Slashdot Mirror
TV Really Might Cause Autism
Alien54 writes "Cornell University researchers are reporting what appears to be a statistically significant relationship between autism rates and television watching by children under the age of 3. The researchers studied autism incidence in California, Oregon, Pennsylvania, and Washington state. They found that as cable television became common in California and Pennsylvania beginning around 1980, childhood autism rose more in the counties that had cable than in the counties that did not. They further found that in all the Western states, the more time toddlers spent in front of the television, the more likely they were to exhibit symptoms of autism disorders. The Cornell study represents a potential bombshell in the autism debate."
Everyone, let's say this together. Come on, I want everyone to join in. Let's all yell it at the top of our lungs until the media hears us. Ready? Here we go:
CORRELATION DOES NOT EQUAL CAUSATION!
CORRELATION DOES NOT EQUAL CAUSATION!
CORRELATION DOES NOT EQUAL CAUSATION!
Now stop reporting on every correlation between disease X and social variable Y as though it were somehow equivalent to a randomized double-blind study on the effects of Y on X. Thank you.
I call bullshit.
That is, it is entirely possible (and plausible) that a correlation exists. However I'd interpret it in the reverse way. That is, the study shows just that children born with autism are more likely to spend time watching TV (knowing the features of autism, this is entirely possible).
Moreover, the existence of a correlation does not show necessarily a cause-effect relationship. Do you remember Lisa Simpson showing Homer a rock that protected from tigers?
This kind of papers are what my collegues call "scientific pornography" -papers thrown up just to stir up controversy, but based on very fragile assumptions and with a few data inflated as much as possible. Quite a common occurrence, sadly, these days.
-- Patent no.123456: A way to personalize
I gave up on TV years ago
No, you didn't. You still watch TV, you just use different hardware. This has some advantages (fewer ads, less 'TV network' crap), but it's still TV.
Okay, that's a little petty, but here's a real difference - people without TVs choose what to watch, when to watch it. People with TVs so often just sit down in front of it and then vegitate, accepting whatever crap is shovelled to them. I'm sure there are plenty of people out there (yourself included?) who will then jump up and shout "But I only turn it on when there's something I really want to watch!" The difference is that people without TVs choose to watch less - they are generally more selective. Speaking for myself, I have a lot of fun times with my wife and daughter - playing games, reading books, talking. I kind of feel sorry for people who miss out on that sort of thing because "There's something really good on TV".
Oh arse
Before anybody starts jeering stupidly and making wise about this subject, perhaps people should read this article from Scientific American: http://www.sciam.com/article.cfm?chanID=sa006&colI D=1&articleID=000B7F38-893D-152E-88E283414B7F0000
Now for some of the usual comments people tend to spew out:
Correlation is not causation
This is true - but correlation indicates that there MAY BE a causation. Thus, when things are strongly correlated and there are other reasons to suspect a causal connection, it is well worth researching further.
Increased awareness
Perhaps 'increased awareness' of autism means that we discover more cases that were not previously recognised? Perhaps, but I don't think it is very likely. Full-blown autism is not something you overlook. It is a serious disorder that in most cases means lifelong disability, and it is unlike any other psychiatric disorder. The increased awareness, I suspect, mostly means that now we spot more of the milder cases, but it is not my impression that this is what this research is about.
So why is it that people on this list are hostile to the idea that maybe TV can contribute to the emergence of autism? My guess is that this is because people on the list tend to be heavy consumers of passive entertainment, like TV and computer games; you don't want to hear that it may be bad for you.
If you have read the article I referred to above, you will know that autism probably has a lot to do with the development of 'mirror neurons' in the brain; a neural system that makes us able to imitate what other people do. Like all neural systems, the mirror neurons need to be trained, and TV is probably not a very good role model for that, at least not if you are already weak in this area. So it is actually quite reasonable to suspect that watching too much TV at an early age may contribute to the development of autism.
In Canada we have warning labels on cigarette packs. Big warning labels. Cigarettes cause cancer, etc. So, naturally, some dollar store entrepreneur creates fake warning labels.
;-)
... it's obviously not a perfect analogy, but I've been debugging way too long to care]
e -against-the-mighty-machines-day-9-of-no-tv.
Anyway, when I was a stereotypical angry young philosophy student, I thought it would be fun to make my own fake warning labels to put on my cigarette packs. So, who did I turn to? Hume, of course.
