“Cheshire” is not a scientist, not even a former one. Lucky guy. Undeterred by the opinions of the scientific elites and their academic power, Cheshire does a great job in exposing falsified science, on Twitter, on PubPeer, and also on my site (see this previous guest post).
Now, Cheshire stumbled over a learned society journal, Bioscience Reports, published by the Biochemical Society in UK. It seems the journal has been infiltrated by fraudsters, many papers from 2019-2020 period contain falsified data. Most of them happen to fare from China, which might indicate a pattern of a paper mill. Such a mill generates fictional research papers out of thin air, complete with fabricated data and fake figures, the paying customer is usually some Chinese hospital clinician who needs English-language peer reviewed research papers for promotion, that being the law in China.
It is however not enough to generate a fake paper, the work of fiction must also be submitted to a journal and pass peer review. Some paper mills seem to offer also this service, including throw-away email addresses and even ORCID accounts. A number of scholarly journals with respectable publishers became, unwittingly or not, easy prey before. A suspicion arises each time if a corruptible editor might have helped.
The Editor-in-Chief of Bioscience Reports is presently Wanjin Hong, professor at A*STAR, Singapore. Who replied to my inquiry about 50 in part very fraudulent papers in his journal with:
“My term as EiC is ending soon so I will leave this to the new team and the office.”
The journal’s Managing Editor, Zara Manwaring, explained this surprising statement:
“I can confirm that Professor Wanjin Hong’s term as Editor in Chief is not ending as a result of these flagged papers, this was coincidental. So far we have not identified a common handling Associate Editor for these papers, but we will be continuing to review the flagged papers to see if a common Associate Editor emerges. We intend to make a further announcement when we have more information“
Manwaring also stated:
“The Journal has been investigating a number of submissions that we suspect are from paper mills. We’ve identified several patterns in these submissions and have been rejecting these without review. While our current processes have screened out some paper mills, we are learning from the current cases and are in the process of considering revised policy and requirements for submitted papers. We will also be looking into all of the flagged papers to determine if there are any additional patterns we can see in these publications, which we can then further use to determine whether a submission may have come from a paper mill.“
But now, over to Cheshire.
Can You Stand On Your Head?
More Misadventures of a Non-Scientist
By “Cheshire” (@Thatsregrettab1)
Dr. Schneider kindly invited me to again share some of my experiences as a non-scientist in the world of scientific research integrity. In this post, I delve into the dizzying array of problem papers published in the last 12 monthly issues of journal Bioscience Reports.
Published by Portland Press for the London-based Biochemical Society, a banner now warns readers that there are “potential issues” with the “scientific validity” of “some articles….” oh, here, read it for yourself:
Here’s that text if you can’t see the image:
Portland Press and the Editorial Board of Bioscience Reports has been made aware of potential issues surrounding the scientific validity of some articles published in the Journal. In accordance with COPE guidelines, we are investigating these as a priority and are contacting the authors of the affected articles to obtain additional data and information that will inform our ongoing investigation. We will provide updates on our investigation and take appropriate action on the affected papers as soon as possible.
All papers at Bioscience Reports are peer reviewed with an Associate Editor and two independent reviewers assessing the paper. Additional measures have immediately been put in place during the editorial process to carry out additional checks on papers. Please be assured that Portland Press are doing all that they can to resolve these issues and correct the scientific record as soon as possible.
This is a long British-y polite way of saying: we’ve been publishing crap for the last 12 months (at least).
This is a pretty standard image overlap and we see these all the time in many journals. I looked at this paper and found another problem with a subsequent figure:
The author did reply to Indigo saying the error was because of “no systematic folder management,” but later clarifying that they meant to “apologize for the confusion of file naming.” While the two errors might be the result of sloppy image file management, I take author responses like this with a grain of salt. Fortunately, when the author contacts the journal, their editorial process will do their best to determine if this was an error (erratum) or evidence of misconduct (retraction). Right? One would hope.
I was not already familiar with the journal Bioscience Reports but found that this journal was published by the Biochemical Society.
“Founded in 1911, the Biochemical Society exists to advance molecular bioscience, promoting its importance as an academic discipline, from grassroots level to government policy, and highlighting its role in positively effecting societal challenges.”
Sounded impressive, until I started reading more papers.
One from 2017, doesn’t look like an accident to me (repeated image, although annotated differently and with different horizontal dimensions), but what do I know? I’m no scientist:
Intrigued by finding this second paper, I decided to be more systematic in my review.
I started looking at journal papers published in 2020 about 3 weeks ago and have moved backward chronologically to almost the beginning of 2019. I have now posted about 50 Bioscience Reports papers on PubPeer that may have image problems (list at the end). Some are probably accidental, some are probably deliberate, on some I’m probably wrong. But 50.
As the list grew, I started tweeting about my findings and soliciting input from followers when I was not sure:
And to my surprise, the journal’s publisher responded on Twitter:
It was really nice to hear from the publisher at all and also that they were looking into these based only on my tweaking them on Twitter. I hadn’t (yet) corresponded with them via email; more on that in a minute. Most journals and editors don’t interact with me on Twitter, partly because I support Leonid’s efforts (if not always his style), although some will via DM or email. Most authors don’t even reply on PubPeer, let alone on Twitter. Not very many take it as far as David Sabatini:
“PubPeer may have had a role at some point, but as far as I can tell, is now a platform for resentful, anonymous, petty, failed scientists to harass those who actually make discoveries“
Or Gary Stacey:
But back at Bioscience Reports I found still more, helped by Forensically software and Hoya Camphorifolia.
