The Impact of Micro-level User Interventions on Macro-level Misinformation Spread

TL;DR

This study quantitatively clarifies the gap between micro-level user interventions and macro-level misinformation spread, and demonstrates the limitations of evaluating misinformation countermeasures based solely on individual-level effectiveness.

Abstract

User interventions such as nudges, prebunking, and contextualization have been widely studied as countermeasures against misinformation, and shown to suppress individual users' sharing behavior. However, it remains unclear whether and to what extent such individual-level effects translate into reductions in collective misinformation prevalence. In this study, we incorporate user interventions as reductions in users' susceptibility within an empirically calibrated network-based misinformation diffusion model, and systematically evaluate how intervention strength, scale, timing, and target selection affect overall misinformation prevalence through numerical simulations and theoretical analysis. The simulation results show that, while all interventions reduce misinformation prevalence as their strength increases, as misinformation becomes more contagious, achieving a given level of prevalence reduction requires substantially stronger interventions. Furthermore, under empirically estimated intervention levels, even adjusted intervention designs, such as expanded scale, earlier deployment, strategic targeting, or combinations of interventions, yield limited collective effects. This study quantitatively clarifies the gap between micro-level user interventions and macro-level misinformation spread, and demonstrates the limitations of evaluating misinformation countermeasures based solely on individual-level effectiveness.

Figures (6)

• Figure 1: Empirical and simulated cumulative retweet counts over time. Gray lines represent individual empirical cascades, the solid black line represents their mean, and the dashed red line represents to the CTIC simulation with the estimated parameters. Vertical dashed line indicates $t=48$.
• Figure 2: Distributions of content-level suppression rates $e(a)$ for each intervention. Numbers indicate mean values.
• Figure 3: Misinformation prevalence (top row) and relative prevalence (bottom row) as functions of intervention strength and contagiousness for (a) nudging, (b) prebunking, and (c) contextualization interventions. Red dotted lines indicate the critical curves predicted by the QMF approximation.
• Figure 4: Misinformation prevalence (top row) and relative prevalence (bottom row) as functions of intervention strength and intervention scale for prebunking (a), and intervention strength and intervention timing for contextualization (b). Red dotted lines indicate critical curves predicted by the QMF approximation.
• Figure 5: Differences in misinformation prevalence relative to random target selection, $\Delta \rho_{\mathrm{X}}(\varepsilon_{\mathrm{pre}}, \delta_{\mathrm{pre}})$, for prebunking interventions with (a) degree-based, (b) susceptibility-based, and (c) distance-based target selection strategies. Red dotted lines indicate the critical curves predicted by the QMF approximation.