A Nature journal has announced that it is conducting a “priority” investigation into a new paper claiming that women in science fare better with male rather than female mentors.
The article, “The association between early career informal mentorship in academic collaborations and junior author performance,” appeared in Nature Communications on November 17, and was written by a trio of authors from New York University’s campus in Abu Dhabi.
According to the abstract:
increasing the proportion of female mentors is associated not only with a reduction in post-mentorship impact of female protégés, but also a reduction in the gain of female mentors. While current diversity policies encourage same-gender mentorships to retain women in academia, our findings raise the possibility that opposite-gender mentorship may actually increase the impact of women who pursue a scientific career. These findings add a new perspective to the policy debate on how to best elevate the status of women in science.
The paper — which shares at least a zip code with the kind of objectionable material publishers have been taking recent pains to purge from their pages — drew immediate flak on Twitter, where commenters like Joshua Miller, of the University of Alberta, expressed a mix of anger and disappointment at the research and the journal.
Oana Carja, of Carnegie Mellon, wrote:
Daniela Witten, a biostatistician at the University of Washington, wrote:
And Leslie Vosshall, of the Rockefeller University and the Howard Hughes Medical Institute, posted a link to an open letter to the journal stating that Nature Communications has an “ethical duty” to retract the article.
Nature Communications reacted quickly — by journal standards — tweeting on November 19 that:
The journal had also posted peer reviewer comments, which show that the authors were cautioned about their having initially imputed causality to their findings, that their definition of a mentor-mentee relationship based on co-authorship might be flawed, and other shortcomings of the manuscript.
As one reviewer wrote:
The authors would be on more solid ground if they just referred to their analysis as the benefits to junior faculty from working on teams of junior and senior faculty. This would make the analysis closer to Li et al. (2018) and the Guimera (2005) analysis of “newcomers” and “incumbents” but the analysis would still be original in scope and scale and in an emphasis on how the success of junior authors is associated with teamwork between junior and senior faculty. (BTW: This change would require a complete change in title.)
However, one of the four reviewers declared:
This is a well done paper. The findings are not terribly surprising, but the analysis is unusually exhaustive. …
Neither of two corresponding authors of the paper, Bedoor Al Shelbi and Talal Rahwan, immediately responded to a request for comment.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
Here we go again
This honestly is terrible, but does not surprise me in this age of cancel culture. I wish folks questioned benefits of female leadership/mentorship with equal gusto.
If there is something wrong with the paper it should be corrected/retracted. This is science, not popular media and among the worse offenders are academics who wrap their identity in a bubblewrap and are too afraid to walk into a room full of needles of scientific reality.
It’s right there in COPE guidelines. Articles should be retracted if the data are no longer reliable, in cases of misconduct or, most importantly, if enough people on Twitter are offended.
The way this paper measure ‘success’ was by number of publications and citations… one component of casual sexual discrimination (shown previously) in STEM is the lack of citation of female colleagues and their work. This adds a bump to the male mentor statistic that really makes me question whether the original authors understand causality at all. It’d be like me putting out a paper about how much greater men are at e-sports and basing it on the fact that there are many more men than women in the arena at this time and just ignore all the sexism that discourages women from joining.
Nope, read the paper. They only considered papers written without the mentor as a co-author and were comparing women with other women. Unless you think women were being discriminated against for having written papers with other women in the past, your criticism is baseless.
There is no indication of unreliable data in the note from the journal and if I missed it why is the journal “thinking” about the paper, just retract it. Why are reviewers wrong and if they are what was the editor doing?
The complainers on Twitter keep saying flawed analysis and bring BS of causality in social science (there is no such thing). The authors have said “association” in their title. The simple rule here is in any regression analysis (or similar methods) one can specify something differently and it can support or not support the hypothesis. The list of analytical checks in social science papers is nothing but a social construction. Look at the journals in applied psychology and you will see the silly drama. Can economists replicate RCTs they so proudly walk around showing off?
If social sciences, which already is questioned all the time, goes this route, it will only be glorified data journalism that no one reads.
