I recently came across this archetypal bad science news headline: "Drinking Sugar-Sweetened Beverages Daily Linked to Diabetes." Although science news outlets like ScienceDaily intend to provide scientific pabulum to amateurs and non-scientists interested in current scientific discourse, the sensationalizing voice that many of their contributers take when reporting research findings can leave the masses intellectually anemic and perhaps more ignorant about what science is really like than before they started reading these articles.

The article itself is actually a rather well written news release put out by the American Heart Association (AHA) to summarize the presentations at its 50th Annual Conference on Cardiovascular Disease Epidemiology and Prevention. Alas, headlines by themselves matter, as the research firm Outsell Inc. elucidated in its most recent News Users' report, which claims (among many other findings cohesively valued at US $1,295.00 for full access to the report):

"a full 44 percent of visitors to Google News scan headlines without accessing newspapers� individual sites."

I assume the head honchos on the board of the AHA are more likely to blame for the headline than its doctorate degree-holding authors. But I see this kind of impudent headline all the time in science reporting. Real scientists (presumably like the ones whose research is discussed in the previous link; here is the title of their actual journal article) don't make such strong claims about the the implications of their research--at least, not when they're writing for any audience who isn't signing their checks. Instead, the convergence of evidence from their research and many others allows us to evaluate the body of knowledge to make a claim like, "drinking sugar-sweetened beverages daily linked to diabetes." And to be fair, the AHA conference was the type of place where people do just that.

But it's not just the conceited headlines like this that bother me. Headlines that seem to reiterate the obvious as if it were novel have a certain bayesian effect on the public perception of science: the more frequently "science" articles tell us something we think that we already know, the less likely we are to value "science" as a valuble source for information about the natural world. If an editor or headline writer believes the "average" reader might not already assume the results advertised in the headline, then at least explicitly address the well-established belief that it affirms, in this case, that for years we've known that increased in sugar consumption increases the risk of diabetes, and recent research further supports that belief.

I understand the need to make headlines catchy and concise enough for hurried readers to get drawn in. As it is, ScienceDaily headlines are fairly long in my RSS reader. But I find NY Times science headlines to be much more appropriate and specific than the ScienceDaily headlines. This may be because ScienceDaily aggregates its stories from many sources, resulting in a less sagacious news source than a newspaper (Hopefully! I don't know if that comparison is true for my hometown newspaper, the San Francisco Chronicle).

I'm thankful for the freedom of information and the broad access to decent science news that websites like ScienceDaily give me, but there's something to be said for discretionary editorialship, which is best provided by people who are employed exclusively for this purpose. Look for my entry tomorrow (link forthcoming) that discusses the importance of preserving professional journalism and the effect of the blogeneration on our perspective about news. Until then, check out Bad Science, It's like the more professional and frequently updated version of FogMag!

Published on May 4, 2010 in Science