This guest post comes from a non-scientist, not even a former scientist or a citizen scientist or someone who always wanted to be a scientist. Cheshire is namely a person with a proper job, and always was. But they recently decided to apply their common sense and their good eye for image irregularities to flag bad science on PubPeer and to the alert journals and publishers, inspired by the legendary Elizabeth Bik. Let’s hope Cheshire will in turn inspire others, because research integrity needs less peer reviewed literature on the topic, but instead more activism, by scientists and non-scientists alike. It needs every helping hand to actually expose fraud, incompetence and quackery. This is Cheshire’s guest post.
Down The Rabbit Hole
The Misadventures of a Non-Scientist
By “Cheshire” @Thatsregrettab1 on Twitter
Dr. Schneider kindly invited me to share some of my experiences as a non-scientist in the world of scientific research integrity. In this post, I’ll briefly explain how I got here and then recount a few of my adventures along my short (so far) journey. A note: I have posted here and elsewhere without using my real name in order to keep my professional and private Internet profiles separate, to the extent that this is possible. I appreciate Dr. Schneider’s support of that preference.
While I have no scientific background, I’ve had a layperson’s curiosity about science, medicine and technology for many years and regularly read blogs and news sites on these subjects (a few: Ars Technica, Neuroskeptic, Science Daily, Phys.org, Retraction Watch). The last is the most relevant to the story here, as I read with incredulity their November 2018 story about Dr. Richard Maximus Fleming. Until I read that story, I was primarily a passive reader of that site, although I posted occasional questions and snarky comments.
Dr. Fleming, aka Dr. who?
As I recounted in some of my comments on that Retraction Watch blog post (there I post under the name “Regret”), I was unaware of Dr. Fleming until that article but was curious enough about this apparent modern-day Renaissance Man that I hopped on Google to read more about him. And that started this adventure for me.
In the Retraction Watch comments, I shared some of what I’d discovered, namely that Dr. Fleming seems a bit, shall we say, inconsistent about his credentials and affiliations. But surprisingly, Dr. Fleming responded to questions and comments on Retraction Watch with entertainingly meandering vagueness and arrogance. This engagement petered out in late November 2018, but the discussion got me to sign back up for Twitter (where I started following scientists, especially those interested in research integrity), and to comment on and also to post a few of Dr. Fleming’s “papers” on PubPeer.
To my great enjoyment, Dr. Fleming responded to the PubPeer posts (on PubPeer and Twitter) and has continued to not only publish his “research” in low quality journals but to strike out against those questioning his research or background. He challenged Dr. Elizabeth Bik to a debate on national TV and responded to one of my comments on PubPeer with:
“I have seen these same KETO stalkers before using the same fungal name identifiers attacking science that they did not like so they could promote through stalking, bullying and harassment their agenda. Actinopolyspora Biskrensis, Paraburkholderia Zhejiangensis and Lyophyllum Pulvis-horrei are not only stalkers, harassers and bullies but cowards. I have reported you on these comments and am recommending that you be prosecuted accordingly.”
Long story short – Dr. Fleming got me hooked.
Twitter and PubPeer communities
In my initial posts on PubPeer, all of which were moderated by the PubPeer moderators, I was focused on Dr. Fleming’s published “research,” however, during this process I discovered belatedly the rules of posting on PubPeer as several of my posts about Dr. Fleming’s felony conviction and FDA debarment and his fuzzy CV were either not posted or were removed after posting for violating those rules. I might have actually read those rules at some point 😊, and I was later upgraded to a status where I could post an article without prior moderation. I also started seeing other interesting posts on PubPeer where authors had apparently manipulated images in published papers, and I was fortunate enough to bump into Dr. Bik on Twitter (@MicrobiomDigest). I’m sure most readers of this article are already familiar with her and her work, so I won’t waste words on that.
Dr. Bik was not only supportive of my original commenting on Dr. Fleming, she was willing to help give me guidance on PubPeer best practices, and help me understand what I was seeing with individual images before I posted concerns on PubPeer, which many authors seem to immediately view as a black mark on their record. I’ve had similar support from Dr. Schneider and some other named or pseudonymous Twitter users, most notably @TigerBB8, @SmutClyde and @mlsmith55. Thanks to you all and apologies to those I forgot to acknowledge… there are definitely more.
