EGU2026 – Presentation about the Skeptical Science Experiment – Image for illustrative purposes only (Image credits: Unsplash)
Vienna, Austria – Scientists at the European Geosciences Union General Assembly shared fresh data this week on an experiment that probed the real-world impact of online climate myth debunking. Skeptical Science, a nonprofit founded in 2007 by John Cook, ran the study across its database of more than 250 peer-reviewed rebuttals. Conducted from November 2021 to July 2025, the effort revealed both promising shifts in reader perceptions and unexpected challenges in swaying the most convinced visitors.
Crafting a Real-World Test for Debunking
The team designed the experiment to capture genuine reactions from site visitors. Those arriving via Google searches for English-language rebuttals encountered an invitation to join a brief survey. After consenting, participants rated their agreement with a statement tied to the rebuttal’s topic – either a fact or a myth – on a six-point scale before and after reading.
Timing added another layer: the system recorded how long readers spent on each page. This setup mirrored everyday browsing, ensuring results reflected typical engagement rather than controlled lab conditions. A companion paper published open access in Geoscience Communication on April 2 details the full methodology.
Scale and Starting Points of Participants
Over nearly four years, the experiment reached massive numbers. More than 858,000 visitors saw the survey prompt, with 13,432 agreeing to the pre-reading questionnaire. About 6,261 followed through to the post-reading version, split nearly evenly between fact and myth statements: 3,146 for facts and 3,115 for myths.
Entrants largely arrived open to climate science. Nearly half – 46.3% – fully endorsed facts or rejected myths from the start. Still, the group included undecided readers and a notable share of skeptics, offering a broad snapshot of online seekers.
Shifts in Belief: Progress Amid Puzzles
Overall, the rebuttals delivered gains. Belief in myths dropped across the board, including among initially dismissive readers who strongly backed myths or rejected facts. This held even for those starting skeptical, pointing to potential in bolstering allies against misinformation.
Yet results showed complications. For some rebuttals and reader subsets, accuracy dipped post-reading. Highly certain fact-agreers occasionally rated lower afterward, a counterintuitive outcome the team termed a possible backfire effect. To unpack this, researchers grouped rebuttals with at least 50 completions into top and bottom performers based on perception shifts.
Patterns emerged clearly:
- Top performers (positive shifts): Consistently offered a clear replacement fact and highlighted the myth’s logical fallacy.
- Bottom performers (negative shifts): Often skipped replacement facts and rarely addressed fallacies.
These insights stemmed from post-hoc analysis, as the streamlined one-question survey prioritized participation over deeper motives. The findings echoed broader research on effective communication while tailoring lessons to Skeptical Science’s model.
Relaunch and Road Ahead
The experiment informed a website overhaul now underway. New rebuttals will adopt a fact-myth-fallacy format to boost debunking success. Volunteers drove the effort: John Cook supplied research and statements, Doug Bostrom handled backend tech, and Collin Maessen with Timo Lubitz managed programming.
Plans call for restarting surveys post-relaunch with refinements, including targeted questions on reader intent. The full presentation, titled “Results of the Skeptical Science experiment and impacts on relaunched website,” appears in session EOS4.1 on Geoethics. Access the abstract at EGU26-4110.
These results underscore a key truth in science communication: tools like rebuttals can equip the willing and nudge the wavering, but precision matters. As climate debates intensify, such experiments offer a roadmap for amplifying evidence over echo chambers.
