Hierarchical Modelling and Affective Dimensions of Social Nature Improve Explanation of Human Econom
Poster No:
1082
Submission Type:
Abstract Submission
Authors:
Jaewon Kim1, Su Hyun Bong1, Dayoung Yoon1, Bumseok Jeong1
Institutions:
1KAIST, Daejeon, Daejeon
First Author:
Co-Author(s):
Introduction:
Psychological and psychiatric models are often challenged by heterogeneity, where classification models lack sufficient dimensions to account for diverse mental states and behavioral phenotypes. In a previous study, we explored the utility of social affective dimensions and demonstrated that these dimensions could enhance the specificity required to identify participants with distinct emotional responses to shared, dynamic rewards in the ultimatum game (UG).
Hierarchical Bayesian modeling assumes a common parent distribution of parameters for each group. This approach provides a "shrinkage effect," enabling stable parameter estimation and the identification and explanation of individual differences. By combining these two techniques, we conducted an exploratory analysis to investigate whether hierarchical modeling, augmented with social affective dimensions, significantly improved model fit for dynamic economic decisions made by 476 human participants. We also proposed a follow-up experiment to provide confirmatory evidence.
Hierarchical Bayesian modeling assumes a common parent distribution of parameters for each group. This approach provides a "shrinkage effect," enabling stable parameter estimation and the identification and explanation of individual differences. By combining these two techniques, we conducted an exploratory analysis to investigate whether hierarchical modeling, augmented with social affective dimensions, significantly improved model fit for dynamic economic decisions made by 476 human participants. We also proposed a follow-up experiment to provide confirmatory evidence.
Methods:
A total of 476 participants played a 30-trial, one-shot UG as responders. Rewards were dynamically structured across three levels of offer fairness. Participants rated their expected offer amounts and evaluated four affective dimensions both before and after receiving offers, followed by decisions to either accept or reject the offers.
We specified four different Bayesian models using the JAGS language. Each model shared a backbone structure: trial-by-trial decisions were sampled from a Bernoulli distribution, with the probability parameter pi,t determined by a utility function and a decision temperature parameter κ. The models differed in their utility functions:
1. Reward pursuit model: Included a reward prediction error (RPE) sensitivity parameter (γj).
2. Inequity aversion model: Included an Inequity aversion parameter (δj).
3. Reward and inequity model: Included both γj and δj.
4. Reward and emotion model: Included an emotion prediction error (EPE) sensitivity γj,emo.
Each model was specified in two forms: an independent model and a hierarchical model. Model comparisons were performed using Deviance Information Criterion (DIC) values. Group labels for hierarchical modeling were based on results from the previous study.
We specified four different Bayesian models using the JAGS language. Each model shared a backbone structure: trial-by-trial decisions were sampled from a Bernoulli distribution, with the probability parameter pi,t determined by a utility function and a decision temperature parameter κ. The models differed in their utility functions:
1. Reward pursuit model: Included a reward prediction error (RPE) sensitivity parameter (γj).
2. Inequity aversion model: Included an Inequity aversion parameter (δj).
3. Reward and inequity model: Included both γj and δj.
4. Reward and emotion model: Included an emotion prediction error (EPE) sensitivity γj,emo.
Each model was specified in two forms: an independent model and a hierarchical model. Model comparisons were performed using Deviance Information Criterion (DIC) values. Group labels for hierarchical modeling were based on results from the previous study.
Results:
Model comparisons revealed that the inequity aversion model provided a better fit, with a DIC value difference exceeding 35,000. The hierarchical inequity aversion model had a significantly lower DIC value (14,304) compared to the independent inequity aversion model (37,191). The best model overall was the hierarchical model accounting for sensitivity to both RPE and EPE, with a DIC value of 4,222.
Conclusions:
These findings indicate that hierarchical Bayesian modeling, incorporating group labels derived from social affective dimensions, offers an improved explanation of individual differences in economic decision-making. We also propose a confirmatory experiment design that contrasts social versus nonsocial contexts and incorporates electrophysiological response measurements.
Emotion, Motivation and Social Neuroscience:
Social Interaction 2
Modeling and Analysis Methods:
Bayesian Modeling 1
Keywords:
Computational Neuroscience
Emotions
Modeling
Social Interactions
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
Provide references using APA citation style.
Kim, J., Bong, S. H., Yoon, D., & Jeong, B. (2024). Prosocial emotions predict individual differences in economic decision-making during ultimatum game with dynamic reciprocal contexts. Scientific Reports, 14(1), 11397.
Farrell, S., & Lewandowsky, S. (2018). Computational modeling of cognition and behavior. Cambridge University Press.