Ten simple rules for the computational modeling of behavioral data

eLife, Journal Year: 2019, Volume and Issue: 8

Published: Nov. 26, 2019

Computational modeling of behavior has revolutionized psychology and neuroscience. By fitting models to experimental data we can probe the algorithms underlying behavior, find neural correlates computational variables better understand effects drugs, illness interventions. But with great power comes responsibility. Here, offer ten simple rules ensure that is used care yields meaningful insights. In particular, present a beginner-friendly, pragmatic details-oriented introduction on how relate data. What, exactly, model tell us about mind? To answer this, apply our simplest techniques most accessible beginning modelers illustrate them examples code available online. However, more advanced techniques. Our hope by following guidelines, researchers will avoid many pitfalls unleash their own

Language: Английский

---

Lack of Theory Building and Testing Impedes Progress in The Factor and Network Literature

Psychological Inquiry, Journal Year: 2020, Volume and Issue: 31(4), P. 271 - 288

Published: Oct. 1, 2020

The applied social science literature using factor and network models continues to grow rapidly. Most work reads like an exercise in model fitting, falls short of theory building testing three ways. First, statistical theoretical are conflated, leading invalid inferences such as the existence psychological constructs based on models, or recommendations for clinical interventions models. I demonstrate this inferential gap a simulation: excellent fit does little corroborate theory, regardless quality quantity data. Second, researchers fail explicate theories about constructs, but use implicit causal beliefs guide inferences. These latent have led problematic best practices. Third, explicated often weak theories: imprecise descriptions vulnerable hidden assumptions unknowns. Such do not offer precise predictions, it is unclear whether effects actually not. that these challenges common harmful, impede formation, failure, reform. Matching necessary bring data bear theories, renewed focus psychology formalizing offers way forward.

Language: Английский

---

Generalization guides human exploration in vast decision spaces

Nature Human Behaviour, Journal Year: 2018, Volume and Issue: 2(12), P. 915 - 924

Published: Nov. 6, 2018

Language: Английский

---

Choice history biases subsequent evidence accumulation

eLife, Journal Year: 2019, Volume and Issue: 8

Published: July 2, 2019

Perceptual choices depend not only on the current sensory input but also behavioral context, such as history of one's own choices. Yet, it remains unknown how signals shape dynamics later decision formation. In models formation, is commonly assumed that choice shifts starting point accumulation toward bound reflecting previous choice. We here present results challenge this idea. fit bounded-accumulation to human perceptual data, and estimated bias parameters depended observers' Across multiple task protocols modalities, individual biases in overt behavior were consistently explained by a history-dependent change evidence accumulation, rather than its point. Choice thus seem interpretation input, akin shifting endogenous attention (or away from) previously selected interpretation.

Language: Английский

---

Confirmation bias in human reinforcement learning: Evidence from counterfactual feedback processing

PLoS Computational Biology, Journal Year: 2017, Volume and Issue: 13(8), P. e1005684 - e1005684

Published: Aug. 11, 2017

Previous studies suggest that factual learning, is, learning from obtained outcomes, is biased, such participants preferentially take into account positive, as compared to negative, prediction errors. However, whether or not the error valence also affects counterfactual forgone unknown. To address this question, we analysed performance of two groups on reinforcement tasks using a computational model was adapted test if influences learning. We carried out experiments: in experiment, learned partial feedback (i.e., outcome chosen option only); complete information outcomes both and unchosen were displayed). In replicated previous findings valence-induced bias, whereby relative contrast, for found opposite bias: negative errors taken account, positive ones. When considering bias context it appears people tend confirms their current choice.

Language: Английский

---

Computational noise in reward-guided learning drives behavioral variability in volatile environments

Nature Neuroscience, Journal Year: 2019, Volume and Issue: 22(12), P. 2066 - 2077

Published: Oct. 28, 2019

Language: Английский

---

A computational reward learning account of social media engagement

Nature Communications, Journal Year: 2021, Volume and Issue: 12(1)

Published: Feb. 26, 2021

Social media has become a modern arena for human life, with billions of daily users worldwide. The intense popularity social is often attributed to psychological need rewards (likes), portraying the online world as Skinner Box human. Yet despite such portrayals, empirical evidence engagement reward-based behavior remains scant. Here, we apply computational approach directly test whether reward learning mechanisms contribute behavior. We analyze over one million posts from 4000 individuals on multiple platforms, using models based reinforcement theory. Our results consistently show that conforms qualitatively and quantitatively principles learning. Specifically, spaced their maximize average rate accrued rewards, in manner subject both effort cost posting opportunity inaction. Results further reveal meaningful individual difference profiles media. Finally, an experiment (n = 176), mimicking key aspects media, verifies causally influence posited by our account. Together, these findings support account offer new insights into this emergent mode

Language: Английский

---

The computational roots of positivity and confirmation biases in reinforcement learning

Trends in Cognitive Sciences, Journal Year: 2022, Volume and Issue: 26(7), P. 607 - 621

Published: May 31, 2022

Language: Английский

---

The interpretation of computational model parameters depends on the context

eLife, Journal Year: 2022, Volume and Issue: 11

Published: Nov. 4, 2022

Reinforcement Learning (RL) models have revolutionized the cognitive and brain sciences, promising to explain behavior from simple conditioning complex problem solving, shed light on developmental individual differences, anchor processes in specific mechanisms. However, RL literature increasingly reveals contradictory results, which might cast doubt these claims. We hypothesized that many contradictions arise two commonly-held assumptions about computational model parameters are actually often invalid: That generalize between contexts (e.g. tasks, models) they capture interpretable (i.e. unique, distinctive) neurocognitive processes. To test this, we asked 291 participants aged 8–30 years complete three learning tasks one experimental session, fitted each. found some (exploration / decision noise) showed significant generalization: followed similar trajectories, were reciprocally predictive tasks. Still, generalization was significantly below methodological ceiling. Furthermore, other (learning rates, forgetting) did not show evidence of generalization, sometimes even opposite trajectories. Interpretability low for all parameters. conclude systematic study context factors reward stochasticity; task volatility) will be necessary enhance generalizability interpretability models.

Language: Английский

---

Improving the Reliability of Cognitive Task Measures: A Narrative Review

Biological Psychiatry Cognitive Neuroscience and Neuroimaging, Journal Year: 2023, Volume and Issue: 8(8), P. 789 - 797

Published: Feb. 25, 2023

Language: Английский

---