Statistically significant findings are more likely to be published. Because of bias and manipulability in data analysis, much of the research published in journals is irreproducible.
Registrations in OSF are time-stamped, read-only versions of a research plan that explain how data will be collected, manipulated, and analyzed in advance. Providing this Registered Report to a participating journal effectively creates two peer-review processes.
Articles resulting from accepted Registered Reports are accepted by the journal regardless of outcome after the peer reviewed quality checks are met.