Among retractions in general, honest mistakes only make up around 40 per cent or less of the total.

From Science Fictions by Stuart Ritchie.  Page 67.

The Retraction Watch Database isn’t a perfect list: some retractions might have been missed, since journals vary widely in the extent to which they acknowledge or highlight retracted articles. It’s also important to note that a retraction doesn’t necessarily mean fraud – many papers are retracted because the authors noticed a mistake and withdrew the paper themselves. Other retractions are more ambiguous: for instance, in early 2020 the Nobel Prize-winning chemical engineer Frances Arnold announced that her team were retracting a paper on enzymes from Science because the results wouldn’t replicate and because ‘careful examination of the first author’s lab notebook … revealed missing contemporaneous entries and raw data for key experiments.’82 Whether this implied mere error or something worse on behalf of that lead author, a student in Arnold’s lab, is unclear. Arnold’s admission was painfully candid: ‘I apologise to all,’ she tweeted. ‘I was a bit busy when this was submitted, and did not do my job well.’

Among retractions in general, honest mistakes only make up around 40 per cent or less of the total. The majority are due to some form of immoral behaviour, including fraud (around 20 per cent), duplicate publication and plagiarism. The number of retractions is also increasing over time, though this might not imply an increase in fraud: rather, it might mean that journal editors are getting wiser to it, or that authors are more willing, like Arnold, to admit that they screwed up.

In the same way that a small number of lawbreakers commit a disproportionate number of crimes in society, the Retraction Watch Database shows that just 2 per cent of individual scientists are responsible for 25 per cent of all retractions.