Few of us have the time or expertise to sift through all of the scientific papers published every day to determine which research is important and relevant to our lives.
In this sense, science journalists play a valuable role – selecting and publicising research that’s of interest and importance to the wider community.
But with deliberate scientific fraud on the rise and with a number of media reports about less-than-sound science having been published recently, perhaps it’s time to introduce some guidelines for science reporting.
The Japanese stem cell debacle
One of the most pertinent examples of bad science reporting was a newspaper article, published earlier this month, about stem cell research.
The article, published in the English edition of the Japanese newspaper Yomiuri Shimbun on October 12, reported that induced pluripotent stem cells (often abbreviated iPS stem cells) had been used to successfully treat a person with terminal heart failure.
According to the front-page article, the patient was healthy eight months after being treated by Japanese researcher Hisashi Moriguchi at Harvard University. The newspaper declared this to be the “first clinical application of iPS cells”.
Sadly, the report began to unravel almost immediately.
A Harvard Medical School spokesman, when contacted by Nature, denied any such procedure had taken place, emphasising:
“No clinical trials related to Dr. Moriguchi’s work have been approved by institutional review boards at either Harvard University or MGH [Massachusetts General Hospital].”
A Harvard communications officer, when contacted by Science, said Moriguchi was a visiting fellow at MGH for two months in 1999, but that “he has not been associated with MGH or Harvard since that time.”
Cause to be cautious
In hindsight, there were certainly reasons to be suspicious of the newspaper report. Moriguchi claimed to have reprogrammed stem cells using just two specific chemicals (microRNA-145 inhibitor and TGF-β ligand, in case you’re wondering), but stem cell researcher Hiromitsu Nakauchi responded saying he had “never heard of success with that method”.
What’s more, he’d never even heard of Moriguchi until his study was published.
The article in which Moriguchi presented his “research” also included several counts of plagiarism, most notably several paragraphs lifted straight from a 2007 research paper by the joint winner of the 2012 Nobel Prize in Medicine, stem cell researcher Shinya Yamanaka.
In short, the reported research is clearly bogus. Its prominent announcement on the front page of Yomiuri Shimbun is both a major embarrassment to the newspaper and the other news organs who blindly picked up the story.
Perhaps more importantly, it’s a source of confusion to readers internationally.
“Due diligence” in science reporting
The Moriguchi case is but one example of bad science being reported in the media. So how can such lapses be prevented in future?
Ideally, all science, finance and medical journalists would be fully accredited, as is the case in many other professions. But this is clearly unrealistic. Instead, we propose a set of standards for “due diligence” in scientific journalism.
Namely, we propose a minimal set of procedures that science journalists should follow before publishing any report on scientific research:
Contact each of the institutions claimed by the lead researcher, at least, to verify the story.
Virtually all reputable institutions have press representatives whose responsibility it is to be aware of recent work (or, if they are not, to contact the researcher to learn specifics).
Verify the work has been published by (not just submitted to) a reputable peer-reviewed journal or conference proceeding. In most cases this can be done by checking the journal’s or conference’s website.
From time to time news outlets report, often informally, on research that has not yet been published, but such announcements should very clearly state that the work is yet to be peer-reviewed.
Contact a reputable researcher in the same field who was not a co-author of the study in question and include, in the report, the name of the researcher and some brief comments.
This step is particularly important if the research in question purports to be a significant breakthrough in the field, or if it may be considered controversial.
After the article is written, send a draft to the researcher and to each other person mentioned centrally in the article, with a reasonable opportunity for them to assist with technical wording and fact-checking.
This is often difficult to do given the constraints of deadlines, but it is an important way to prevent unintentionally confusing or misleading comments in the article.
All journalists want to be the first to release a report of some major development, not only in science but in every other topic normally covered by the news media, from politics to football. Carefully following rules such as the ones we have proposed will require precious time and effort.
… but it really matters
The field of science is unique in that the public, by necessity, places a great deal of trust in its practitioners.
In addition, many of the day-to-day details of scientific research are too technical to be fully appreciated, even by members of the public that have significant scientific training.
Thus, both scientists and the public rely on accurate and informative journalism, summarising what is being done in these fields.
With the media’s important gate-keeper role comes an equally important responsibility to perform the “due diligence” required. Otherwise, the public’s confidence in the scientific world will be further eroded, and scientists themselves will be unable to fully understand what is being done outside their own specialities.
The Moriguchi fraud was easy enough to detect and substantiate that by October 19th – just seven days after the Yomiuri Shimbun article was published – the University of Tokyo had already fired Moriguchi (who, incidentally, still claims to be innocent!)
As David Cyranoski wrote in Nature on October 16:
“rarely has such a spectacular scientific claim been debunked so rapidly.”
It could have been prevented with little extra effort.
This is an edited version of an article that originally appeared at Math Drudge.