Stardate 20011009.0733 (On Screen): Dean sends in this link to an article that claims that two researchers have shown a correlation between violence on TV and violence in the world, along with a claim that the story is being suppressed. I'm not sympathetic. Yet again it needs to be said: "Correlation does not demonstrate causation." It's
Post Hoc fallacy, yet again. (This keeps popping up in these kinds of studies.)
"What we are finding is that when people watch a program with violence or sex, they think about violence and sex," said Bushman. "The sex and violence registers much more strongly than the messages the advertisers are hoping to deliver."
Or that when people are thinking about violence and sex they decide to watch TV shows about it as a safe outlet for those feelings. (By the way, why did "sex" creep into there?) They claim that their correlation shows that more TV violence begets more violence in the real world, despite such things as the fact that violent crime in the US has declined for the last few years. Suppose, just suppose, that for some unidentified reason there is more of a tendency for Americans now to get violent feelings -- but that most of them are watching violent television shows as a vicarious way of getting rid of those feelings. In that case, decreasing television violence would increase real-world violence as those people no longer had a safe outlet for those feelings. It's not that I consider this likely, it's that they haven't disproved this with their study. Their conclusion may be true, it's just that their evidence doesn't prove it. (discuss)
