Replication and reproduction context
Reports can show whether a preprint is a documented repeat attempt and whether its cited foundations have documented replication or reproduction attempts.
Source and matching
We synchronise the open FORRT Library of Reproduction and Replication Attempts (FLoRA). Each source row links an original work to a replication or reproduction and records the repeat study's characterised outcome.
A report checks the paper DOI, its known published-version DOI, and DOI-shaped references found in the cached manuscript full text. Matches are exact after DOI normalisation; titles and author names are not used to force a match.
What the labels mean
- Successful, failed, or mixed reproduces FLoRA's structured interpretation of how the repeat-study authors described the result.
- Replication uses new data to repeat a prior study or experiment.
- Reproduction reanalyses the original data or computational workflow.
Scoring policy
This preview signal does not alter Evidence, Trust, Novelty, run confidence, or publication proxies. Any future scoring use would require a separate validation decision demonstrating that the specific interpretation is reliable and applicable.
Freshness and failure states
The source is refreshed after a successful download passes schema, size, and record-count validation. An incomplete refresh preserves the last good dataset. The API distinguishes a successful check with no match (not_found) from a check that could not run (unavailable).
Known limitations
- FLoRA is curated and useful, but it is not a complete index of every repeat attempt.
- Most records are replications; computational reproductions have much lower coverage.
- DOI-only matching misses books, datasets, software, malformed references, and works without a DOI.
- A DOI found in manuscript text is treated as a cited-work candidate; extraction does not by itself establish how the source was used.