About the data
ReproTrail shows what is on record about a paper after it was published: notices, replication attempts, data and code, and the papers that followed it.
Where each part comes from
- Paper details, citations and references: OpenAlex. The map shows the most-cited and the newest papers citing it, plus its references.
- Retractions, corrections and expressions of concern: Crossref, which carries publishers’ own notices and the Retraction Watch database.
- Replication outcomes: FReD, the FORRT Replication Database (CC BY 4.0), which lists 2,848 attempts to replicate or reproduce 2,578 original papers. Outcomes are FReD’s, grouped here as successful, failed, mixed or other.
- Data and code: DataCite records from repositories such as Zenodo, OSF, Dryad and Figshare that are linked to the paper as its data, code or materials.
- Possible follow-ups: later papers whose title, or the sentence where they cite the paper, describes replicating or re-examining it (citing sentences from Semantic Scholar). A “likely follow-up” also shares the paper’s topic; “mentions replication” does not. Both are leads to read, not verdicts.
What it cannot tell you
- A missing record means nothing was found in these sources, not that nothing happened. FReD covers psychology and neighbouring fields best.
- Citations are not endorsements. A paper can be cited heavily because it was wrong.
- Possible follow-ups are found from wording, not results. A paper titled as a replication may report success or failure; open it to see which.