A new analysis has affirmed what many in the science audience already knew; mainstream media prefer weak observational studies. It's why you're reading this article here instead of the New York Times.
And that is not just in regards to social psychology correlations made using surveys of college students or sociology mysticism, it happens in medical coverage too. The examination found that observational studies get far better coverage than actual randomized controlled trials, which are what should really be important to most people.
The authors compared the coverage of 75 medical/clinical journal articles in the top five newspapers (The Wall Street Journal, USA Today, The New York Times, Los Angeles Times, San Jose Mercury News)(1) against 75 articles in the top five medical journals by impact factor during the same period (ending 2011). They found that the preference for observational studies over clinical trials was clear; 75% versus 47%.
That's no surprise, a lot of clinical trial papers will be too boring or incremental to get press coverage - journalism costs money and editors know they need to provide bang for the buck. What was a concern was that newspaper journalists and editors seemed to really prefer studies that had weak methodology. So newspaper headlines like “Statin-takers Less Likely to Die from Cancer” failed to ever note that they were drawing dubious causal conclusions from observational data. That sort of thing is fun for social sciences but in medical coverage it can be dangerous - and with limited space, it means the public isn't getting coverage of real medicine.
Distribution of study design ratings for clinical investigations from the media and medical journals. (A) The media covers inferior quality study designs than those published in (B) high impact medical journals, p = 0.003. doi:10.1371/journal.pone.0085355
And, the authors say, it is systemic bias. Not bias as in that newspapers prefer one particular type of article but that they prefer things that make for catchy headlines, even if the methodology is suspect. They note the obvious impact; after media coverage of invasive group A streptococcal (GAS) disease, testing by worried parents for “the flesh eating bacteria” went up. We know what that does to health care costs.
It isn't just on the public side. What happens in the research community when observational studies get a lot of coverage? More people go into the area and that means more taxpayer money spent on studying Acai berries and gluten-free foods and whatever else makes the LA Times science and health section this month.
Obviously it's a tricky balance; flesh-eating bacteria is likely more interesting than lots of other studies that are released every day. But it would also help if journalists read more than press releases.
Citation: Selvaraj S, Borkar DS, Prasad V (2014) Media Coverage of Medical Journals: Do the Best Articles Make the News? PLoS ONE 9(1): e85355. doi:10.1371/journal.pone.0085355
(1) PLOS One insists its 10,000 articles per year are peer reviewed but editing, much less peer review, would have caught the mistake in calling one of those top 5 newspapers, the San Jose Mercury News, by the wrong name, the San Jose Mercury Times. At least they didn't also list the New York Today and USA Journal.
In a journal article about suspect coverage of journal articles, this is irony.
- PHYSICAL SCIENCES
- EARTH SCIENCES
- LIFE SCIENCES
- SOCIAL SCIENCES
Subscribe to the newsletter
Stay in touch with the scientific world!
Know Science And Want To Write?
- Part I: Bee Deaths Mystery Solved? Neonicotinoids (Neonics) May Actually Help Bee Health
- On integrating out short-distance physics
- Is The Micropower Revolution Here?
- Chronic Pain Isn't All In The Brain
- Big Data Could Be A Big Problem For Workplace Discrimination Law
- Violence, Sex And Taboo: The Original Brothers Grimm Fairy Tales Back In Print
- Like Collaboration And Intelligence In Humans? Thank War
- "Could be, but we have continuity of the fossil record with later life forms evolving from earlier..."
- "I think they have all been corrected by now - in the next blog post there is a link to the updated..."
- "There are fews other typographical details (as the title 3.2 vs 3.3 : 5-sigma vs 5$\sigma$)...I..."
- "Yes, I agree Luigi - it was a typo, I replaced f(x)dx with f(t)dt in eq. 1Cheers,T...."
- "nice visit http://www.clubsideeffects.com..."
- Study reveals significantly increased risk of stillbirth in males
- Moderate coffee consumption may lower the risk of Alzheimer's disease by up to 20 percent
- New research supporting stroke rehabilitation
- The artificial pancreas shown to improve the treatment of type 1 diabetes
- The Lancet: Leading medical experts call for an end to UK postcode lottery for liver disease treatment and detection