Hierarchical Bayesian estimation and hypothesis testing for delay discounting tasks

Vincent, Ben

doi:10.3758/s13428-015-0672-2

Cited by 53 publications

(63 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The approach of relying on previous parameter inferences to determine priors for related models is becoming more frequent in cognitive modeling. Some recent examples include Gu et al (2016) in psychophysics, Gershman (2016) in reinforcement learning, Vincent (2016) in the context of temporal discounting, Wiehler et al (2015) for different clinical sub-populations in the context of gambling, and Donkin et al (2015) in the context of a visual working memory model. In an interesting application of the latter model, Kary et al (2015) used vague priors for key parameters, and used the data from the first half of their participants to derive the posterior distributions.…”

Section: Previous Data and Modelingmentioning

confidence: 99%

Determining informative priors for cognitive models

2017

View full text Add to dashboard Cite

The development of cognitive models involves the creative scientific formalization of assumptions, based on theory, observation, and other relevant information. In the Bayesian approach to implementing, testing, and using cognitive models, assumptions can influence both the likelihood function of the model, usually corresponding to assumptions about psychological processes, and the prior distribution over model parameters, usually corresponding to assumptions about the psychological variables that influence those processes. The specification of the prior is unique to the Bayesian context, but often raises concerns that lead to the use of vague or non-informative priors in cognitive modeling. Sometimes the concerns stem from philosophical objections, but more often practical difficulties with how priors should be determined are the stumbling block. We survey several sources of information that can help to specify priors for cognitive models, discuss some of the methods by which this information can be formalized in a prior distribution, and identify a number of benefits of including informative priors in cognitive modeling. Our discussion is based on three illustrative cognitive models, involving memory retention, categorization, and decision making.

show abstract

Section: Previous Data and Modelingmentioning

confidence: 99%

Determining informative priors for cognitive models

2017

View full text Add to dashboard Cite

show abstract

“…We The discount factors were derived from the fully endogenized rational inattention model (R2) after fitting the free parameters to data from [14]. also compared against standard quasi-hyperbolic discounting (QH; [21]), and several variations of hyperbolic discounting, including the basic functional form (H0), and generalized versions that incorporate magnitude-dependent discounting and choice stochasticity (H1-H3; [22]). The fully endogenized rational model decisively won the model comparison, with a protected exceedance probability (PXP) greater than 0.99.…”

Section: Applications To Prior Experimental Resultsmentioning

confidence: 99%

“…where α is the inverse temperature and ω is a lapse probability, capturing occasional random responses (see also [22]). All subsequent models share the same choice probability function.…”

Section: Model-fitting and Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

Rationally inattentive intertemporal choice

Gershman

Bhui

2019

Preprint

View full text Add to dashboard Cite

Discounting of future rewards is traditionally interpreted as evidence for an intrinsic preference in favor of sooner rewards. However, temporal discounting can also arise from internal uncertainty in value representations of future events, if one assumes that noisy mental simulations of the future are rationally combined with prior beliefs. Here, we further develop this idea by considering how simulation noise may be adaptively modulated by task demands, based on principles of rational inattention. We show how the optimal allocation of mental effort can give rise to well-known phenomena in intertemporal choice, such as the magnitude effect and the interaction between sign and magnitude effects. In a re-analysis of two prior data sets, and in a new experiment, we reveal several behavioral signatures of this novel theoretical account, tying choice stochasticity to the magnitude effect. We conclude that some aspects of temporal discounting may result from a cognitively plausible adaptive response to the costs of information processing.

show abstract

“…Forward Model Model Inversion are for a single class of modeling such as sequential sampling models (Matzke et al, 2013;Singmann et al, 2016;Vincent, 2015;Wabersich & Vandekerckhove, 2014;Wiecki, Sofer, & Frank, 2013). An exception is the Variational Bayesian Analysis (VBA) MATLAB toolbox (Daunizeau, Adam, & Rigoux, 2014), which allows users to fit and compare various models with variational Bayesian algorithms.…”

Section: Model Inversionmentioning

confidence: 99%

Revealing neuro-computational mechanisms of reinforcement learning and decision-making with the hBayesDM package

Ahn

Haines

Zhang

2016

Preprint

View full text Add to dashboard Cite

Word count (Abstract): 276Word count (Main text): 8,362Abstract Reinforcement learning and decision-making (RLDM) provide a quantitative framework, which allows us to specify psychiatric conditions with basic dimensions of neurocognitive functioning. RLDM offer a novel approach to assess and potentially diagnose psychiatric patients, and there is growing enthusiasm on RLDM and Computational Psychiatry among clinical researchers. Such a framework can also provide insights into the brain substrates of particular RLDM processes as exemplified by modelbased functional magnetic resonance imaging (fMRI) or electroencephalogram (EEG).However, many researchers often find the approach too technical and have difficulty adopting it for their research. Thus, there remains a critical need to develop a userfriendly tool for the wide dissemination of computational psychiatric methods. We introduce an R package called hBayesDM (hierarchical Bayesian modeling of Decision-Making tasks), which offers computational modeling on an array of RLDM tasks and social exchange games. The hBayesDM package offers state-of-the-art hierarchical Bayesian modeling, where both individual and group parameters (i.e., posterior distributions) are estimated simultaneously in a mutually constraining fashion. At the same time, it is extremely user-friendly: users can perform computational modeling, output visualization, and Bayesian model comparisons-each with a single line of coding.Users can also extract trial-by-trial latent variables (e.g., prediction errors) required for model-based fMRI/EEG. With the hBayesDM package, we anticipate that anyone with minimal knowledge of programming can take advantage of cutting-edge computational modeling approaches and investigate the underlying processes of and interactions between multiple decision-making (e.g., goal-directed, habitual, and Pavlovian) systems.In this way, it is our expectation that the hBayesDM package will contribute to the dissemination of advanced modeling approaches and enable a wide range of researchers to easily perform computational psychiatric research within their populations.Performing computational psychiatric research, however, especially computational modeling, is a challenging task for many clinical researchers or those with limited quantitative skills. Computational modeling involves multiple steps including designing/adopting laboratory tasks, building a theoretical framework of the task with a set of assumptions and mathematical equations, formulating multiple computational models based on the assumptions, estimating model parameters of each model, and quantitatively comparing the models of interest (e.g., Busemeyer & Diederich, 2010;Wiecki et al., 2015). It is a pressing issue how to train clinical researchers in mental health (e.g., psychiatrists and clinical psychologists) so that they can receive in-depth training across several related fields including cognitive science, advanced statistics, and neuroscience (Montague et al., 2012). For the dissemination of Computational Ps...

show abstract

Hierarchical Bayesian estimation and hypothesis testing for delay discounting tasks

Cited by 53 publications

References 42 publications

Determining informative priors for cognitive models

Determining informative priors for cognitive models

Rationally inattentive intertemporal choice

Revealing neuro-computational mechanisms of reinforcement learning and decision-making with the hBayesDM package

Contact Info

Product

Resources

About