TV & Violence: Strong Bond or Weak Link?

A new study reports more than one hour of television per day may make adolescents more prone to violence.

It's a conclusion that's about as silly as Saturday Night Live's immortal Emily Litella commenting about "violins on television."

Researchers from Columbia University and the New York State Psychiatric Institute "tracked" about 700 children from adolescence to adulthood. Correlations between adolescents who watched more than one hour of TV per day and violent acts committed in their young adulthood were reported in the journal Science.

"Our findings suggest that, at least during early adolescence, responsible parents should avoid permitting their children to watch more than one hour of television a day. That's where the vast majority of the increase in risk occurs," said researcher Jeffrey Johnson.

An editorial accompanying the Science study advocates "interventions for adolescents … because a heavy diet of media violence contributes to a societal violence rate that is unnecessarily obese."

I've reached a different conclusion, however: Science's editors should be ashamed for once again allowing the prestigious journal to be the mouthpiece for junk science-fueled political correctness.

The researchers acknowledged their study was the first to investigate the potential for TV viewing to have long-term effects on aggressive behavior.

One study, though, is not science. At best, it's a data point — and it's not clear this study even measures up to that standard.

The reported statistical associations are pretty weak.

According to the study, the rate of violence committed by 16- to 22-year-olds who watched more than an hour of TV per day at age 14 increased by 58 percent. The corresponding increase among 30-year-olds who watched more than an hour of TV per day at age 22 was 57 percent.

But the rule of thumb for such statistical studies is that increases on the order of 100 percent or less are viewed as statistical noise since they may readily be explained by poor quality data and chance.

Inexplicable inconsistencies among the results reinforce the notion that they are statistical nonsense.

Boys (16-22), but not girls (16-22), who watched TV for more than one hour at age 14 were statistically associated with increased violent acts. Thirty-year-old women, but not men, who watched TV for more than an hour per day at age 22 were associated with increased violent acts.

The researchers don't really know that TV viewing had anything at all to do with violent acts.

The researchers interviewed the families of the study subjects several times between 1975 and 1993. They interviewed the study subjects in 2000 about aggressive acts committed and collected data from law enforcement sources concerning arrests and charges for adult criminal behavior.

But the researchers didn’t collect data on whether the violent acts committed by the study subjects were motivated, inspired or otherwise linked directly or indirectly to television viewing. They hoped to accomplish this feat with statistical correlations alone.

Statistical correlations can be persuasive if very strong. But that's not the case here.

The researchers further seem to believe they can simply wave a statistical wand over their data to make alternative explanations for violent behavior — such as history of aggressive behavior, childhood neglect, family income, neighborhood violence, parental education, psychiatric disorders, and others — disappear.

The junk science-taint of the study is sealed by the overzealous and even somewhat paranoid comments in the editorial.

"Despite the consensus among experts, lay people do not seem to be getting the message from the popular press that media violence contributes to a more violent society," opined the Iowa State researchers who have a history of trying to link video games with violent behavior.

"This inaccurate reporting in the popular press may account for the continuing controversy long after the debate should have been over, much as the cigarette smoking/cancer controversy persisted long after the scientific community knew that smoking causes lung cancer."

Now I get it. TV viewing increases the risk of violent behavior because smoking increases the risk of lung cancer. And the only reason kids still watch so much TV is that the media is misleading parents!

The researchers buried a disclaimer at the end of their study, "It should be noted that a strong inference of causality cannot be made without conducting a controlled experiment, and we cannot rule out the possibility [of other explanations] for these associations."

Hmmm … now what would Emily Litella say about this study? Oh yeah … "Never mind."

Steven Milloy is the publisher of , an adjunct scholar at the Cato Institute and the author of Junk Science Judo: Self-defense Against Health Scares and Scams (Cato Institute, 2001).

Respond to the Writer