Which intervention works best—for whom?

What it does

The paper Toward a Mechanistic Understanding of False News Sharing tested four widely used interventions head-to-head: accuracy prompts, warning labels, social-norm nudges, and media-literacy tips. This companion app puts those results in your hands. Choose an audience—by age, political orientation, and other demographic and psychological characteristics—and the app updates a live visualization of how effective each intervention is for that group.

Why it’s useful

Interventions that help one audience can do little—or backfire—for another, so a single “best” intervention rarely exists. Under the hood, the app uses a drift-diffusion model to break each sharing decision into its parts: the initial leaning toward or against sharing, and the way people weigh a story’s content before acting. Seeing which part an intervention moves—and for whom—helps in designing targeted strategies (for example, for older conservatives) rather than applying one blanket fix.

Good to know

The tool is descriptive: it visualizes the patterns present in the study’s dataset and does not make out-of-sample predictions. It was built by Alan Tump, a co-author of the paper, and is released open-source in the spirit of treating research data as a public good.