A Ph.D. student found a fundamental flaw in research published in 2014, and upon which she was building her research, which led to a retraction by the original author.
Susanne Stoll, at the Department of Experimental Psychology at University College London, said she thought she had made a mistake when trying to replicate the results of previous work in brain research.
“When I first stumbled upon the error, I had no clue what I was dealing with. I thought I had accidentally messed up my experiment because I obtained similar results in fairly distinct experimental conditions,” Stoll said in emails with VOA Student Union. “There was no extraordinary feeling I experienced. Errors are everyday business in science after all.”
Stoll tried to duplicate the results and checked the programming codes to determine where she had erred.
“The idea here is that if the analysis pipeline works the way it should, an analysis on random data should not produce the type of result I observed,” she said of research written about in a paper by Benjamin de Haas and colleagues while at the Institute of Cognitive Neuroscience at University College London in 2014.
“However, when I ran the analysis on random data, I obtained similar results as with the original data,” Stoll said.
After consulting colleagues, they determined she had revealed an error in the original research. They encouraged her to explore and correct it. When she reached out to de Haas, he was eager to help Stoll correct what they determined was an oversight.
He “was as keen as everyone else to figure things out despite knowing that he might need to retract his paper,” Stoll said. The process to reexamine one’s work is tedious, she said, but necessary to correct the error, which other papers reprinted.
“We apologize to the scientific community for any inconvenience caused...Finally, we would like to thank our colleague Susanne Stoll, who first pointed out the problem to us and plans to publish in due course a more general exposition on the difficulties of this approach,” they wrote.
Mistakes are difficult for many professionals to admit. Retractions in science and medicine are not looked upon favorably, but they are necessary to clear the record about the process, accuracy and validity of medical and scientific research.
And they occur more often than most people realize.
Retraction Watch first published Stoll’s story. The staff has identified dozens of retractions of published results regarding the worldwide outbreak of the coronavirus responsible for the COVID-19 disease. The ailment has caused more than 1.8 million deaths worldwide, including more than 357,000 in the United States, according to the Johns Hopkins University.
Some of the journals that have seen papers retracted include The New England Journal of Medicine, The Lancet, Annals of Internal Medicine, Journal of Public Health, Asian Journal of Psychiatry, Korean Journal of Anesthesiology and Chinese Journal of Epidemiology.
Retraction Watch is a function of the Center for Scientific Integrity, whose mission is “to promote transparency and integrity in science and scientific publishing, and to disseminate best practices and increase efficiency in science.”
Who funded the research?
In addition to protecting the integrity of science, Retraction Watch sheds light on the funding process.
“Most retractions live in obscurity in Medline and other databases. That means those who funded the retracted research — often taxpayers — aren’t particularly likely to find out about them,” it says in answers in its frequently asked questions section. “Nor are investors always likely to hear about retractions on basic science papers whose findings may have formed the basis for companies into which they pour dollars.”
Sciencemag.org quoted Retraction Watch in a 2018 article that said retractions had increased tenfold between 2000 and 2014, likely because of increased editorial oversight. Still, 60% of retractions are because of fraud, Sciencemag.org said.
Fraud occurs when a researcher asserts information that has not been scientifically proven or vetted. One famous case, according to the science journal Nature, is that of Yoshihiro Sato, who for nearly 20 years "fabricated data and forged authorships — prompting retractions of more than 60 studies," it reported its website in 2019.
“There is certainly a stigma attached to retractions, which are often exclusively associated with fraud,” Stoll said. “However, both an oversight and fraud can lead to retraction. It is important to distinguish these cases, and I hope that our case encourages other researchers to correct their honest mistakes and that this moves science forward.”