Looking at the Retraction Watch Leaderboard, knzhou commented:
the main common feature among the top 10 isn't that they're Japanese, it's that they're almost all medical researchers. Medical research has by far the highest retraction rate of any part of science.
Why does medical research have by far the highest retraction rate of any science?
That the authors with the highest number of retractions tend to be in the medical field does not alone indicate that the medical field has a higher retraction rate than other fields. Data from 2019 at this link from the NSF (https://ncses.nsf.gov/pubs/nsb20206/publication-output-by-field-of-science) shows that the "health, biological and biomedical" field publishes several times as many articles as any other field. It shouldn't really be a surprise that when a plurality of articles are in the medical field, a plurality of top-retracted authors are also in the medical field - you could select authors by any criteria you like and likely find that many of them are in the medical field.
Furthermore, looking at just the top 10 most-retracted authors is using a highly anomalous group of authors - the linked article states that the top fifteen offenders accounted for half of all retractions. The top-most retracted authors simply are not representative of the general group of authors, no matter what field they're in. Concluding that the medical field has a higher retraction rate than other fields because the top retracted researchers are medical is like concluding that Norwegians are in general better at chess because the world's top chess player happens to be Norwegian.
The data linked in the other answer (Grieneisen, M. L., and Zhang, M.. "A comprehensive survey of retracted articles from the scholarly literature." PloS one 7.10 (2012): e44118.) does suggest that the medical field does have a higher-than-average retraction rate, as evidenced by the fact that it is above the 1:1 line. But this phenomenon is not unique to the Medical field - Chemistry, Life Science, and Multidisciplinary Science all account for approximately 1/3 more retracted articles than their background prevalence. In fact, if you draw a line on the plot from the origin to the "Medicine" point, as shown in red below, each of those other three fields lie above that line - they retract articles at a higher rate than the field of Medicine.
Overall, I agree that the medical field does retract more than its share of articles. But the evidence in the question, as well as from my subsequent investigation, doesn't actually suggest the rate of retraction is "by far the highest of any field". The number of articles retracted in medicine is by far the highest of any field, but this is true largely due to the simple fact that the number of articles published in medicine is by far the highest of any field. But the rate at which medical articles are retracted is on par with, or even surpassed by, several other fields.
First, the best evidence I have seen that medical science has an above average retraction rate was published over a decade ago in Grieneisen, M. L., and Zhang, M.. "A comprehensive survey of retracted articles from the scholarly literature." PloS one 7.10 (2012): e44118. In particular, Figure 2 (reproduced under CC-BY) provides a striking picture:
Admittely, a plot with fresher data would be appreciated. One thing that is clear from more recent publications is that the retraction rate of medical articles has increased during the last 20 years, see for example Fernandes, Bianca Barros Parron, et al. "The Secret Life of Retractions in Scientific Publications." Principles and practice of clinical research 9.1 (2023) (PMID 37693831) and references within. That editorial also highlights the different rates among medical subfields: oncology has 2.8 times the number of retractions of anesthesia despite the top three on the RetractionWatch "leaderboard" being anesthesiologists. The editorial also claims that the majority of retracted medical articles were retracted for plagiarism, and that the retraction rate for plagiarism has been rising. (The same may have happened also in non-medical fields, but I haven't seen comparable results.)
The causes of differences between fields are less clear. It has been noted that, on a journal level, there is a link between retraction rate and impact factor. Besides the potential increased willingness to bend the rules in order to publish in a high prestige venue, the higher impact factor may increase the readership of and scrutiny on individual articles. If the same principle is extended across entire fields, well, medicine tends to have the highest impact factors, which could inflate the retraction rates.
The following are additional potential causes that I consider plausible, but which I'm not aware of compelling evidence for. First, there is the perception that the consequences of research that is wrong can be quite dire in medicine compared to some other fields, which could lead to higher post-publication scrutiny and willingness of journals to retract flawed papers. Second, there is the issue that, at least from the perspective of a physicist, medicine studies horrendously complicated systems under conditions that often can't really be controlled, so the rate of honest errors and nonreproducible results may well be higher than in some other fields. Training (e.g., in good research practices and methods) could be another factor, but not one I have any real insight into.
Some possible factors: more scrutiny, high number of medical papers, higher financial stakes (big pharma) and medical authors often don't have much research training.
Not my field, but could it be that there is just more examination of claims due to the potential harm that failed research can cause? People can die when medical research is done poorly or the evidence doesn't support conclusions. Buffy (Commented on Jul 31 at 20:21)
Note that this is a leaderboard sorted by the number of individual retracted articles. Maybe something about medicine lends itself to enabling multiple publications based on a single set of data, meaning that one instance of fraud can have a wider range of repercussions than in other disciplines. (Reddit user pacific_plywood)
My guess is that most doctors haven't really gone through a research program in grad school, and then they got tossed into it. An MD != a PhD. (Reddit user Eccentric755)
There's also a ton of money in pharma, so the motivation to cheat may get big. Scott Seidman Commented 2 days ago
