A confidence interval is a range of values that are reasonably compatible with a study’s data. Most papers report a 95% interval. A narrow interval means the estimate is fairly precise; a wide one means the true effect could be quite different from the headline number.
The key check is whether the interval includes “no effect”. In a 2025 meta-analysis, the Toronto LSD alcoholism trial had an estimate of g 0.21 with an interval from −0.55 to 0.99, which includes no effect (Keighley et al., 2025). In a 2026 DMT depression trial, the difference in response at two weeks had an interval that crossed zero (Erritzoe et al., 2026). By contrast, a 2012 pooled analysis of six LSD trials reported an odds ratio of 1.96 with an interval of 1.36 to 2.84 (Krebs and Johansen, 2012).
Small studies produce wide intervals, so a striking odds ratio from a few dozen people may mean little. The American Statistical Association lists confidence intervals among the estimation methods statisticians use to supplement or replace p-values (ASA, 2016).
Sources
- Keighley et al., 2025 (PMC)
- Krebs and Johansen, Journal of Psychopharmacology, 2012
- Erritzoe et al., Nature Medicine, 2026
- ASA statement, 2016
Last checked .