According to a new study, withdrawal is not effective in reducing online attention to problematic research papers Proceedings of the National Academy of Sciences (PNAS).
The research does not support the notion that withdrawal withdraws risks the Streisand effect, or that it inadvertently exacerbates bad research. Instead, the study demonstrates the limits of what withdrawal can do to prevent the spread of poor science.
Most notably, the study found that when withdrawing a problematic paper, the public’s attention shifted to a relative trickle and that it focused on the event of withdrawal, not the questionable research. (Withdrawal timelines vary greatly, but they usually take months or years, even when initial proof of withdrawal is mandatory.)
MK-Agnes Harvat, an assistant professor of communication at Northwestern University, says it is too late to intervene in critical references to problematic papers.
Co-author Daniel Romero, an associate professor of information at the University of Michigan, emphasized that the study did not focus on withdrawal, but only on articles after publication and before and after withdrawal. And one key point is that withdrawn papers receive more critical attention across the platform before comparatively withdrawn papers, he said.
That said, withdrawn papers also receive relatively more critical mention (negative attention) on social media before being withdrawn, suggesting that social media users are able to identify problems with them.
Impact for public discourse
Importantly, this new paper does not focus on academic quotations, but focuses on research on “highly curated” platforms such as social media (Facebook, Reddit, Twitter and more) and news websites, Wikipedia and other repositories and research blogs. (Previous research has shown that withdrawal studies continue to be cited, in many cases the withdrawal is unnoticed.)
Of course, many educators use social media, so they are making some attention to the study address. But when it comes to focusing on platforms for the public, research has important social debates that can be influenced by misinformation.
Most withdrawn papers, good or bad, don’t get much attention, but some get visibility that spread across social media, news, knowledge repositories, and blogs, Harvat says. Withdrawal studies linking childhood vaccines to autism, for example, or recently withdrawn COVID-19-related papers suggest that “the disadvantages of highly distributed problematic research may be substantial,” he added.
Critical and uncritical attention
Use Withdrawal watchHorvát, Romero, and the third co-author, Hao Peng, analyzed the amount and type of attention (“important” or not) of the 3,851 withdrawn papers obtained in the last 10 years. They then looked at the same number of authors, published at the same time from the same journal, and focused on uninterrupted papers to compare with the effect of author research.
The study found that 2,830 withdrawn and 13,599 control papers, including tracking windows of at least six months, received more attention after being published on all types of platforms. On average, the papers obtained are most frequently mentioned on social media, then through the news. They get almost the same amount of attention in blogs and knowledge. (Researchers have tracked the level of attention via Altmetric, which follows the spread of scholarly content online.)
Across a variety of platforms, withdrawn papers were not discussed significantly more than control papers immediately before their withdrawal, according to the study. About 94 percent of them received no mention last month before withdrawing.
Twitter is responsible for 80 percent of Altmetric mentions. Compared to the control paper, the finally withdrawn papers received a higher fraction of critical tweets, suggesting that Twitter could be an effective tool for highlighting troublesome research.
Ivan Oransky, its co-founder Withdrawal watch“It is understandable that the documents that have been withdrawn have received more attention in general,” he said on Wednesday. Previous research has shown that recall occurs frequently in high-citation articles published in high-impact journals, he said (this is also mentioned in the research paper), and his experience over a period of more than 12 years. Withdrawal watch “His best guess as to whether a paper will be withdrawn is how much attention it gets.”
Oransky says one of the findings of the new study is that “journal and peer reviewers are not doing a really great job of figuring out the issues they should be.” He added, “This is another reminder that pre-publication peer reviews are completely inadequate. And I would argue that it’s getting worse, because there are a lot of papers that need peer review and not enough peer reviewers. ” )
It is true that the volume of publications continues to increase and many peer reviewers are refusing the referee’s request due to burnout and other issues. Regarding the peer-to-peer crisis, James Stacey Taylor, an associate professor of philosophy at the College of New Jersey, suggested an unusual solution in his recently published book, Market with Limits: How Academia Commodification Debates Stop (Routledge): Authors are required to pay a “bounty” to the journal referee for each error they are reviewing, whether the paper is accepted for publication or not.
“Probably less than $ 10 for a bibliographic error, that is, $ 100 for a source that has been misrepresented,” Taylor explained. “Of course, researchers can reduce their liability to zero by being very careful.”
Calling educators gracious “hunters” is unlikely to “fix the error, but it will certainly reduce it,” Taylor said of the idea, which could be controversial, given that the idea of paying reviewers in any way remains controversial.
Of PNAS In the paper, Taylor said he would be interested to know more about the papers drawn for the theft and that with the results of the false research, there is a greater tendency to subsequently “contaminate” public debate (reasons for withdrawal have been studied elsewhere, but not entirely related to the amount of attention.) Papers).
Outside of that, Taylor said he’s not sure “online sources should really be concerned with their shelf life.” Withdrawals – when observed – effectively serve their purpose in academic discussions, he added.