TUESDAY, 22 SEPTEMBER 2020In 1975, the American physicist and mathematician Jack H. Hetherington named his cat a co-author on one of his papers. He did this to avoid having to correct his accidental use of 'we' instead of 'I'. While this fake feline author was harmless, fake authors and papers are an emerging problem in the scientific world.
A growing number of papers have been discovered in which images are manipulated or even completely faked — the scientific equivalent of photoshopping your photos to make yourself look better. While the very nature of science is to criticise research methods and findings, we don’t often consider the scandalous possibility that the results are entirely fabricated.
In response to this, a group of eagle-eyed researchers are dedicating their time and effort to investigating the problem and exposing the guilty parties.
One of the biggest sleuths is microbiologist and scientific integrity consultant Elisabeth Bik. Bik is a distinguished author in the field of gut bacteria, and founded a blog that compiles scientific papers on microbiomes. After receiving her PhD at Utrecht University, Bik became the Science Editor at a biotechnology company, and then joined a medical company as their Director of Science. But beyond all of this, Bik is working hard to maintain the integrity of the scientific community.
Bik first stumbled upon the issue of plagiarism in science when reading an online book. She noticed that the book had plagiarised phrases from other scientific papers, including her own! After this incident, Bik kept a closer eye on the literature for similar cases.
When she began to look through peer-reviewed papers instead of just online books, Bik discovered that, in several papers, microscope images of protein gels and biological tissues had been manipulated to steer the results in a certain direction. These edits might have been easily overlooked, as such images are complex in nature. However, Bik noticed that the features from one area of a figure had been copied and pasted into another area. Many studies use images as a main source of data, so the fact that such image manipulation could be happening in many other fields is a big concern for research integrity.
As the list of offending papers lengthened, a formal study was launched. In 2016 Elisabeth Bik, Arturo Casadevall and Ferric C. Fang examined over 20,000 papers. As many as 4% of these papers were deemed suspicious. Examples of figure manipulation included cropping, cutting and pasting sections of an image, and duplicating data. Such plagiarism turned out to be so widespread that Bik has recently peeled herself away from her research to investigate papers with ‘suspicious figures’ in a full-time capacity.
After reading a potentially offending paper, Bik contacts the journal in which it is published. In most cases, however, the journal does not respond and no changes are made to the publication. This shows that the journals are neglecting their responsibilities of ensuring effective, transparent peer-review and taking action if its quality is brought into question.
Bik has taken to voicing her findings on Twitter (@MicrobiomDigest) or PubPeer, an online forum for peer-reviewing published papers, highlighting the problematic images. Other users can discuss the research with her, find more papers with suspicious features and, importantly, drum up attention by encouraging conversations with academics and the journals themselves. Given the severity of these offences, it is surprising that the papers were successfully published in the first place.
More disturbingly, what Bik initially thought were unrelated incidents turned out to belong to a network of deception connecting multiple suspicious papers. Digging further, Bik and a group of online collaborators unearthed a ‘paper mill’ run from China. This service allows users to pay to have their name on a paper published in a seemingly reputable journal. The results are photoshopped, the text is stolen, and the paper is rushed through peer-review via the collusion of corrupt editors. Bik and others noticed this connection from the observation that many published papers seemed to be replicas of one another, and the ‘authors’ were not contactable. More recently, Bik’s group found advertisements online for 'your name on a peer-reviewed paper', presumably by a similar paper mill.
This is not where the fraud ends. Predatory conferences convince scientists, usually those who are young and inexperienced, to pay registration fees and give presentations. When the attendees arrive, however, the conference itself turns out to be a sham event. There is no clear management, the alleged guest speakers often do not show up, and the presentation topics have little consistency both in content and intended audience. The wrongdoing is not only on the part of the organisers: sometimes researchers intentionally attend these conferences. A record of attendance at these events both raises their profile and makes them seem more experienced without any effort on their part.
There are a number of potential causes for all of these types of misconduct. In any research environment there is pressure to publish. Sometimes, quantity is favoured over quality. A high volume of research output can lead to greater opportunities, higher paid jobs, and a greater profile for those tied up in these scams. For some, these schemes are a quick and dirty way of having a more impressive publication list. Other causes include a general lack of control, rigour, or peer-review within the journal.
It is also important to think about how scientific literature is consumed by modern researchers. The publishing industry has changed dramatically in recent times. There used to be fewer journals, and one could receive these in print and be assured of their reliability. Now, there are more papers being published than ever. New journals have been established all over the world with different peer-review regulations. Literature aggregators and search engines such as Mendeley and Google Scholar are becoming more popular, so scientists often do not pay attention to the actual journals in which these papers are published. Furthermore, up-and-coming methods of distributing scientific findings such as preprints are competing with the main journals. Preprints are versions of scientific papers that are available online but precede peer-review and publication, making them more susceptible to false information. Overall, with readers interacting with the publications in a range of new ways, and research being published at increasing rates, it is difficult but more important than ever to sort the real from the fake.
Bik has made a number of suggestions to try and resolve the situation. Firstly, every journal should employ their own forensic image analyst, or artificial intelligence should be developed that can detect misleading figure manipulations and similarities between papers. Secondly, on a wider scale, national and international scientific policy must clamp down on fraudulent publications. Furthermore, scientists should be encouraged to do their own post-publication peer-review, just as Bik does, to reinforce the culture of mutual criticism which the scientific community strives to have.
Despite these problems, Bik stresses that science is still good at its core. ‘Don't get fooled into thinking that science is broken’, she says on Twitter. ‘Always ask critical questions. If someone claims that vaccines kill more lives than save lives, don't just shrug or walk away. Teach, convince, give evidence. Be a beacon of knowledge in your environment’.
Liam Ives is a 1st year PhD student in Material Science and Metallurgy at Selwyn College. Artwork by Erin Slatery.