In response to this, Carrie and Michael had an interesting exchange on possible junk science.

First, Carrie linked to these two studies with evidence linking toddler TV viewing to attention and learning disorders:

Christakis DA, Zimmerman FJ, DiGiuseppe DL, McCarty CA. Early television exposure and subsequent attentional problems in children. Pediatrics. 2004;113:708-713.

TV Has Negative Impact on Very Young Children’s Learning Abilities (Arch Pediatr Adolesc Med. 2005;159:619-625)

Michael then responded,

See–this is why I need to write that book. Those two papers you cite are terrible–exhibit A for bad research. Seriously.

It’s very difficult to infer causal relationships from observational data like that. As you know, there could be a lot of confounding factors. But it’s difficult but not impossible. One can do a range of things, such as including more variables in the analysis, adjusting for their impact in a flexible way, and so on.

In the Pediatrics paper, we took the same data and found that the effect of TV was very sensitivity to the addition of even a few covariates. And the ones we added should have been included in the first place. We also found that the effect of TV was only at extremely high levels (>10 hours / day). It’s hard to believe that those “effects” were really effects–one has to wonder about those families letting kids watch that much TV being just different.

Now, one might say, this paper went through peer review at a good journal–isn’t that evidence of quality? The problem is that reviewers often share the same biases as authors. It’s also hard to get methodologists to evaluate papers for these applied journals . And the American Academy of Pediatrics has a policy statement urging parents not to let their little kids watch much TV. So, I think they give papers like this the benefit of the doubt in the review process.

We’re writing a paper offering counter-analyses, but we had an earlier version that was more of a direct critique. Pediatrics rejected it. Of course, one of the reviewers was an author of the original paper. I’m not sure they were quite objective. I’ve attached an earlier version of our response. We’ve extended the analysis still further, and we just don’t find any effects. It’s probably overkill. The whole thing is just a myth.

Of course, there’s the myth that living in single-parent family is bad for a child. 8-) Just published a paper on that, too. Of course, the press ignored it.

Michael’s papers are here and here.