Predicting human cooperation: sensitizing drift-diffusion model to interaction and external stimuli
Lucila G. Alvarez-Zuzek ⋆, Laura Ferrarotti ⋆, Bruno Lepri, Riccardo Gallotti · 2024
Read Predicting human cooperation: sensitizing drift-diffusion model to interaction and external stimuli by Lucila G. Alvarez-Zuzek ⋆, Laura Ferrarotti ⋆, Bruno Lepri, Riccardo Gallotti (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-16121.txt.
Introduction
Lucila Alvarez-Zuzek, Laura Ferrarotti, Bruno Lepri, and Riccardo Gallotti modelled human cooperation in two thousand twenty-four in the Prisoner's Dilemma game. People adjust choices as group mood and social contact shift, but which parts of a contact sway the choice to cooperate or defect was not clear. They used a drift-diffusion model of decision making.
A Bayesian layer tracks how the model's settings evolve after each kind of contact with other players. Fit on past data, the model forecast the group's cooperation rate on unseen data well. The team then probed three what-if cases: steering a partner, rewards and punishments, and time pressure. The tool may help design pro-cooperation rules.
Catalog record
| CREATOR | Lucila G. Alvarez-Zuzek ⋆, Laura Ferrarotti ⋆, Bruno Lepri, Riccardo Gallotti |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 79 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-16121
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Predicting human cooperation: sensitizing drift-diffusion model to interaction and external stimuli Lucila G. Alvarez-Zuzek ⋆, Laura Ferrarotti ⋆, Bruno Lepri, Riccardo Gallotti Abstract As humans perceive and actively engage with the world, we adjust our decisions in response to shifting group dynamics and are influenced by social interactions. This study aims to identify which aspects of interaction affect cooperation-defection choices. Specifically, we investigate human cooperation within the Prisoner’s Dilemma game, using the Drift-Diffusion Model to describe the decision-making process. We introduce a novel Bayesian model for the evolution of the model’s parameters based on the nature of interactions experienced with other players. This approach enables us to predict the evolution of the population’s expected cooperation rate. We successfully validate our model using an unseen test dataset and apply it to explore three strategic scenarios: co-player manipulation, use of rewards and punishments, and time pressure. These results support the potential of our model as a foundational tool for developing and testing strategies aimed at enhancing cooperation, ultimately contributing to societal welfare. $\star$ $\star$ footnotetext: The authors contributed equally to the work, and they are alphabetically sorted. Introduction Decision-making, intended as the fundamental capability of selecting the most favorable course of action among various options is an essential part of our daily life [ 1 ] . Although fundamental, the human decision-making process is indeed difficult to model, due (among other factors) to the complexity of brain processes [ 2 , 3 ] , the influence of cognitive biases [ 4 , 5 ] , emotions [ 6 , 7 ] , social and environmental factors [ 8 , 9 ] , unconscious mechanisms [ 10 , 11 ] , and individual differences [ 12 , 13 ] . The study of human decision-making [ 14 , 15 ] is still an open and vast multidisciplinary field that draws on insights from various scientific disciplines, including psychology [ 16 , 17 ] , neuroscience [ 18 , 19 ] , economics [ 20 , 21 ] and public health [ 22 , 23 ] . Our work focuses on social decisions , intended as choices or judgments made by individuals within the context of social interactions or group dynamics [ 8 , 24 ] . Social decisions can encompass a wide range of behaviors [ 25 ] , including cooperation [ 26 , 27 ] , competition [ 28 , 29 , 30 ] , altruism [ 31 , 32 ] , reciprocity [ 33 , 34 ] , conformity [ 35 , 36 , 37 ] , and conflict resolution [ 38 , 39 , 40 ] . The study of social decisions often involves examining how individuals navigate social situations, make choices that affect themselves and others, and respond to the influence of social factors on their decision-making processes [ 41 ] . Social decisions are integral to understanding the dynamics of human societies and the complex interplay between individual preferences and social structures [ 42 , 43 , 44 ] . Particularly, given the importance of cooperation as an asset in human societies [ 27 , 45 ] , being able to deploy policies that enhance it as a preferred choice is crucial. Cooperation is considered fundamental in the development of successful human societies, through fostering social harmony and stability [ 46 , 47 ] , mutual aid and sharing of resources [ 48 , 49 ] , economic prosperity [ 50 , 51 ] , cultural and technological advancement [ 52 , 53 ] , social evolution and adaptation [ 54 ] , among others. Conversely, the lack of cooperation might cause compromising scenarios with respect to societal welfare [ 55 , 56 ] . In this manuscript, we explore the social decision-making process of cooperation-defection in the Multiplayer Iterated Prisoner’s Dilemma game (MIPD) [ 57 ] . MIPD is an extended version of the classic two-player non-zero-sum Iterated Prisoner’s Dilemma game (IPD), a benchmark among iterative social dilemmas, providing a model for situations in which individual rationality leads often to sub-optimal collective equilibria [ 58 ] . Iterative social dilemmas [ 59 ] is a controlled and versatile framework for studying social decision-making and cooperation, as shown for instance by [ 60 ] . These scenarios offer insights into the strategic considerations, incentives, and dynamics that shape individual and collective behaviors, making them invaluable for researchers seeking to understand the foundations of cooperation [ 61 , 62 , 63 ] . Moreover, the iterative nature of these games allows researchers to observe the evolution of cooperation over time. By studying repeated interactions, it is possible to explore how cooperation behavior develops and adapts within a population [ 64 , 65 ] . We employ a sequential sampling model, Drift Diffusion Model (DDM) [ 66 , 67 ] , to represent the human binary decision process, when choosing either cooperation or defection. Starting with the model proposed and developed by Ratcliff in [ 67 ] , a DDM assumes a one-dimensional random walk representing the accumulation in our brain of noisy evidence in favor of one of the two alternative options. The process models the temporal evolution of a random variable x(t) from a point x(0)=z\cdot a , where z\in[\,0,\,1\,] represents the initial bias of the individual towards one of the two possible alternatives, embodied by the extremes of the interval [\,0,\,a\,] . At each time step, as information is randomly collected in favor of one of the two choices, this accumulation is represented as increases or decreases in x(t) (with information accumulation rate \nu ), up until a decision is made when one of the boundaries x=a or x=0 is reached (see Materials and Methods for a mathematical description of the model). From a neuroscience perspective, the DDM free parameters (i.e., z , a , \nu , t_{0} ) have the following interpretation: the bias z is related to the prior inclination (the pre-existing opinion) the individual has about the two options, the height of the barrier a quantifies cautiousness in response, the drift rate \nu is related to the quality and velocity of the information extracted from the stimulus, hence to the task complexity, and the non-decision time t_{0} models the preparation time that individuals need before starting the decision process. The accuracy of this model is measured with the response time probability density functions (PDFs). By fitting the response time PDFs from the data, we can obtain an estimation of the free parameters characterizing the DDM. The universality of DDMs allowed scientists to use it in a wide spectrum of different topics, in order to understand the cognitive mechanisms behind human decision-making in different contexts. Previous research used the model to describe and explore human cooperation-defection in the IPD-MIPD scenarios [ 68 ] . DDMs have also been used to describe human decisions in the context of altruism [ 69 ] , food choices [ 70 ] , moral judgments [ 71 ] , and more. Notwithstanding this, in the mentioned literature, such models were employed as descriptive models. By employing the classic DDM, we can describe a behavior present in the collected dataset, but we cannot simulate any changing behavior due to mutated environmental