In journalism, we often joke that three cases of a phenomenon is a trend. If that’s the case, the trend of late 2019 and early 2020 would appear to be authors announcing retractions on Twitter.
In December, Joscha Legewie took to social media to say he had been made aware of an error that had caused him to retract a just-published paper on police shootings and the health of black infants. Nobel Prize winner Frances Arnold did something similar just a few weeks ago.
And now, the authors of a 2016 study on the social networks of spiders have retracted the paper after finding irreconcilable problems with their data — and the first author tweeted about it.
In doing so, she was following in the foosteps of the editor in chief of the journal that published the paper, who had himself retracted a paper several years ago. Read on for more.
The paper, “Individual and group performance suffers from social niche disruption,” had appeared in The American Naturalist in 2016. Per the abstract:
Here we track individual personality and performance and collective performance among groups of social spiders where we manipulated the familiarity of the group members. We show that individual personalities, as measured by consistent individual differences in boldness behavior, strengthen with increasing familiarity and that these personalities can be disrupted by a change in group membership. Changing group membership negatively impacted both individual and group performance. Individuals in less familiar groups lost weight, and these groups were less successful at performing vital collective tasks.
But when a question from a reader prompted the authors to revisit their raw data, their article quickly unraveled.
According to the retraction notice:
The authors hereby retract the article “Individual and Group Performance Suffers from Social Niche Disruption,” published in the June 2016 issue (pp. 776–785) of The American Naturalist. After receiving a question from a reader about the publicly available data, the authors noticed irregularities in the raw data, which were collected in the laboratory of the third author. Unfortunately, the anomalies in the raw data cannot be sufficiently explained nor corrected. As such, the results drawn from these data can no longer be considered reliable. We regretfully retract this article, whose results we cannot stand behind in full confidence.
To her credit, the first author, Kate Laskowski, of the University of California, Davis, announced the retraction on her Twitter feed and provided some background on the fate of the paper:
In the last tweet in the thread, Laskowski thanks Daniel Bolnick, editor in chief of The American Naturalist, for being “so absolutely supportive” during the process. That doesn’t surprise us: Bolnick went to great lengths to explain what went wrong when he had to retract a paper in 2016.
‘[S]ource of the anomalies in the data is currently unclear’
Jonathan Pruitt, of UC Santa Barbara, was the third author of the 2016 paper on which Laskowski was lead author. We emailed him for comment but have not heard back.
One question we had, of course, was whether the contaminated data made it into any other of Pruitt’s publications. Although we don’t know the answer, we do know that Pruitt has cited the now-retracted article more than a dozen times since 2016, including in a 2020 paper. At the very least, those papers deserve some scrutiny from the editors to see if they, too, require remediation. [See update at end of post.]
Laskowski told us that:
The source of the anomalies in the data is currently unclear. The universities where Jonathan collected the data have been contacted about this and so I would not like to comment with what would be my personal opinion. I do have other papers with Jonathan that I am re-analyzing and am already in contact with the relevant journal editors about next steps.
She added:
This has been a tough situation so it is a relief to see the positive response so far.
Update, 1530 UTC, 1/24/20: Pruitt has still not responded to our request for comment, but he did tweet on January 17 saying that his team would be retracting three papers:
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].
As of today (1/20), the paper is still listed in Pruitt’s website: https://labs.eemb.ucsb.edu/pruitt/jonathan/publications/1001
However, it is no longer on Laskowski’s website.
UCSB website. Pruitt is no longer at UCSB but rather at McMaster Univ in Canada. His website there links to Google Scholar, which still lists the paper.
People are applauding her for checking the data after being asked for it? As first author, wasn’t that supposed to be done prior to submitting the manuscript for publication?
Another paper involving the third author of this paper shows statistically implausible data anomilies
https://pubpeer.com/publications/F62202DB4DB0804CD2E65C25764549#1
Curiously, there is a reply from someone, denoted as author Noa Pinter-Wollman, with the end of the message tagged ” – JNP”, perhaps the third author of this paper?
@Matilda, you’re correct. This seem like the normal job of an investigator. I find it also unfortunate that this paper (and maybe others??) helped to gain a coveted professorship.