Never Trust a Headline: I read enough science articles on the Web to know that headlines are more often than not misleading. And I know why that is, because in our increasingly noisy travels through the world of media, our attention can only be grabbed by big surprising statements. So "Scientists think that maybe" is never going to get as many readers as "Studies show!!!" Reading the article would often clear up the hyperbole, but gads, who reads articles anymore? You take in the headline and the first sentence or two on your RSS reader and then hit the share button. (I am entirely guilty of this. Hey, I've got Tweets and blogs to feed.) So we never get to the "well, really, it's just a couple of nerds talking over lunch, but it could be true!" part, and if a PR person is doing the writing, that part might not even get written. This week's case in point, as dissected by Emily Willingham on The Biology Files: Autism and type 2 diabetes linked because "both are increasing"? (I very much want a study done, or at least a series of alarmist articles, on the correlation between autism and basketball-shorts length proposed by a commenter on that post.)