Preregistration means writing down the research questions and the analysis plan, and posting them publicly, before the outcomes are observed. Its purpose is to separate tests of predictions from ideas generated after seeing the data, a mistake that ordinary biases such as hindsight make easy (Nosek et al., 2018).
For clinical trials, the equivalent is registering the trial, with its primary endpoint, in a public registry such as ClinicalTrials.gov before enrolment. In psychedelic science outside trials, preregistration is increasingly common: a 2022 microdosing study described itself as a preregistered field and lab study and reported null results on emotion processing (Marschall et al., 2022).
Preregistration helps readers spot outcome switching, where a study quietly promotes a secondary result that happened to look good. It doesn’t guarantee a good study: a preregistered design can still be small, unblinded or badly measured. Readers can compare the registered plan with the published paper.
Sources
Last checked .