"Preprints have enormous potential to revolutionize peer review and science"
Meet Leonid Schneider, a cell biologist, science journalist, prolific cartoonist, and a passionate communicator. After gaining 13 years of research experience in molecular cell biology, Leonid began talking about science through his blog For better Science, where he explores current issues in academia using an investigative approach. Leonid considers his blog a form of much needed social activism in academia. His Facebook page Science Cartoons displays his art and sense of humor when representing subtle aspects of the scholarly publishing process. He also publishes his seminar presentations about various aspects of scholarly publishing on SlideShare. For a while, he also regularly contributed online editorials and print features as a freelancer for Lab Times (English) and Laborjournal (German). After acquiring a PhD in Biological and Biomedical Sciences from the University of Dusseldorf in Germany, Leonid held several postdoctoral positions as a stem cell researcher.
“Candid” and “passionate” are the words that come to mind when you read Leonid’s views on scholarly publishing. In this conversation, Leonid shares some of his personal views on some issues in academia and how they could be addressed, e.g., what he feels about the traditional peer review model, the potential of preprints, open science, the ethics of research and journal publishing, vigilantism, and whistleblowing. Towards the end of the interview, Leonid also talks about how researchers, publishers, funders, and institutions can drive positive change in scholarly publishing.
Could you please share your story with us? You are a biomedical researcher by academic training. How did you become an independent science journalist?
I guess I am what some people I write about call a “failed scientist”. Obviously they published big papers in big journals and drew lots of funding, and I did not. When I was still in research, I noticed that producing reliable science is not really the same as publishing oodles of papers or getting into elite journals. Back then, I actually had no clue about how widespread research misconduct and data manipulation are. I really did envy all those lucky colleagues who made what I thought were groundbreaking discoveries and deservingly published in Nature and Cell. So eventually I started to figure out what was really happening and I left lab research, since I saw no point in what I was doing. It got so bad I intensely disliked entering a lab. At that time I was already writing journalism pieces as a side job, and then I decided to set up my own website. It is certainly not a business, but a form of social activism. For better science.
Some of the questions that follow are based on the About me and contact section of your blog. How do you intend to promote “reproducibility, honesty and fairness in academic research”?
What I do is to give voice to scientists who care about such things like research integrity and ethics, reproducibility of results, and good scientific practice in general. Obviously the traditional route of letters to editors is not working for them, and the usual science journalism is also not really a place of contact, as it is very rarely critical of anything in academia except the perceived shortage of funding. This is why I specifically invite guest posts from scientists (including former scientists) to critically discuss published papers, or just share their experiences. There are so many good scientists out there, but they have no voice. On my site they do get their voice, even if some of them prefer anonymity.
When we read scientific news, we only hear from those who are in power, who made it into a big journal, who got the largest chunk of funding, who hold top faculty positions. Are those really objectively the best scientists? If so, why do they abandon their own breakthrough discoveries so quickly and move on to the next hot-air claim? Why is so much of elite published work irreproducible? And what about those “sloppy” figures: is there any correlation to research irreproducibility? I think there is.
Other readers bring me evidence of data manipulations which I then publish. Those are mostly people with a very good eye for image duplications and some solid ethics principles. Incidentally, those manipulations are often found in the breakthrough papers by the stars of science, same people we admire in the news.
I used to be surprised about why hardly any other medium brings critical investigative articles about what really goes on in academia, but now I know that this is really difficult and could involve legal battles. I want to thank here all my donors, for their generous help and support.
Interestingly, your view – “we need to achieve unconditional data sharing, transparency in academic research and editorial processes, most importantly in the peer review” – is in line with the theme of the recently concluded Peer Review Week 2017, Transparency in Review. How can we go about achieving these things?
First of all, the incentives for change cannot come from the top. Unfortunately this is what is currently happening. Commercial publishers are leading the publishing revolution, including peer review. I am a bit worried about this, also because peer review is probably the most cumbersome, time- and money-consuming part of the scholarly publishing process. By the way, this is despite peer reviewers doing their job for free, which some scientists wish to change, but which I think is part of their academic work spectrum. So if publishers wish to revolutionize peer review which costs them so much, I worry: into which direction? What if they actually want to abolish it, and legalize predatory or vanity publishing?
Traditional peer review obviously does not work, it actually never did. I believe the problem is its confidentiality. Honestly I do not understand why some people want to increase it by having double or triple blind peer review, I think it will introduce additional levels of paranoia and nastiness. On the other hand, publishing of peer review reports improves their quality and makes personal and unfair attacks disappear, and this is despite there being no naming of the reviewers themselves. I see peer review transparency as key, the more transparency the better. This is, however, a grassroots movement, only scientists themselves can start with it, by engaging with their peers journal editors, and of course with funding agencies, or simply starting to peer review preprint publications in the open. There is no better peer review than the open, signed and discussible review of a preprint. I do hope more and more scientists will engage in it.
