|
Current affairs
|
|
Better science reporting
Over on BMJ Blogs, BMJ investigations editor Deb Cohen discusses how to improve reporting of science and medicine in the media (see http://blogs.bmj.com/bmj/2011/07/22/deborah-cohen-on-improving-health-reporting/). It's a hot topic, and one near to my heart. As she points out, one important point is to establish how good the research actually is, and that's a question that isn't always asked.
At Best Health, the BMJ's patient information website, we always include a section where we ask the question: 'How reliable are the findings?' It means we always have to check that the findings are actually statistically significant, that the study was properly blinded, that we don't imply causality to an observational study. You can see our latest research stories on http://besthealth.bmj.com/x/set/news/latestnews. You need a few things to do this - experience, training in interpreting research, a suspicious frame of mind - but first you need the actual research paper. Amazingly, few journals actually send out the paper, or links to it, when they send out the press release. The BMJ does, of course, and big US journals like JAMA. But lots rely on you to request a copy, and when you have a tight deadline, that holds things up considerably. As a result many of the news stories that appear in the media are based on the press release, not on the actual study.
|
Recent Entries
Most Recent Tags
|
They need to understand that science cannot 'prove' anything, but make other things most unlikely, and that it is a grdaul, process of advancement by tiny steps that test a hypothesis and provide new data for a new one.
They love to fetaure the maverick, the rebel (read complete barking loony) becaus thye think it makes good news.
John