor, low-risk refuge, C. elegans innately exhibits circa-strike behaviors. With experience, it learns post- and pre-encounter behaviors that proactively anticipate threats. These defense modes intensify with predator lethality, with only life-threatening predators capable of eliciting all three modes. SEB-3 receptors and NLP-49 peptides, key stress regulators, vary in their impact and interdependence across defense modes. Overall, our model system reveals fine-grained insights into how stress-related signaling regulates defensive behaviors.
-
- Neuroscience
Several recent theoretical accounts have posited that interoception, the perception of internal bodily signals, plays a vital role in early human development. Yet, empirical evidence of cardiac interoceptive sensitivity in infants to date has been mixed. Furthermore, existing evidence does not go beyond the perception of cardiac signals and focuses only on the age of 5–7 mo, limiting the generalizability of the results. Here, we used a modified version of the cardiac interoceptive sensitivity paradigm introduced by Maister et al., 2017 in 3-, 9-, and 18-mo-old infants using cross-sectional and longitudinal approaches. Going beyond, we introduce a novel experimental paradigm, namely the iBREATH, to investigate respiratory interoceptive sensitivity in infants. Overall, for cardiac interoceptive sensitivity (total n=135) we find rather stable evidence across ages with infants on average preferring stimuli presented synchronously to their heartbeat. For respiratory interoceptive sensitivity (total n=120) our results show a similar pattern in the first year of life, but not at 18 mo. We did not observe a strong relationship between cardiac and respiratory interoceptive sensitivity at 3 and 9 mo but found some evidence for a relationship at 18 mo. We validated our results using specification curve- and mega-analytic approaches. By examining early cardiac and respiratory interoceptive processing, we provide evidence that infants are sensitive to their interoceptive signals.