Preprints have an enormous potential to revolutionize peer review and science in general, and I see with a degree of worry that publishers are trying to corner that market. I hope it will not go the way of Open Access, where publishers quickly took control first of the Open Access movement and then of Open Access policies of funders and national states.
Data sharing is another issue. Published science for which the original data is unavailable is useless. Even clinical trial data can be safely anonymised before release; in fact it is the patient groups which advocate data sharing, while clinicians oppose it, allegedly on their patients’ behalf. Times are a bit too rough to take scientists on their word of honor alone. The solution is to mandate data sharing, and I think only research funding agencies can do that; journals prefer to act totally helpless. They would never retract a paper which refuses to share data, if its authors threaten a lawsuit.
We need real robust open science, but what we get instead is some publishers redefining what open science is. They say it is exactly the same as author-pays open access publishing. Open science means open data. Unconditionally, there are no real excuses.
You have, on more than one occasion on your blog, alluded to a climate of fear in the scientific community. Could you elaborate? Is this a pessimistic view, or do you believe there is hope and things could get better, if they are not so today?
The climate of fear is what brings people to my site, what makes them share my articles by email or word of mouth communication instead of sharing on social media, and which makes them sign their guest posts with “anonymous.” Globally, this fear is what prevents people from peer reviewing preprints or doing any post-publication peer review at all. Certain elite scientists actually boast about how much power they have and make it quite clear what they can do to those criticizing their work.
It is incredibly easy to crush an academic’s career by placing some phone calls or dropping some emails, and people are well aware of this. You write one critical review on PubPeer or elsewhere, and even if you did not sign it, your name can be guessed, because science is a village, if you look at any specific field of research. People know each other, they peer review each other’s manuscripts and grant applications, they sit on each other’s tenure committees. Suddenly you don’t get conference or seminar invitations, your grant proposals and papers are rejected by anonymous reviewers, and come evaluation time with your faculty: your head rolls.
It’s not just research misconduct or simple bad science which academics are afraid to criticize for fear of retribution. Sexual harassment is also something which only a woman with career suicide intentions will report. Never mind the general abuse and humiliations early career researchers have to endure in silence.
This is what the sacred confidentiality of peer review caused: it makes punishment and destructions of critics very easy. Are scientists really proud of the monster they created? Are they consoling themselves that feeding that monster will let them be eaten last?
I’d now like to elicit your views on some of the ethical aspects of research and its communication. In one of your SlideShare presentations, you talk about how scientists manipulate data. Could you think of why they might feel the need to do so? How can this problem be addressed?
There are basically two reasons to manipulate data. Simply put: one is to bend the truth, another is to steal it.
In the first case, scientists try to force the reality to adhere to their visions and ideas of a paper. Many researchers start a project with its outcome already set in stone, but this is not how science works. In fact, this is the whole reason why you do experiments: to test your theories and to continuously adjust those to your results. Not vice versa. But for some researchers, if the experimental data does not fit, or is lacking, it is invented - sometimes in Photoshop during figure preparation, but mostly already in the lab - and others will never get to know it, unless someone on the inside blows the whistle. There is a huge spectrum of data manipulation, from “minor” cherry-picking, “clean-up” of raw data, or avoidance of control-experiments (since those might overthrow a pet theory) to outright fraud and data fabrication.
The second case is the scoop. If you see unpublished results at a conference, you can claim all the fame and merit for yourself by scooping that clueless colleague of yours. Naturally, that colleague invested months or even years to obtain all those results, time which you don’t have. This is where data manipulation or fabrication helps. The crazy thing is: even if you do get caught, you won’t be found guilty of fraud, because your results are scientifically reliable. Only thing, those are in reality not your results, but stolen ones. But who knows this, and who cares?
Platforms like PubPeer are tremendously helpful in exposing data manipulations; in fact PubPeer revolutionized science a lot. But why do I ever say platforms? There is only PubPeer, and nothing else, which I see as a huge problem. But what will happen if someone enforces the shutdown of PubPeer, e.g., the French CNRS which employs its owners? Not so far-fetched, there are quite a number of very senior CNRS figures who have their papers flagged on PubPeer.
We had precedence already, when Jeffrey Beall was forced to delete his list of predatory publishers, because a certain European publisher, unhappy about its placing on the list, pressured Beall’s academic employer, the University of Colorado for long enough.
The so-called research integrity Watchdogs of RetractionWatch decided against whistleblowing articles, or post-publication peer review, or even overtly crucial reporting, things they did occasionally publish in the past (which now seems distant). Incidentally, the site grows from strength to strength with this new approach.
Well, and I am getting sued, for posting information which certain German professors and doctors don’t want to become public.
I don’t have any good suggestions on how safe whistleblowing can be achieved, without the danger of the site being removed forever. Except maybe if many do it, on different sites, a critical mass will be reached.
Based on your research experience, what are some of the main problems related to journal publishing specific to biomedicine? How do you think these problems could be solved?