On PubPeer I was assigned the name Actinopolyspora Biskrensis by their naming convention (Dr. Fleming helpfully notes above that this is a fungus), and I’ve found the universe of users there and on Twitter to provide a lot of insight into science, research publishing and human nature. I’ve been (mostly politely) slapped down when I’ve flagged papers or images that others did not believe were manipulated and my own judgment has become more refined. While I depended primarily on Dr. Bik and Dr. Schneider for assistance at first, the list of Twitter users willing to help me has grown and I will generally get a prompt response to my “Help with another?” tweets asking for input from experienced scientists.
I have also reported some concerns directly to authors’ institutions and journals via anonymous email. Many of these emails go unacknowledged, which I think is fine, but several journals and at least one institution have acknowledged my email and have reported back to me at a later date about any action taken. Although I have deliberately not pressed for details in this situation: I believe one organization has disassociated themselves from an individual about whom I reported some potentially concerning issues. I will say that none of those who have responded have been dismissive of my concerns, they have all responded professionally and politely… which is nice.
Image concerns posted to PubPeer
Searching through papers I mostly uncover image concerns like those shared below. With help from Twitter users, I first confirm that these are worthy of posting to PubPeer. Since I don’t have a scientific background, I am really really unqualified to comment on the scientific content of the papers I skim, but occasionally I’ll find other types of problems regarding conflicts of interest disclosures, concerns with author affiliations, or image reuse in multiple papers. I have elected not to systematically look for plagiarism, as I lack the professional tools and access to libraries of papers that would make this process more efficient. Some examples things I’ve flagged on PubPeer (rightly or wrongly) follow:
Here’s my typical process: First, I post on Twitter:
In this case, I am successful in getting a “real” scientist (or at least real enough to fool me!) to confirm that they agree this is an image problem (see below),
So, I post this concern on PubPeer, and in this case, get an almost immediate response from the author whose email is tagged in the PubPeer system:
Mission accomplished? I’m not sure. In this case, I can see how it is possible that this was an honest mistake, although I personally do not know how the coloring on the two images would be different if the same image was accidentally reused. The article has yet to be updated on the journal site, so we’ll see if the editors accept the author’s explanation (or whether the journal was even contacted).
Many of my posts to PubPeer are ignored by the authors, even when at least one author has a confirmed email address on PubPeer (as indicated by a check mark). Below is an example of one of those (link).
To a non-scientist, it is difficult to see how this type of image manipulation could occur accidentally during figure assembly, and the similarity is strikingly obvious with the two images appearing in the paper side-by-side. Perhaps the absence of an innocent story explains the lack of response to the PubPeer concern. Sometimes the paper I’ve flagged is published by a lower-quality or predatory journal, which could explain the reason a pre-publication peer review did not catch this and why neither the authors or journal will bother to correct the record. <shrug>
Sometimes after I post a paper to PubPeer, authors and other readers will both respond, and I’ll generally keep my head down as both sides make their cases. As I don’t have the scientific background to add any substance to the debate, I basically stay out of it after my initial post. Here’s one of those (link).
At the end of that thread, the author seems satisfied that they have provided a sufficient response, although in my reading, other commenters didn’t seem as convinced.
I hesitate to send a notification to a journal in these type of cases, because I really don’t know whether any of the concerns discussed in the thread are worthy of their attention. But I have had correspondence with journals about cases I have reported and they have generally been appreciative if not always supportive of my concerns.
Another image problem, another PubPeer post, another ghosting by authors. Probably a simple error, so not sure why authors don’t respond on PubPeer since they have confirmed emails.
Some of my PubPeer posts get author replies that make even a non-scientific amateur to roll my eyes, and in this thread, Dr. Bik chimes in to indicate she didn’t believe the authors either.
“Thank you for your attention. We have verified the images. The images in Figure 5C (Left Panel) and 5D (Left Panel) marked with green border represent vehicle (DMSO) injected controls hence there was a mix-up. However, images in Figure 5B (Right Panel) and 5C (Right Panel) marked with the blue border have been taken at different locations of the same section. Images in Figure 5A (Left Panel) and 5D (Right Panel) marked with the red border are similar. This mistake has occurred inadvertently during the selection of images for preparing the final panel for publication. We communicate to the journal about this inadvertent mistake to notify an erratum to this effect. This correction does not change in any way the content of the study. G Bhanuprakash Reddy”
In reply, Dr. Bik:
“The duplications in Figure 5 do not appear to be simple errors during the preparation of this Figure. These are not simple duplications, but instead appear to show stretching in the panels marked in green and blue, suggesting that these are more than just figure assembly errors. Even if these are errors, it is unexpected that 3 errors were made in a single figure. Could the authors please comment how three inadvertent mistakes could happen within one figure?”
