Belief-Sim: Towards Belief-Driven Simulation of Demographic Misinformation Susceptibility

TL;DR

This paper introduces BeliefSim, a framework that uses belief profiles to simulate demographic differences in misinformation susceptibility, achieving high accuracy in modeling human-like responses.

Contribution

It presents a novel belief-driven simulation method for demographic misinformation susceptibility using LLMs, integrating psychology-informed profiles and evaluation strategies.

Findings

Beliefs significantly improve simulation accuracy, up to 92%.

Simulation captures demographic variations in misinformation susceptibility.

Framework effective across multiple datasets and modeling strategies.

Abstract

Misinformation is a growing societal threat, and susceptibility to misinformative claims varies across demographic groups due to differences in underlying beliefs. As Large Language Models (LLMs) are increasingly used to simulate human behaviors, we investigate whether they can simulate demographic misinformation susceptibility, treating beliefs as a primary driving factor. We introduce BeliefSim, a simulation framework that constructs demographic belief profiles using psychology-informed taxonomies and survey priors. We study prompt-based conditioning and post-training adaptation, and conduct a multi-fold evaluation using: (i) susceptibility accuracy and (ii) counterfactual demographic sensitivity. Across both datasets and modeling strategies, we show that beliefs provide a strong prior for simulating misinformation susceptibility, with accuracy up to 92%.