Are psychedelic users less susceptible to visual illusions? Evidence from the Illusion Game
Dominique Makowski, Marco Sperduti, Oliver Collins, Lucie Daudé, Joy Zogheib, Ewen Kervadec, Bruno Romeo, Pascale Piolino, Baptiste Fauvel
Study at a glance
AI-extracted from the abstract| Characteristics | Observational study |
|---|---|
| Sample size | 118 |
| Population | 118 adults: 44 never-users of classic psychedelics, 39 who used one within the last 30 days, and 35 who used one longer ago |
| Measures | Illusion Game, hierarchical drift diffusion model |
| Key findings | Psychedelic users showed an illusion-sensitivity index 22% lower on average than never-users, equally in recent and past users, and made fewer errors, consistent with the idea that psychedelics relax prior expectations. The authors caution that the absence of a clear link with use recency limits interpretative confidence and generalizability. |
Abstract
Classic psychedelics are thought to relax prior expectations, which should weaken visual illusions. Evidence is scarce, especially outside the acute state, and error rates, the usual measure of illusion sensitivity, also depend on overall performance. We tested 118 adults, of whom 44 had never used a classic psychedelic, 39 had used one within the last 30 days (the "afterglow" period) and 35 longer ago. They completed the Illusion Game, a response-time task of judging which of two targets is larger while the strength of an illusion (Ebbinghaus, Müller-Lyer, vertical-horizontal) varies independently of the real difference between them. A hierarchical drift diffusion model was fitted to every choice and response time. Both user groups set wider decision boundaries than naïve participants, gathering more evidence before responding, and made fewer errors. We then computed an index of illusion sensitivity independent of overall accuracy: the effect of illusion strength on performance relative to that of the real difference. This index was 22% lower in users on average, equally in recent and past users, and unrelated to age, language, student status, psychological flexibility, distress or the intensity of the last experience. It correlated moderately across illusions, less than general performance did. Psychedelic users appeared somewhat less sensitive to illusions, as relaxed priors would predict, but the absence of a clear link with use recency limits interpretative confidence and generalizability.