Warns against fake research
PhD student Andreas Volstad was delayed by several months because he discovered something strange in the studies he was using for his research. "This can have consequences for both research and clinical practice," he says.
What researchers refer to as literature reviews or meta-studies have been considered some of the most robust and reliable forms of research.
These are large studies that summarise the conclusions of research within a particular field.
Such studies are also highly significant in practice. They are used, among other things, to determine the best medical treatment for a disease.
However, several researchers now warn that more and more fake research articles are sneaking into these summaries.
This could jeopardise trust in research, they argue.
Noticed something strange
Norwegian researcher Andreas Voldstad is one of those who has had bad experiences with such literature reviews.
Voldstad is pursuing a PhD in psychiatry at the University of Oxford in the UK.
During his work, he and his colleagues conducted a literature review of studies on mindfulness.
That is when they noticed something strange, he tells the journal Science.
Suspicious effect sizes
The language in some articles seemed to come from an automatic translation program.
Some articles lacked adequate reporting.
In a few articles, it was primarily the enormous effect sizes that raised concern. Some described that the effect of mindfulness was between 10 and 20 times greater than those documented in other studies.
They also became concerned about the characteristics of some journals, Voldstad writes in an email to sciencenorway.no.
"We realised we had to take our concerns seriously since we aimed to reach credible conclusions," he writes.
Could influence conclusions
Eventually, Voldstad and his colleagues included all the studies in their analysis while documenting their concerns about some articles.
Additionally, they performed a separate analysis excluding the potentially problematic studies.
"This had a significant impact," he writes to sciencenorway.no.
Voldstad believes studies that are potentially untrustworthy can influence the conclusions of literature reviews.
"This can have consequences for both research and clinical practice," he writes.
Shocked
Several other researchers told Science they had encountered the same issues as Voldstad and his colleagues at Oxford.
During the process, Voldstad contacted Dutch researcher Rene Aquarius, who had experienced the same issues in another research field.
When Aquarius reviewed research on stroke treatments, he was shocked, he told Science.
Most viewed
The first red flag appeared when he saw the sheer number of studies published in this area. He had expected between 50 and 60 but found 600.
When he and his colleagues examined the studies, they discovered numerous problems with many of them.
Among other issues, they found that the same data had been used in multiple articles, while researchers claimed they were from different experiments, Aquarius told the website Retraction Watch.
Months of delays
The extra work of scrutinising the research literature caused Andreas Voldstad several months of delays.
He notes that several tools have recently been developed to assess the credibility of literature reviews.
To ensure credible results, such tools should become standard, he believes.
"That way, credibility can be assessed systematically instead of addressing concerns afterwards, as we did," he writes.
A major issue for research
Both Science and Nature, two leading scientific journals, have recently revealed that fraudulent research articles are an increasing problem.
Experts suspect these articles come from paper mills that offer researchers the chance to have their name listed as an author on a research article for payment, even if they did not contribute to it.
These mills have been found in countries where researchers are motivated to publish as many scientific articles as possible, such as China, Russia, India, and Iran, according to The Conversation.
Strong research cultures
Helene Ingierd, director of the National Research Ethics Committees, believes it is crucial to have mechanisms in place to detect and retract fake articles.
"But it's also crucial to foster strong research cultures to prevent these kinds of issues from happening, ensuring that researchers themselves understand that this is not in line with good scientific practice, she said in an interview with Khrono in 2022 (link in Norwegian).
She believes the problem partly stems from the publication pressure in the research system, which contributes to questionable practices.
———
Translated by Alette Bjordal Gjellesvik
Read the Norwegian version of this article on forskning.no
Subscribe to our newsletter
The latest news from Science Norway, sent twice a week and completely free.
