The study is linked to in the second paragraph. It appeared in what seems to be a reputable, peer-reviewed journal and controls for socioeconomic status.
I see this all the time online: a study is published, and a chorus of commenters echo "correlation isn't causation" without actually reading the study and seeing that these other factors have been accounted for.
There are plenty of things wrong with science reporting, but when many readers are apparently too lazy to exert the minimal effort to verify that their misgivings were indeed addressed by the researchers, one can't exclusively fault the media.
I have to disagree with both of you. The original criticism was incorrect, because this study does control for income and other socioeconomic stuff. But "correlation isn't causation" is still a valid criticism of this study. To prove causation you would have to have some random effect controlling where people live.
Just because it's peer-reviewed doesn't mean the article proves what it claims to. Unless you have a natural experiment it's hard to prove causation in situations like these. So you get a lot of published research that just proves correlation.
In passing, while playing with Google Trends, there seems to be a long term inverse correlation between media coverage and search volume for terms such as 'study' and 'research', as well as for the term 'statistics'.
http://www.google.com/trends?q=statistics,+study,+research&#...
The annual cycle seems to correlate strongly with college schedules :)
I see this all the time online: a study is published, and a chorus of commenters echo "correlation isn't causation" without actually reading the study and seeing that these other factors have been accounted for.
There are plenty of things wrong with science reporting, but when many readers are apparently too lazy to exert the minimal effort to verify that their misgivings were indeed addressed by the researchers, one can't exclusively fault the media.