Scientific journals may have believed that they had their fact checking down to a fine science, but lately, their credibility has come under fire.
The reliability of research publications has been called into serious question following Science’s confirmation earlier this month that South Korean scientist Dr. Hwang Woo Suk fabricated evidence published in the journal stating that he had cloned a human cell. Yale School of Medicine Dean Robert Alpern said he does not think this negative stigma will affect Yale researchers, but he said it might influence the way Americans perceive science abroad.
“I think fraud has always gone on — it’s a small minority of researchers, but it’s always been there,” Alpern said. “I think what this has done is probably unfairly create suspicion about science coming from Asia.”
But despite the skepticism, David Corcoran, assistant editor for the science desk of The New York Times, said journalists will continue to use science and medical journals as major sources of news.
“While the Hwang scandal is certainly embarrassing, it’s hard for a journal or any news outlet to guard against someone determined to pull off a hoax,” Corcoran said. “As for ensuring that our own reporting is accurate, we can’t run our own experiments, but we do make a practice of seeking comment from experts not involved with the research in question.”
Cell biology professor Carl Hashimoto said he does not think the controversy and any new policies instituted at various journals will affect Yale scientists in terms of getting their work published and accepted. But Hashimoto said he does think the Hwang incident has brought attention to an important issue.
“[The scandal has] made scientists more aware of the issue of data manipulation, what is legitimate and what is not, and of the critical need to teach this concept to students and trainees,” he said in an e-mail. “In fact, it is taught now as part of an ethics course that all students are required to take at the beginning of their graduate studies, as mandated by the [National Institutes of Health].”
While most Yale researchers are not directly feeling the effects of the scandal, some are working to prevent flawed results from being published in the first place. Immunology professor Ira Mellman, who is also the editor of the Journal of Cell Biology, developed a way to use Adobe Photoshop to examine whether images have been manipulated. After publicizing the system in the past several years, Mellman said that, ironically enough, Science was the only journal that decided to adopt it. Unfortunately, he said, they did so a little too late.
“We would have picked up on at least one of the anomalies in [Hwang’s] paper,” he said.
Mellman said that even the 1 percent rate of data manipulation that he and Executive Editor Michael Rossner found in the papers they have received since the system was developed in 2002 is unacceptable.
“I found it kind of alarming that even that fraction was occurring,” he said. “We have to take steps as a community to deal with it.”
Mellman said that while fraud is a major issue for scientists, it is an even bigger one for journalists, and that this new level of skepticism should have been present in the journalistic community all along, especially when it comes to new research.
“When reporting on information that is brand new, a rather significant portion of it will change,” he said. “It’s incredibly important for the public to be educated about science, and journalists should not just treat science as acts of truth. It’s a developing story. Journalists shouldn’t oversell things.”
But not all scientists see it this way. Cell biology professor Peter Novick said that while tools for detecting fraud are important, journalists have to trust that the majority of scientists do not have such intentions.
“Science has to rely on the integrity of the scientists, with the understanding that sooner or later the truth comes out as people try to repeat the results,” he said in an e-mail. “This is why science moves forward, not because scientists are perfectly honest or because the mechanisms to detect dishonesty are perfect.”
Mellman said that although a case of the magnitude of Hwang’s only occurs once every 10 to 15 years, it should serve as a lesson for reporters to seek further information before publishing anything as truth.
Some researchers said that the pressure by the scientific community to publish completely original papers can also contribute to the manipulation or outright falsification of data. Brian Wayda ’07, who serves as editor in chief of Yale Scientific, said scientists should put greater emphasis on replicating and confirming previous work.
“When the mindset is ‘publish or perish,’ there’s little incentive or recognition given to researchers for merely confirming someone else’s results,” he said. “From my observation, Yale professors bring a healthy skepticism when looking at published studies in their fields. The problem is you don’t win fellowships and acclaim for verifying other scientists’ work.”
Hashimoto said the questions that have come up as a result of the Hwang scandal are important because they will help define scientific ethics for future generations of scientists.
“I do think that journalists should be generally critical of what they read in scientific publications, just as scientists are,” he said. “In the end, the whole scientific enterprise hinges on the personal integrity and honesty of the scientist doing the experiment, so it is most important that we teach students the proper way of doing science.”