Integrating Persuasion Theory into the Epidemiological Modelling of Health Misinformation Spread on Social Media

Mkululi Sikosana; Sean Maudsley-Barton; Oluwaseun Ajao

cs.SIAug 16, 2026

Integrating Persuasion Theory into the Epidemiological Modelling of Health Misinformation Spread on Social Media

Authors: Mkululi Sikosana, Sean Maudsley-Barton, Oluwaseun Ajao

Organizations: Department of Computing and Mathematics, Manchester Metropolitan University, Manchester M15 6BH, United Kingdom

Abstract

This study presents a hybrid epidemiological and behavioural framework to simulate the spread of health misinformation on social media. We extend the classical Susceptible--Infected--Recovered (SIR) model to a six-compartment structure (SIRMMM), incorporating Misinformed Susceptible (MS), Misinformed Infected (MI), and Misinformed Recovered (MR) compartments to better reflect the dynamics of the misinformation lifecycle. To account for individual-level behavioural variation, we extend the SIRMMM model by integrating psychological signals from the Elaboration Likelihood Model (ELM), including sentiment polarity, engagement metrics, and cognitive effort, which dynamically modulate the misinformation transmission rate, yielding the ELM-SIRMMM framework. Model parameters were estimated using the FibVID dataset, which captures COVID-19 misinformation on Twitter. Generalisability was tested on two additional datasets: MC-Fake (emotional misinformation) and Monant (general health misinformation). Results show that the ELM-SIRMMM model enhances both predictive accuracy and dynamic realism. On FibVID, it decreases RMSE by 5.5%, delays the misinformation peak from day 150 to day 160, and increases its peak prevalence from 6% to 7%. On MC-Fake, it accurately reproduces a flash-rumour pattern, infecting 38% of users by day 45 and achieving 97% misinformation recovery, all while maintaining model accuracy. In contrast, minimal behavioural signal variability in the Monant dataset leads to marginal benefit, with only a 3% peak and 57% of users remaining susceptible. These findings suggest that structural elaboration alone is insufficient. Functional realism in modelling misinformation spread requires dynamic psychological inputs that vary meaningfully across time and contexts.

Figures & tables

Symbol	Description
$\beta$	Baseline transmission rate (SIR)
$\gamma$	Recovery rate from misinformation
$\beta_{m}$	Misinformation transmission rate (SIRMMM)
$\gamma_{m}$	Disengagement rate from misinformation
MS	Misinformed Susceptible
MI	Misinformed Infected

TABLE I: Summary of Model Parameters & Descriptions

Dataset	Domain	Size and Time Span	Behavioural Features Used
FibVID	COVID-19 Twitter misinformation	$\sim$ 360,000 tweets (Mar–Jul 2020)	Sentiment (TextBlob polarity), Engagement (likes + retweets), Cognition (word count)
MC–Fake	Multidomain news rumours	$\sim$ 28,000 cascades (Jan–May 2021)	Sentiment (polarity), Engagement (likes + retweets), Cognition (word count)
Monant	Health forum discussions	$\sim$ 370,000 threads (Jan–Apr 2021)	Sentiment (polarity), Engagement (upvotes – downvotes), Cognition (word count)
Feature normalisation: Each feature is z-score normalised and interpolated to a daily grid aligned with model time steps.

TABLE II: Overview of Datasets & Behavioural Features Used

Fig. 1: Compartment trajectories for plain vs. ELM-SIRMMM across three datasets. Simulations illustrate differences in peak timing, prevalence, and user transitions across misinformation environments.

Observation	Empirical Evidence	Implication
Signal dependence of $\beta_{m}(t)$	Only datasets with rich psychological variance (panels 2 and 3) display pronounced ELM effects.	Behaviourally responsive models require feature-rich corpora for calibration.
Correction capacity vs. peak infection	Panel 3 combines the largest MI apex (38%) with the highest MR endpoint (97%).	High $\gamma_{m}$ can offset an initially severe outbreak [ 23 ] .
Residual susceptibility as latent risk	MS remains above 55% in Monant despite a small MI pulse.	Quiescence does not imply resilience; MS constitutes a reservoir for resurgence.
Parameter congruence $\neq$ functional equivalence	$\beta$ -coefficients for FibVID and Monant point in the same direction, yet dynamics diverge.	Structural similarity alone cannot ensure cross-context validity [ 2 , 24 ] .

