Or subscribe now to read with unlimited access for as low as $10/month.
A free account provides you access to a limited number of free articles each month, plus newsletters, job postings, salary data, and exclusive store discounts.
In the midst of a renewed national debate over fracking, a study last month from Colorado State University at Fort Collins and the University of Colorado at Boulder gave the public welcome assurances that the oil-extraction technology is safe.
The study, according to a summary prepared by the University of Colorado and repeated widely in the news, found the major group of chemicals used in fracking to be as safe as common consumer items like toothpaste and ice cream.
That should “give comfort” to anyone alarmed by the reported dangers of the controversial method, also known as hydraulic fracturing, the Colorado Oil & Gas Association said in a statement.
The problem, according to a lead author of the study: That’s not what the study actually found.
Instead, the research was an effort just to test a new method of chemically analyzing the fluids that are pumped into the ground during the fracking process, said the author, Thomas Borch, an associate professor of environmental chemistry and biogeochemistry at Colorado State. The study drew very limited conclusions, if any, about the actual safety of those chemicals, Mr. Borch said.
The stream of misleading media assurances about the fracking study is noteworthy. But according to a study published on Tuesday in Britain, it isn’t unique.
The British study examined 462 news releases on biomedical and health-related science topics issued in 2011 by 20 leading universities in Britain, along with the research papers and 668 news articles associated with those releases.
The analysis, published in the journal BMJ, was led by researchers at Cardiff University. The Cardiff team found problems with more than one-third of the news articles—exaggerations of advice flowing from the studies, unsubstantiated claims of cause and effect, and unwarranted inferences that animal-research findings could be applied to human beings.
In the overwhelming number of those incidents, the press releases describing the research contained the same types of exaggerations that turned up in the news. The chance that exaggerated advice would appear in a news article was 6.5 times higher when the press release contained exaggerated advice, 20 times higher in cases involving exaggerated causal claims, and 56 times higher with instances of exaggerated inferences to humans from animal research, the authors wrote in BMJ.
Previous studies have reached similar conclusions, in the United States and abroad. The problem reflects failures by multiple parties, including the journalists who write the news articles and the scientists who, in most cases, are required by their institutions to approve the press releases in advance, the Cardiff team wrote.
“The blame—if it can be meaningfully apportioned—lies mainly with the increasing culture of university competition and self-promotion, interacting with the increasing pressures on journalists to do more with less time,” said the authors of the BMJ study, led by two professors of psychology at Cardiff, Petroc Sumner and Christopher D. Chambers.
In an accompanying commentary, Ben Goldacre, a research fellow in epidemiology at the London School of Hygiene and Tropical Medicine who writes regularly about media coverage of science, suggested that universities and journals hold scientists more accountable for exaggerations in news releases about their work.
All news releases describing research findings should have named authors, including both the scientists and the press officers, Mr. Goldacre said. Just as scientists are expected to correct errors in journal articles, they should be expected to fix errors in media coverage, he said.
Hopes for such reforms got a boost this week with the announcement of the revival of financing for HealthNewsReview.org, a project operated by Gary J. Schwitzer, an adjunct associate professor of public health at the University of Minnesota in Minneapolis. The project, which analyzed the accuracy of news articles covering science, lost its foundation support in July 2013.
But this week it received a $1.3-million grant from the Laura and John Arnold Foundation that will finance its work for two years. As part of the grant, the project will be expanded to study both news articles and the press releases that precede them, Mr. Schwitzer said.
“Nothing could set the table better for what my new grant will allow us to do,” he said of the BMJ study.
The numbers found by the Cardiff researchers reflect a problem that is continuing and perhaps worsening as universities face growing financial pressures to bring attention to their research activities, Mr. Schwitzer said.
One of the most egregious recent examples, he said, involved a study from the University of Exeter that found that hydrogen sulfide can help strengthen mitochondria, which generate energy supplies in animal cells. The university’s news release described hydrogen sulfide as having the “smell of flatulence” and said that the health of mitochondria can be associated with various human diseases.
That led, Mr. Schwitzer said, to a series of news articles by major publications talking of medical cures associated with “fart gas,” including one in Time magazine that first appeared with a headline suggesting that “Smelling Farts Might Prevent Cancer.”
Last month’s Colorado fracking study also led to several misleading headlines, Mr. Borch said.
News outlets, including USA Today and The Washington Post, largely followed the theme set by a news release issued by the University of Colorado at Boulder. The headline of the release read: “Major class of fracking chemicals no more toxic than common household substances.” A subsequent headline in The Washington Times read: “Chemicals used in fracking deemed harmless, study shows.”
Mr. Borch said the University of Colorado had given him a copy of its proposed press release ahead of distribution, and he said he had urged his co-author at the Boulder institution, E. Michael Thurman, a research scientist, to revise it.
Beneath its headline, the release did state that the fracking chemicals deemed “no more toxic than common household substances” represented only a part of the mixture pumped into the ground to extract petroleum. But the emphasis on safety seemed unwarranted, given the scope of the research, and the release did not reflect possible additional dangers from the particular mix of those chemicals that is found in fracking fluids, Mr. Borch said.
Mr. Thurman declined to talk about the matter, referring questions to Laura Snider, the University of Colorado media representative who wrote the press release. Ms. Snider acknowledged that the university had “struggled over the headline a bit,” but she defended the release over all as a fair attempt to describe a complicated situation.
Another expert in science and the media, William Kovarik, a professor of environmental communications at Unity College, in Maine, called the widespread misreporting of the Colorado study a striking example of a nationwide lack of scientific literacy, driven heavily by extensive shifts in the news media, and he said he planned to make an example of it for his undergraduate classes.
Mr. Borch said he too regarded it as a lesson. For press releases issued by Colorado State University, Mr. Borch said he already makes a practice of working directly with the university’s press officers to ensure that each release fully reflects the findings of his research. “Yes, it takes work,” he said, “but I think that’s such an important process.”
One irony, said Mr. Sumner and Mr. Chambers, the Cardiff team’s leaders, is that their data suggested that exaggerations in the reporting of scientific findings in press releases did not produce any meaningful increase in the number of news articles.
But exploring that effect was not a main objective of their study, the authors said. That question will next be tested in a planned randomized trial of the effects of modifications in the style and content of press releases, they said.
Paul Basken covers university research and its intersection with government policy. He can be found on Twitter @pbasken or reached by email at paul.basken@chronicle.com.
