In spite of the best efforts of statistics faculty everywhere, people constantly want to infer causality from correlations. Even though we can point to ridiculous examples, such as the strong positive correlation between ice cream sales and the murder rate (both increase in warm weather), some people just don’t get it.

Now we have what may be one of the worst contributions to our understanding of autism, a paper entitled “Does Television Cause Autism?” Good grief. The authors claim that autism rates are rising at the same time access to cable TV is increasing.

Where do I start?

First of all, the authors are professors of management and economics at Cornell with no expertise whatsoever in mental health. The lead author told Slate magazine, “I asked around and found that medical researchers were not working on this, so accepted that I should research it myself.” You know, sometimes there are very good reasons why medical researchers are NOT working on something. For instance, they know something about autism, the timing of its onset (most people assume it is present in many cases from birth), its symptoms, and some of its more likely causes. TV is not on this list for some very sound scientific reasons.

Second, most researchers believe that recent increases in autism rates have more to do with a loosening of diagnostic criteria than with a real increase in prevalence. The jury is still out on this one, as some environmental factors, such as mercury emissions, are suspected as having a possible role.

I am not a fan of network television, and I think that parking young children in front of the TV is a huge mistake. TV can have lots of negative effects on kids, but most relate to TV time being used as a substitute for things children SHOULD be doing, like playing, exploring, and interacting with parents, siblings and peers. The American Academy of Pediatrics recommends limiting children’s television exposure to 1-2 hours per day as a maximum.

We still have more questions than answers about autism. Lots of parents are desperate for answers. I think it’s highly unlikely that the answer will be in the amount of television children watch.

 


6 Comments

MunkeyChowFan · October 20, 2006 at 1:31 pm

It’s surprising to me that a prestigious University like Cornell would even allow profs in managament and economics make such a leap across disciplines. How did their research get approved in the first place? Wouldn’t it have to be approved by the APA as well since it involves autism?

Laura Freberg · October 20, 2006 at 3:30 pm

As far as I know, the Waldman et al. study has not been published. They will be presenting their paper today (Friday) at a National Burea of Economics conference. Normally, papers must undergo peer review before publication, not APA approval per se. Whatever the editors and reviews of a publication deem appropriate will be published. Why this study has received any serious attention at all remains a mystery to me.

cableguy · October 22, 2006 at 3:25 pm

It sounds like similar issue to the increase in diagnosis for ADHD: “a loosening of the diagnostic criteria”. Nicely worded.

emhughes · October 30, 2006 at 11:36 pm

What a pecuiler correlation. Did these “management and economic” professors not read the latest studies that suggest Autism is present at birth (as cited by you)? Why would they spend so much time on a study not in their related field?

tomkielymft · December 17, 2006 at 2:16 am

I’m not sure the MIND Institute would support any correlation in this study. And I think it’s fair to say that the Combating Autism Act (that was recently signed off by The House of Representatives and our US Senate and awaiting the President’s sign off YAY!) will not be funding any future studies done by this lab.

With regard to study/research/treatment for Autistic Spetrum Disorders, sadly this is not the craziest thing I’ve seen.

Encephalon, issue 9 « Migrations · October 22, 2006 at 7:25 pm

[…] Laura, on her Psychology blog, also takes a look this paper on television and autism. Laura takes a rather critical view, and suggests that not only is the evidence weak, but there’s several reasons why the basic tenets of the study are fundamentally flawed. (selected by the Neurophilosopher) […]

Comments are closed.