DISCLAIMER: I can't cite the diabetes study, as I read the Times article and lost it, but the study on men and memory can be found at http://www.msnbc.msn.com/id/3032619/vp/22652909#22652909 . I queried the guy who did the study, but he never responded to me.

I would rather "professional" news reporters not report on science than get it so badly wrong. Example: NBC anchorman implied that men can't remember details in their emotional context and that women can't do anything but remember details with their emotional context.

Another example: the old saw about type 2 (insulin resistance) diabetes being linked to obesity continues to be part of the "standard" description of newsreporters. Science news reporters should not use "standard" descriptions without fact checking them. According to the New York Times, the recent studies have indicated that weight gain is often linked to diabetes-treating medications, a timeline ignored in many statistical studies. Also, there are a statistically significant number of people with low body fat who have type 2 diabetes. At the same time, other Times articles report on a strong link between obesity and diabetes, suggesting that reporters are summarizing results, making them a mouthpiece for the latest study, rather than individuals taking responsibility for the accuracy of what they write.

The same, of course, goes for smoking and lung cancer--a statistically significant number of people who smoke get lung cancer within twenty years (after twenty years, the risk drops tremendously for people who have quite smoking). However, a statistically significant number of people don't get lung cancer. Plus, with the rise in lung-cancer rates among non-smokers, these statistics are likely to change (this rise was cited by a physician).