Over-hyping scientific findings often begins in research article abstracts, even before press releases and media outlets get their hands on them.thephotographymuse/Flickr
The scramble for attention in the scientific community -- increasingly through hits, likes, tweets, and the like -- perpetuates a cycle of increasingly over-hyped coverage of medical studies. That attention can have real effects in terms of grants, awards, promotions, etc., in professional realms. While the media receives a lot of flak (some of it which is certainly deserved) for incautious reporting, medical researchers themselves aren't immune to this tendency to sensationalize, either.
Bringing this to our attention is (what else but) a new study. Published in PLoS Medicine, it looks at the findings of randomized controlled studies of experimental treatments, as represented by the articles' abstracts, the press releases that promote them, and the ultimate media coverage.
40% of abstracts were less objective than the full-length articles they were meant to summarize.
Like other studies before it, this one attempts to identify where "spin" (in the way these stories are reported) come from. This bias towards positive and novel findings can occur when the benefits of the new treatment are overemphasized, or when statistically insignificant results are ignored. News media, the study found, contained misleading information that had at some point been spun to sound greater than the actual research results, about half the time.
In attempting to identify the original source of this spin, it could usually be traced back to the press releases that journalists often consult for their information. Beyond that, it was also found in abstracts of the journal articles themselves, 40 percent of which were less objective than the full-length articles they are meant to summarize, according to independent analyses. Ninety-three percent of press releases containing spin were linked to abstracts that also over-hyped findings; press releases about studies that utilized small sample sizes or that were published in specialized journals were also more likely to contain spin.
Every time either the abstract or the press release contained spin, it would carry over to the news coverage.
It's impossible to say whether the spin identified in this study was intentional or unintentional, especially when you consider the pressure put on everyone from researchers to academic journals to media outlets to promote research findings. A different study from a few years back in Annals of Internal Medicine found similar bias in press releases, specifically exaggerations of the findings' importance and under-emphasis of study limitations. This one looked at the press release processes at the highest and lowest-ranked academic medical centers, the heads of which all verified that "media coverage is an important measure of their success." Most also told the researchers that they report their media hit numbers to the administration.
Meanwhile, in Slate last week, Charles Seife looked into Jonah Lehrer's misdeeds back when he was a science blogger for Wired.com -- before he was called out for self-plagiarizing in The New Yorker -- and found a number of instances where he took quotes and phrases directly from press releases without attribution. While Seife acknowledges that many journalists do find this practice acceptable ("Press releases, after all, are meant to be copied"), he cautions that doing so can come perilously close to violating the readers' trust. "Press officers are thrilled when reporters copy their prose," he writes. "What better way for an institution to control a message than by putting words directly in a journalist's mouth?"
Trust and accountability are, of course, the most important issues here, and this study shows how their breakdown can have widespread ramifications. While journalists do well to meticulously evaluate full-length articles and contact multiple external sources, they might not have the time and resources to do so for every headline. Our initial decision to look further into a study usually comes from its press lease, and if a study isn't made publicly available, the best we can do to allow readers to judge it for themselves is to refer them to its abstract.
Everyone involved in spreading scientific and medical knowledge is accountable -- not only to their bottom line, but to the public. When spin is perpetuated, regardless of the source, people's very real expectations of new therapies or disease courses or medical recommendations are misinformed. When this happens, we all share in the guilt.