People are not dumb. Even if – or maybe especially if – news stories don’t point out the limitations of observational studies and the fact that they can’t establish cause-and-effect, many readers seem to get it.
Here are some of the online user comments in response to a CNN.com story that is headlined, “Coffee may cut risk for some cancers.”
* “i love how an article starts with something positive and then slowly becomes a little gloomy. so is it good or not? i’m still where i was with coffee, it’s all in moderation, it ain’t gonna solve your health woes.”
* “The statistics book in a class I’m taking right know uses coffee as an example of statistics run amok. It seems coffee has caused all the cancers and cures them at the same time.”
* “Could it be that instead of having mysterious compounds, coffee drinkers just drink more coffee than they drink alcohol or smoke?”
* “I am so f-ing sick of these studies, or more precicesly how these “risk factors” are interpreted as “facts” by newspaper headlines. If you can’t explain why something happens other than surmising, stop wasting our time.”
* “…correlation IS NOT causation!!!! So people that drink 4 or more cups of coffee have a lower incidence of two certain types of head and neck cancers, and this is supposed to mean that coffee is actually “warding off” these cancers???”
We reviewed the CNN story and a WebMD story – both inaccurate because of the inappropriate use of terms like “benefit…lowers risk…protective effect” – when none of these can be proven by the kind of study they were reporting on.
And in both we referred the journalists and news consumers to a primer on our site, “Does The Language Fit The Evidence? – Association Versus Causation,” that explains why it’s important to get this right.