You can read the rest of the thread at the link.
I do get some feedback on PubPeer posts when my amateur status takes me off course, but this doesn’t happen too frequently thanks to the support of my followers on Twitter:
“would the presence of a splice affect the result? If they were presented as separated lanes, would it affect the interpretation? This would provide context for what sort of “comment” you are requesting from the authors. Note that in these gels the key feature is the pattern of mobility, and not the intensity level (where it is more important to confirm that the samples were run side-by-side).”
I don’t take these valid concerns and comments personally. I am still an amateur and I am still learning.
Below I recount some of the more interesting (to me) episodes in the 6 months I’ve been doing this. Some of these have been written up on Dr. Bik’s and Dr. Schneider’s blogs, so please read their sites for more details.
William Figg – July 2019
I often browse ResearchGate looking for papers with images that might have problems. In early July, I came across the image below.
I posted a “Help with another?” message on Twitter and two followers responded to acknowledge that they agreed this might a problem image, with one pointing out the issue with left-most image. I posted this to PubPeer and the last author acknowledged “unintentional image duplications” and provided a replacement image purportedly from the original thesis of the first author. Again, with the urging of a Twitter follower, I looked at that thesis and the image wasn’t identical to the one published there.
I noted on Twitter that while many of the authors were with an institution in Argentina, the author list included a William Figg affiliated with the National Cancer Institute (US). A quick Google search showed that he is a senior member of that organization and that he was on Twitter, so I tagged him there. To my surprise, he responded on Twitter that he was looking into the matter. With his involvement over the July 4 holiday weekend, the issue was swiftly resolved with the last and senior author pledging to retract the paper due to “multiple issues” (as described by Dr. Figg). While I was pleased that my inquiry resulted in a correction to the scientific record, to my personal dismay, it appears that this paper may be the PhD thesis of the first author, which could result in additional consequences for that person. Hmmm. Dr. Bik had warned me of real-world consequences of this type of work and I’m not super comfortable possibly hurting someone’s probably legitimate career.
One thing that I’ve done on Twitter that seems to catch the attention of some authors whose work I’ve questioned on PubPeer is (what I call) critical humor. No news to anyone reading this that there are some big egos in science, and authors sometimes try bullying tactics in response to PubPeer questions and concerns, while others have responded with implausible explanations for misconduct masquerading as mistakes. I’ve learned that personal criticism will get me put in moderator prison on PubPeer and I was banned once on Twitter, but I’ve found some authors to be more responsive to satire than honest concerns expressed about their work. Note: while the individuals I’ve poked fun at may not appreciate the humor at their expense, in the legal jurisdiction from which I write, laws protect speech that is critical of public figures (scientists publishing their research) and commentary about matters of public concern (scientific and medical ethics), both of which apply here.
For example, Gary Stacey initially dismissed concerns expressed about multiple papers, but after a while, the repeated attention I and others were bringing to the matters got his institution to acknowledge that they were looking into the matters. I created the image below as a tribute to how Dr. Stacey treated Dr. Schneider’s comments on Twitter, which I suspect help goad Dr. Stacey into action.
Dr. Richard Fleming initially had an account on Twitter on which he flogged his “research” (published in low quality and predatory journals, and the vanity press articles he had published. After facing much criticism on Twitter and elsewhere, he dropped off Twitter. Later, a user surfaced that I suspected was a sockpuppet created by Fleming (@Veritas34407030) self-described as “Journalist with experience sifting through BS,” who blocked me after I responded to “Veritas” with a couple of satirical images of Fleming. While “Veritas” stayed in character for a bit, he blew his cover with a July 4 tweet about Fleming’s patented method “FMTDVM,” and “Veritas” has been pretty quiet recently.
I gave Dr. Fleming the nickname “Dr. who?” because of his relentless and shameless (and so far fruitless) self-promotion and created the image below in his honor. The photograph is a still shot of a short acting video posted to YouTube (worth a watch), because yes, he’s an actor too.
Last here is an image I’ve used a couple of times on Twitter to illustrate the point that a figure that appears to have been assembled from multiple images is probably misconduct, not just a mistake.
I’d like to end this post by soliciting reader feedback. Whether or not you have encountered my posts on Twitter or PubPeer before, what do you think about a non-scientist doing this? Should a non-scientist even be allowed on PubPeer? Can you recommend training or references that would help me do a better job (and not distract professionals like Dr. Bik and Dr. Schneider as much)? Any other criticisms, suggestions or ideas?
If you are interested to support my work, 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 of my journalism will be most appreciated!