Learning Social Influence with Unobserved Attention
Abstract
Social influence, the tendency of our social environment to sway our beliefs and behaviors, is often cited as a driver of population-level trends in opinions or beliefs. Many quantitative theories of social influence assume that an agent interacts with its entire social environment. This assumption is often unrealistic. For example, users of online platforms typically see posts from only a small subset of the accounts they follow at any given time. We introduce a simple probabilistic agent-based model of social influence that explicitly models unobserved attention mechanisms. The first of these latent mechanisms is selection: each agent interacts with only one other agent at a time from its social environment. The second is retention: the agent chooses whether or not to ignore the opinion of the other. Our model admits an efficient maximum-likelihood algorithm which allows us to fit it to opinion traces even when these mechanisms are unobserved. On synthetic data, omitting either latent mechanism reduces the overall strength of social influence governed by the inferred interaction kernel, which hides the evidence for the underlying nonlinearity of the social influence. We study a selection of previously analyzed empirical datasets and curate several new ones. We show that models that include the latent mechanisms are often more predictive on these datasets; indeed, they typically improve prediction far more than the choice of interaction kernel does. Modeling latent attention increases held-out predictive likelihood on 14 of these 15 datasets. Each mechanism changes the size of the inferred kernel in a predictable direction, while its effect on the kernel's shape depends on the data.
Related
- Causal Inference with Unobserved Confounding: A Mixture Learning Perspective
- Infectious behaviour: Simulating the effects of communication and social influence on pathogen transmission in crowds
- Robust Design and Evaluation of Predictive Algorithms under Unobserved Confounding
- Sharp Bounds on Continuous Treatment Effects under Unobserved Confounding
- Doubly valid and doubly sharp sensitivity analysis to unobserved confounding for survival outcomes
BibTeX
@misc{thompson2026learning,
title = {{Learning Social Influence with Unobserved Attention}},
author = {William Thompson and Moyi Tian and Linh Huynh and Alexander Wiedemann and Vicki Modisette and Philip S. Chodrow and Heather Z. Brooks},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.12616},
eprint = {2610.12616},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.12616},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning Social Influence with Unobserved Attention
AU - William Thompson
AU - Moyi Tian
AU - Linh Huynh
AU - Alexander Wiedemann
AU - Vicki Modisette
AU - Philip S. Chodrow
AU - Heather Z. Brooks
PY - 2026
DA - 2026/10/08
PB - arXiv
DO - 10.48550/arxiv.2610.12616
UR - https://arxiv.org/abs/2610.12616
AB - Social influence, the tendency of our social environment to sway our beliefs and behaviors, is often cited as a driver of population-level trends in opinions or beliefs. Many quantitative theories of social influence assume that an agent interacts with its entire social environment. This assumption is often unrealistic. For example, users of online platforms typically see posts from only a small subset of the accounts they follow at any given time. We introduce a simple probabilistic agent-based model of social influence that explicitly models unobserved attention mechanisms. The first of these latent mechanisms is selection: each agent interacts with only one other agent at a time from its social environment. The second is retention: the agent chooses whether or not to ignore the opinion of the other. Our model admits an efficient maximum-likelihood algorithm which allows us to fit it to opinion traces even when these mechanisms are unobserved. On synthetic data, omitting either latent mechanism reduces the overall strength of social influence governed by the inferred interaction kernel, which hides the evidence for the underlying nonlinearity of the social influence. We study a selection of previously analyzed empirical datasets and curate several new ones. We show that models that include the latent mechanisms are often more predictive on these datasets; indeed, they typically improve prediction far more than the choice of interaction kernel does. Modeling latent attention increases held-out predictive likelihood on 14 of these 15 datasets. Each mechanism changes the size of the inferred kernel in a predictable direction, while its effect on the kernel's shape depends on the data.
ER -