Sometimes, I’m not sure what I’m seeing is really wrong, and on a few, the author Shengnan Xia has already replied (if not always in a convincing way):
“We are also very sorry for this problem. To verify our research results, we re-run the WB experiment and re-tested the protein, and contacted the magazine to correct the picture.”
As I went along, I started postulating about why this journal is so messed up. Could it be a papermill?
Or maybe it’s a predatory journal?
Then the journal’s Managing Editor Zara Manwaring got in touch with me:
Many thanks for agreeing to speak with us about these papers. I firstly wanted to let you know that we will be looking into all of these as a priority and will be taking appropriate action according to COPE guidelines and the advice of the Editorial Board. The Journal is taking this very seriously and we are planning to make several changes to prevent the publication of further fraudulent data and duplicated images. We’ve immediately introduced additional checks upon initial submission to the journal and upon acceptance to try and identify these prior to publication. We are also currently reviewing all of the figures of the papers that have been accepted and/or are currently with our production department.
The Journal has been investigating a number of papers that we suspect are from paper mills and we have been able to identify and reject many of these before peer review. Patterns that we have seen include:
- authors with multiple accounts, each with a different non-institutional email address and a single submission
- cover letters that follow the same layout and content pattern
- unrelated papers with identical reviewer suggestions
- biomarker papers that include 1-3 tables and 2 figures (usually a ROC curve)
Their list includes many of the red flags previously identified by Elisabeth Bik, as well as the pseudonymous Smut Clyde, Morty and Tiger BB8, and discussed in The full-service paper mill and its Chinese customers. This is probably a different paper mill than those previously identified, but the tactics appear similar.
The breadth of the problem caused me to continue to poke at the poor journal, suggesting their tagline “Your Home for Sound Scientific Research” must be ironic:
And taking some morbid satisfaction in seeing that one of the papers featured in the Most Cited list on their site had numerous problems too:
Below are some of the more interesting “finds” (with the acknowledgement that I may be wrong). Clicking on the image should send you to the PubPeer post.
The journal has problems, that is clear, but what are they to do about it? Here were my initial recommendations/opinions for the journal (original typos corrected) as shared with the managing editor:
“[M]y thoughts and experience in these efforts are limited. In one of my tweets I urged you to contact Dr. Elisabeth Bik. She has guided me so that I’m able to catch and properly flag many things, but her experience and skills are far beyond my own. If I were you, I would seriously seek to engage her professionally.
That said, here are some thoughts, somewhat in order of how I approach papers in your journal.
- Authors affiliated with Chinese* hospitals: Based on experience and information from Chinese researchers, these are high risk papers. Apparently MDs at Chinese hospitals need to get research published in international journals (non-Chinese) in order to advance their career. Unfortunately, most lack the time, training, fluency in English, and lab resources to get research published so they turn to outside labs and paper mills. As you note, most of these authors publish only a single paper.
- Chinese researchers: Researchers in China receive cash bonuses for getting papers published. Although this is rumored to have recently changed, some prolific authors from China seem to have gotten away with sloppy or fraudulent research for a long time. If I find one bad paper from a researcher, I’ll often find more… this is generally true for researchers from any country.
- Images: My only expertise (such as it is) is in finding problematic images. For papers that are mostly genetics, statistics, or math, I’m not much use. If I were you, I would turn to people with expertise in finding misconduct in each of these areas, such as Nick Brown or James Heathers on statistics.
- Western blots: In papers from 2012 and before, standards were different and the PubPeer moderators frown on us flagging older papers with splicing issues. Post 2012, I do look for evidence of differential splicing and flag those. I seldom catch manufactured blots, but as you’ve seen some have repeating backgrounds or reused bands and I’ll flag those when I catch them.”
* A note: references to “Chinese” researchers has been challenged on Twitter as phrasing that might be interpreted as racism. My view is that I am here identifying the country of origin of the affiliated institution (China) not a researcher’s ethnicity, which I don’t even know. You may disagree with this view… good for you.
The journal did get back to me, but apparently my suggestion that they hire Dr. Bik was not acted upon. Coincidentally or maybe ironically, another Portland Press title, The Biochemist, just ran a profile piece on Dr. Bik. The lack of self awareness is strong in this one. We will have to see how the journal deals with this, but based on their correspondence with me, I anticipate a number of Expressions of Concern to show up soon and probably some retractions… although in following COPE guidelines, journals usually take a long time to take down papers.
I do regret my suggestion that someone at the journal should be fired over this, but there seems to be something really broken here. How is it possible that a non-scientist, in their free time, for free, can identify ~50 possible papers with errors in about 3 weeks. It’s frustrating, but some of us have anger management issues.
I do want to thank the large number of Twitter users who responded to my pleas “Help with another?” That list includes: @TigerBB8, @SmutClyde, @OKalliokoski, @schneiderleonid, @MicrobiomDigest, @mortenoxe, @mlsmith55, @science_surf, @brnncaetano, @akaknijn
OK, enough flogging that dead horse.
List of papers posted (so far):
If you are interested to support For Better Science, you can leave here a small tip of $5. Or several of small tips, just increase the amount as you like (2x=€10; 5x=€25). Your generous patronage will be most appreciated!