So, my cigarette packs had a big warning: "Correlation Does Not Imply Causation" on them. I thought it was a good joke, by philosophy joke standards anyway.
Now, I knew perfectly well that in this case even though it did not imply it, it was in fact true. Of course cigarretes caused cancer. In many cases correlation is, um, correlated with causation. But I was 18 so I didn't care; I thought it was funny.
What a joke.
So, the point is: correlation is a start. If there is a correlation, you should look for ways to establish whether causation exists or not.
Now, you usually cannot do real proper experiments on humans with smoking (starting with a large random set of non-smokers, making half of them smoke their entire lives, and seeing how many of each group died of cancer). The ethics boards at the university wouldn't approve
So, do you just give up and say "thank you for smoking" or "well, we'll never prove anything according to David Hume then". No, you don't. There are statistical tools like factor analysis which let people smarter than me figure out how much of A is (probably) caused by B, etc.
Anyway, I have a 2 year old son now, and stuff I thought was funny at 18 is certainly not funny anymore over a decade later. I quit smoking. I certainly wouldn't give my son a cigarette, ever.
However, if there is a strong correlation between TV and autism, I have to wonder whether I am in effect doing something similar. What if further anaylsis proves (as much as you can prove anything) that it is indeed a cause?
What would I have done??
[yeah, yeah, there's a mountain of evidence in one of the cases vs. one study in another
"Correlation Does Not Imply Causation" does not mean act insanely. You have only ever seen gravity by correlation, but you still believe in it. (Yes you do. Wipe that smirk off your face.)
Now, coincidentally, I also cancelled cable TV after reading Gregg Easterbrook's original Slate article. Obligatory blog whoring: I blogged about it at http://peterarmstrong.com/articles/2006/10/08/rag
Do I think there is conclusive, Hume-would-be-proud proof that TV causes autism. No.
Do I think that TV is good for young children?
Would I give my 2 year old son a cigarette?
As a pediatrician trained in child development (and the parent of an autistic teenager), I've got a strong interest and background in this, and I can tell you quite plainly that the paper is crap.
This is a spectacularly good example of really stupid statistical games. I only skimmed it (Acrobat Reader blew up on me as I tried to save it, and I'll get another copy later), but these people did the following amazing things:
1. Accept as fact that autism itself is increasing (as opposed to the diagnosis of autism). This is possible, but contentious and somewhat controversial. I'll spare you the full story, but the general opinion is that while the disorder is more common than it once was, changes in diagnosis (and benefits for diagnosis) make it hard to do more than guess at the actual rate of increase.
2. Consider de novo a hypothesis "that early childhood television watching is an important trigger for the onset of autism." They do note that nobody else has bothered to consider this, but don't spend much time wondering why. Apparently, they're special. Perhaps because nobody has measured this in a useful manner? They do admit this, but they find a solution!
3. Because there are no good numbers for early television watching, they use precipitation as a proxy for television watching. Apparently, if it rains, you're likely inside with the tube on. They do show a strong positive correlation between rainfall and autism. Yep, that's right - rain causes autism.
4. But wait - it can't be the rain, it has to be the television! That's what we started trying to prove, anyway, so it's important to stay focused. They try it another way: they consider the availability of cable. They show that autism correlates with the availability of cable. No, really, it does. Of course, diagnosis of a LOT of chronic developmental syndromes increases with affluence, because of the increased availability of medical care and the reduction in "grab-bag" diagnoses like "mental retardation". But still, it must be the cable.
5. Having neatly done all the "proof" they require, they then proceed to tear the numbers apart and "prove" that 40% of autism in California is triggered by early television watching, while only 17% is triggered in PA. Why, we don't know, but it appears that rain, or cable, or maybe just TV is more powerful in CA than in PA. Or something like that.
I don't have time for a complete fisking right now, but I may do it later. Aside from the basic methodologic errors (confusing correlation with causation, adopting a highly questionable proxy indicator without validating it, and spending almost no time ruling out confounding factors or tainted data), there remain the dozens of smaller tactical problems that should have sidelined this turkey. I assume the peer reviewers, if there are any, were on drugs.
This paper will be a bombshell, all right. I'll use it over and over again as I explain to medical students and colleagues that you don't have to have much in the way of actual brains to write a scientific paper. Or, as I said about another paper in journal club once, "the font is nice, and I like the layout of the tables. It's a shame the actual science is such garbage."