I don’t think there is a specific problem of biomedical publishing. It appears as most corrupt on PubPeer, but this is because it is so reliant on image data. Anyone with a good eye can spot data manipulations without having a clue about what the figure actually shows. This is different with other natural sciences, where you need to be a specialist to find data rigging.
But on the other hand, exactly this ease of vigilantism offers the opportunity to clean up biomedicine. Mandate data openness and act on evidence of data manipulations. It is not so difficult actually. There is a huge community of people willing to help weed out bad science. The problem is, they are viewed as enemies of science, or as trolls. Incidentally, I myself was recently certified as a non-peer reviewed “internet troll” by a director of a major plant science research center in the UK.
Let’s go back to revisit your experiences during your research phase, as an author trying to get published. What were some of the challenges you faced? Do those challenges still exist?
My own experiences as scientist were that traditional academic publishing is completely rotten, and back then we also had no clue what a preprint might be (strangely even now many biomedical researchers don’t know). So the only advice I can give early career researchers who care about research integrity: don’t try to compete in high-impact publishing; you will most likely lose against your dishonest peers. You won’t get your scientifically solid paper in a top-shelf journal unless you agree to cut corners or if you are very, very, lucky. Do you want to take that gamble? Your decision. Otherwise, chose the journals where you know and trust the editors and the papers published there. And most importantly: preprint! It offers you scoop protection and gets your work out and into your CV before it enters the battle of traditional peer review. You might even get good reviews on your preprint, and establish important contacts in this way. Preprints can save and even make your career.
You invite suggestions for journalistic stories on your blog. What kinds of suggestions have you received? Any interesting one you could share with us? Also, on a broader level, have these suggestions shaped your journalistic approach?
As I said above, I always welcome guest posts mostly on post-publication peer review, whistleblowing, or your general past experience with academia. I learned of things which I previously would have never deemed possible. Really, you won’t believe how bad the secretive academic environment can get.
Most recently, a citizen scientist from the US, a retired industrial plastics researcher, contacted me about a paper in Science, and I published his guest post. That Science paper was retracted because its Swedish authors were found guilty of research misconduct and most probably fabricated their data on fish larvae eating plastics. But that citizen scientist pointed out how shaky the peer review process at Science was in the first place, and that this paper should have never been accepted, regardless of its later misconduct findings. Scientifically, it made very little sense as it was. But the central message was so exciting (juvenile fish eat plastic pollution as junk food!) that the elite journal overrode all objective quality concerns. It would be very helpful to see that paper’s peer review reports. But of course they are confidential.
As a science journalist passionate about driving the change for better science, what is your message to each of these stakeholders in scholarly landscape: researchers, journals and publishers, funders, academic institutions, science journalists?
I’ll make it point-by-point then.
Researchers: Stop being afraid of each other! Share data! Seniors: Take responsibility for the early career researchers you mentor! Start advocating with funders and faculties for a change in how science is evaluated! Lead with example and do open peer review, especially on preprints! Juniors: Push for preprints and post-publication peer review; do it, don’t rely on others! Network with people who value reproducible and open science, and not those who publish in the biggest journals.
Journals and publishers: Stop influencing or controlling scholarly communication policies! Your services are only needed on the technical side. Scientists can do peer review and all science-related editorial tasks themselves, thank you very much. PS: you don’t own science and its publications, you just print them.
Funders: Mandate unconditional data sharing! Stop rewarding by impact factor! Sanction misconduct, because universities or journals won’t do that. Take back our money from data manipulating cheaters (or their uncooperative universities) and give it to their honest peers instead. If unsure if this is the right way, just ask the tax-paying public on who should get our money. Start printing your own research! You have enough money; you can pay publishers for the technical service. Your scientists will do the rest - editing, peer review, gratis. Do everything transparently, or it will be a huge cock-up.
Academic institutions: Stop covering up misconduct. You do not look too great when you do it. Everyone laughs at your pathetic attempts to declare that all those data manipulations your professor did were non-existent or accidental. Cooperate with external investigators and funding institutions. Honesty and integrity will eventually pay out more than trying to save those bits of funding your cheaters dishonestly obtained. Print your own research in your own journals, just like suggested for funders above.
Science journalists: Stop trusting scientists as if it was a religious sect and all they say is a gospel. Get proper peer opinion, not just some pro-forma one. Scientific reporting is not just for entertainment. Some people, like patients, do take it seriously still. Your gullible reiteration of press releases can damage science enormously, or even get people killed or mutilated. Did you ever think of that? Try to be a bit critical. Read some political journalism to get a clue what this means. Take responsibility for your past reporting. If you extolled some alleged breakthroughs which afterwards proved bunk or fraud, issue a correction. You owe it to your readers.
Thanks for sharing your opinions, Leonid. We appreciate your candor.
Disclaimer: All views expressed here are the interviewee's own and not necessarily endorsed by Editage Insights.