Retraction notices: 72 in the week of Oct. 19, against a median week of 101

The count is retraction notices indexed by Crossref, whose retraction records come from Retraction Watch and from publishers, dated to the week each notice was first seen. The week of Oct. 19, Monday to Sunday, produced 72 notices; the baseline is a median of 101 notices a week over 17 earlier weeks. Crossref reports the week as complete through Oct. 28.

Earlier weeks in the series ranged from 61 notices, in the week of Aug. 31, to 168, in the week of Sept. 28. Set against the 16 complete earlier weeks plotted, the 72 ranks at the 31st percentile, with a median of 96.5 and a typical spread of 26.5 notices a week, which puts it 0.9 spreads below that median.

Notices in the week include one in Water Resources Research dated Oct. 24, one in Biological Trace Element Research dated Oct. 24, one in BMC Women's Health dated Oct. 24, and one in Frontiers in Cellular Neuroscience dated Oct. 23, according to Crossref retraction data.

A retraction notice is a bibliographic record that an article has been retracted; the reasons a journal gives sit in the notice itself, and this count does not sort notices by reason. Crossref's documentation describes the retraction data as drawn from Retraction Watch and from publishers (library/science/crossref-retraction-watch-data.md; https://www.crossref.org/documentation/retrieve-metadata/retraction-watch/).

Room for disagreement

The earlier weeks range widely, from 61 to 168 notices, so a single week's 72 is well inside recent variation. Using only the 16 complete earlier weeks plotted gives a median of 96.5 rather than the 101 of the 17-week baseline, which narrows the gap, and a weekly count also reflects when notices reached the index.