Why do observational cannabis studies often conflict with RCTs?

The frequent discordance between observational studies and RCTs in cannabis research reflects fundamental differences in what each study type measures. Observational studies (surveys, registries, retrospective analyses) capture real-world use patterns but are confounded by: (1) Selection bias — people who choose cannabis may differ systematically from those who don't (healthier, more health-conscious, or conversely, more desperate). (2) Recall bias — self-reported cannabis use and symptom improvement are unreliable. (3) Confounding — cannabis users often make other lifestyle changes simultaneously. (4) Placebo/expectancy effects — patients who believe cannabis will help them report improvement regardless of pharmacological effect. (5) Publication bias — positive observational findings are more likely to be published. RCTs control for these biases through randomization and blinding, which is why they consistently show smaller effect sizes than observational studies. The gap between patient-reported benefits (often dramatic) and RCT findings (often modest) is a consistent pattern in cannabis research.