TABLE III: Cross-cutting Evaluation

Fig. 2: Behavioural signals modulate $\beta_{m}(t)$ in FibVID (dynamic) but not Monant (flat).

Dimension	Observation	Generalisation Issue
Coefficient similarity	FibVID and Monant $\beta_{s}$ , $\beta_{e}$ , $\beta_{c}$ are numerically similar	✓Directional consistency only
Feature variance	Monant features (sentiment, engagement, cognition) had narrow ranges	$\times$ Limited signal range meant that dynamic $\beta_{m}(t)$ modulation was not triggered
$\beta_{m}(t)$ shape	FibVID $\beta_{m}(t)$ is dynamic; Monant $\beta_{m}(t)$ is nearly flat	$\times$ Behavioural modulation was effectively suppressed in the Monant dataset
MI trajectory	Delayed, shaped curve (FibVID); sharp, early spike (Monant)	$\times$ Indicates that persuasion cues in Monant lacked impact on misinformation dynamics
Cognitive/Engagement signals	Monant dataset posts were short with few up/downvotes	$\times$ Weak signal input means the model functionally reduced to static SIRMMM

TABLE IV: Diagnostic Matrix: Monant’s ELM Limitations

Feature	FibVID	MC–Fake	Monant	Interpretation
Sentiment	$-14.35$	$-11.84$	$-14.92$	More negative sentiment increases misinformation spread [ 1 ] .
Engagement	$-32.56$	$-26.47$	$-29.84$	Higher engagement reduces spread (contrary to expectation).
Cognition (Word Count)	$-63.72$	$-59.35$	$-62.10$	Higher cognitive load reduces spread (aligns with ELM theory).

TABLE V: Fitted ELM Coefficients (

\beta_{m}

) Across Datasets

Dataset	Proxy	$\mathbf{\hat{\omega}}_{eng}$	RMSE (ELM)	Evidence used for robustness
FibVID	(a)	$-32.56$	856	Anchored to within-dataset improvement: Plain RMSE $=906$ , so $\Delta$ RMSE $=-50$ ( $\approx 5.5\%$ ).
MC–Fake	(b)	$-26.47$	856	Same negative sign under a news-level proxy without likes, supports sign stability under proxy heterogeneity.
Monant	(c)	$-29.84$	225	Same negative sign under vote-based engagement, interpretation bounded by low behavioural variance and vote sparsity.

TABLE VI: Engagement-term robustness summary in ELM–SIRMMM using dataset-available engagement proxies, consolidating coefficient evidence from Table V and fit/diagnostic evidence from Tables VII and IV .

Metric	Plain (FibVID)	ELM (FibVID)	ELM (MC–Fake)	ELM (Monant)
RMSE $\downarrow$	906	856	856	225
Peak MI day	150	160	$\approx 45$	$\approx 80$
Peak MI % pop.	6%	7%	38%	3%
MR end %	56%	60%	97%	40%
MS end %	44%	40%	$\approx 0$ %	57%
MI end %	$\approx$ 0%	$\approx$ 0%	3%	3%

TABLE VII: Key fit and trajectory summaries of SIRMMM variants (Plain and ELM on FibVID; ELM on MC–Fake and Monant).

Fig. 3: Overlay of observed incidence (blue points), SIRMMM (blue curve), and ELM-augmented SIRMMM (orange curve) on the FibVID dataset.

Dynamical Systems

Dynamical systems studies how states evolve over time, including stability, feedback, chaos, and transitions in physical, biological, and engineered systems.

See all 468 papers

Aug 31, 2026

A Multi-Branch Feature Fusion Approach for Health Misinformation Detection and Propagation

This paper presents a multi-branch fusion framework for detecting and characterising the propagation of health misinformation in online social networks (OSNs). Grounded in the Elaboration Likelihood Model (ELM) and the Theory of Planned Behaviour (TPB), the model fuses transformer-based semantics with rhetorical cues, stance representations, and psychologically motivated proxies in a unified multi-task architecture. In addition to binary classification, we introduce the Cognitive Propagation Score (CPS), an interpretable post-hoc auxiliary score computed from psychologically motivated, text-derived cues capturing argument complexity, emotional intensity, and content-derived virality potential, to support diffusion-risk reasoning when engagement ground truth is incomplete or unavailable. Experiments on three benchmark datasets, Constraint, COVID--19_FNIR, and Monkeypox, show strong classification performance, achieving ROC--AUC up to 0.9999 on COVID--19_FNIR, while propagation-oriented ranking achieves near-perfect agreement when engagement-derived supervision is available (Monkeypox, Spearman's