Can the complainers on Twitter do their own study and disprove it? Will be a better use of their time than engaging in this cancel culture type behavior.
Promote diversity and inclusion, but not at the expense of scientific merit.
I am sorry, you seem to think that papers in an academic journal warrant an intellectual response. This is not true. See below:
“Misogynist white dude prof trolling women in STEM about this sexist paper is everything I hate about academia.
We don’t need to do our own research to publish a rebuttal.
I don’t care about the details of their methods.
Not everything is an intellectual exercise.”
https://twitter.com/lisa_gunaydin/status/1329653164488810496
Indeed, here we go again: Jumping straight to causation from correlation and ignoring confounding along the way. And jumping straight from a flawed analysis to cancel culture, like they have never seen papers causing an uproar upon publication and leading to reexamination of the analysis
Anyone thinking this is just Twitter outrage or people not liking the conclusions should read the reviews. There are significant flaws that reviewers pointed out (but apparently they were not listened to).
https://static-content.springer.com/esm/art%3A10.1038%2Fs41467-020-19723-8/MediaObjects/41467_2020_19723_MOESM2_ESM.pdf?fbclid=IwAR1Erxovbci0rbD1tODJc5W2MJfFvO3bCbzXhea3CHlk_WJ8H5u68-vZnIw
Firstly, I personally disagree with the conclusions of this paper. However, I believe in freedom of speech and I am against censorship. Therefore, I think that paper should not be retracted just because of pressure by the PC police and social media.
The paper may have scientific flaws but I don’t understand how it is misogynistic. If the correlation has been correctly identified, couldn’t it just be attesting to the continued existence of gender discrimination, as the first tweet says? I get that gender discrimination is bad, but I don’t understand the basis of the Twitter complaints.
I think this tweet is an excellent reply that the Nature editors should be considering.
https://twitter.com/jaeger_natalia/status/1329454764665544704
(“Another option is to write a commentary. Get together with other great scientist and point out the flaws in this data analysis. And also suggest changes to academia to be more inclusive to women such as maternity leave, more flexible deadlines. #WomenInSTEM”)
The authors of the paper missed the most obvious conclusion for their data, which is not that female mentors are bad.
The obvious conclusion is that a female scientist who continually seeks out exclusively female mentors is a different kind of female scientist than one who will seek a mentor of either sex. It’s about individual behavior.
Here I explore some strange aspects of the data set used in the problematic AlShebli @NatureComms paper. See Section 13 ‘Summary’ of this web page:
https://danieleweeks.github.io/Mentorship/
or if you want the Twitter Thread version, go to:
https://twitter.com/statgendan/status/1330238657937207303?s=21
How can one have a set of mentors with an average academic age > 200? How can one have 91 mentors?
Always always graph your data!
This is the best analysis. The paper is obviously BS because the data is biologically impossible, but the comments are full of people bemoaning cancel culture and the “PC police” whatever those are. I wish this site’s comments didn’t turn into a cesspool every time anything remotely rightwing or un-PC is questioned.
Cesspool? Not one comment here has argued that the paper or its conclusions are sound. At most they have argued that the proper way to deal with this paper is scholarly rebuttal.
This paper was always obviously rubbish but the reaction is not much better. But let’s just give the right another martyr. That will guarantee victory in the culture wars.
Retraction is reserved for the most egregious misconduct.
Its not appropriate to retract a paper because you dislike its findings.
You can’t retract a paper just because the results are politically incorrect or unpopular. If papers had to be retracted because of their failure to conform with society’s prevalent views on a certain topic, science would have nothing else to contribute to society. A mature and scientifically-minded response to this paper would have been to design a “better” study to challenge the conclusions of the paper, not to shout in anger that the authors of the paper deserve to be burn at the stake, as many scientist, some of them prominent, are doing at the moment.
Let’s be honest, it’s just a sh**ty paper and whoever the editor is should probably lose their job. Seriously, read the paper, the methods are completely atrocious and were appropriately called out by the reviewers. The paper honestly reads like one of those algorithm-written papers that only exist to see if they can get accepted. The editor is bad and should feel bad.