In which bad science negates good

On wednesday we all went off to the Royal Institution hoping to see the big fight. It turned out that the fight of the century -- medic, scientist, and Bad Science writer Bed Goldacre versus Paul Drayson, Baron Drayson of Kensington, Minister for Science & Innovation and Minister for Strategic Defence Procurement, on the topic of whether science and health reporting in the media is bad for you -- was a relatively dull and polite affair of concessions and clarifications. It was one of those "debates" that in a large part boils down to a difference in emphasis: for Ben, the daily flurry of embarrassingly ignorant science stories in the newspapers really grates, really does harm, and thus requires our vigilance, ridicule, and correction; for Drayson, it's the really well researched and well written pieces of science journalism that stand out, really educate the public and engage them, and thus require our praise.

It's a shame that there was so much concession and agreement, really, because that overshadowed the fact that, ultimately, Drayson was completely wrong: science journalism is bad for the reader's health. The most important point made during the evening was Ben's reaction to a headline from that day's Express, regarding a cancer drug. Drayson's reaction had been to have his office phone the scientists who did the work, and confirm that the reporting was accurate; Ben's reaction had been skepticism -- not active disbelief, just a total inability to know whether the story was worth believing. And, indeed, a study that he cited found that few people can know what to believe in the newspaper: they get the impression that scientists keep changing their minds -- one week scientists think that coffee causes cancer, and the next week they think it cures it.

So, sure we should praise and highlight good science journalism. But so long as newspapers are filled with so much of the truly appalling kind, the accurate ones are worth nothing: few readers have the background knowledge, access to the literature, or office staff at their command, that are necessary to know which stories they can believe. Pointing at one good science story is not sufficient to demonstrate that a newspaper can be trusted; pointing at a single ridiculously, childishly, fraudulently bad science story is enough to demonstrate that it can't be.