ρ= 0.9952

) and similarly high ranking alignment under proxy-based supervision on COVID--19_FNIR (

ρ= 0.9954

). Compared with representative literature baselines, the fusion model improves detection on Constraint and COVID--19_FNIR, while Monkeypox remains more challenging, reflecting domain- and signal-specific differences. Ablation analysis further indicates that psychological and rhetorical branches provide complementary gains beyond semantic embeddings. Overall, the framework bridges cognitive theory and neural modelling to improve transparency and to support scalable misinformation monitoring, with future work required to validate CPS against human-centred diffusion judgements.

Mkululi Sikosana, Sean Maudsley-Barton, Oluwaseun Ajao

Department Of Computing And Mathematics, Manchester Metropolitan University, Manchester, UK

Fake News Detection Automated Fact-Checking Health Misinformation Detection Feature Fusion Misinformation Diffusion

Ask PDF

Aug 16, 2026

BERTopic-Virality Prioritisation: A Scalable Framework for Thematic and Comparative Analysis of COVID-19 and Monkeypox Misinformation on Twitter

Health misinformation circulating during pandemics can gain traction rapidly, creating harmful narratives that compete with public health guidance. Most topic-modelling pipelines treat engagement as an external outcome, limiting their ability to prioritise semantically coherent topics that are also rapidly diffusing. We introduce BERTopic-VP, a virality-prioritised topic-modelling framework that combines contextual embedding-based clustering (BERTopic) with a post hoc Virality Prioritisation (VP) layer. The pipeline is complemented by a two-stage hybrid misinformation detection module that fuses a supervised content-based classifier with an external verification signal derived from public-health knowledge bases. Applied to three benchmark datasets, COVID-19_FNIR, Monkeypox, and Constraint, the framework achieves strong classification performance, with F1 up to 0.950 and ROC-AUC up to 0.989, while identifying high-impact clusters under top 1%, 5%, and 10% VP thresholds. For datasets without native engagement metadata, prioritisation is based on a logistic propensity-to-spread score, used as an ordinal proxy for diffusion potential rather than a direct measure of engagement. The results show that integrating semantic structure, virality-aware ranking, and affective-linguistic profiling enables scalable and interpretable comparative analysis of misinformation across pandemics. The proposed framework supports monitoring-oriented early warning by surfacing low-volume but high-risk narratives for analyst review.

Mkululi Sikosana, Sean Maudsley-Barton, Oluwaseun Ajao

Department Of Computing And Mathematics, Manchester Metropolitan University, Manchester, UK

Health Misinformation Detection Topic Modeling Social Media Analysis Misinformation Diffusion Fake News Detection

Ask PDF

Sep 9, 2026

How neighbourhood ideology shapes misinformation belief in densely tied social networks

With the rapid spread of news on social media, understanding the propagation of misinformation is becoming increasingly important. One factor that affects individuals' vulnerability to false information is their ideological predisposition. Despite the large number of agent-based models that focus on social influence as a driver of the spread of false claims, they often fail to explicitly integrate personal ideological biases into belief formation. In this work, we explore how misinformation spreads through the interaction between individuals' ideological biases and social influence. Our model accounts for both the strength of individuals' ideological biases and the extent to which a false claim aligns with their ideology. Social influence modifies the effects of ideological intensity and false claim alignment through network interactions. Notably, the influence of neighbours' ideological intensity on belief is strongly affected by how well those neighbours are connected to one another. These results highlight the importance of considering both network structure and personal ideological biases when modelling misinformation propagation.

Soroush Karimi, Marcos Oliveira, Diogo Pacheco

Department of Computer Science, University of Exeter, UK · Vrije Universiteit Amsterdam, Amsterdam, Netherlands

Opinion Dynamics Computational Social Science Agent-Based Modeling Misinformation Diffusion

Ask PDF

Abstract

Dynamical Systems

Explore similar work

A Multi-Branch Feature Fusion Approach for Health Misinformation Detection and Propagation

BERTopic-Virality Prioritisation: A Scalable Framework for Thematic and Comparative Analysis of COVID-19 and Monkeypox Misinformation on Twitter

How neighbourhood ideology shapes misinformation belief in densely tied social networks