Well the vaccine hypothesis didnlt quite work out, maybe this mud will stick.
I think effect left on the train before cause arrived. Correlation does not demonstrate causality, and there could be numerous reasons for the correlation. Autistic behavior also starts showing up before age 3, confounding television’s role in cuasing it. And let’s not even mention that there are about 20 genes that have been associated with autistic spectrum disorders, and that brain cell morphology changes have been seen before age three.
Vlad/Igor
There’s an excellent commentary in Slate’s Fraywatch. One comment in particular:
Hmmm . . . Christina Ricci was born in 1980 . . . Omigod: Christina Ricci causes autism!!
Although I certainly am happy to see doubt cast on this report, I’m disappointed to see that it was Gregg Easterbrook that put this out in the first place. I’m a big fan of his work, although it’s usually limited to Tuesday mornings.
It seems 1980 coincides with a change in the definition of autism and the confusion in the Press and general public as to difference between autism and ASD .
You left out some letters. It’s auteroticism she causes. The bitch.
Holy cow! Autoeroticism=autism+erotic!
This must mean something!
I went through the paper and I’m not impressed. They use precipitation as a proxy for adult television watching which they use as a proxy for child television watching; in addition, they acknowledge a genetic component to autism, yet they siblings in the same household as independent observations, thus skewing their results.
When they use cable television subscription rates as a confirming proxy for television watching, they use cable television under the theory that kids watch more cable; however, they specifically eliminate areas where broadcast telelvsion isn’t well received, and therefore include many homes that already had significant PBS children’s programming for decades before cable came to be an issue. In addition, I didn’t see any indication that the character of cable television networks was included in their research.
There are other problems with it, some being quite significant, IMO, and I see no sign that it’s been accepted for publication, which means that we can’t be sure it’s been properly vetted.
I hate to be a smartass, but if you follow all the links you’ll see that the “Cornell researchers” are two management grad students. Dunno about ya’ll, but I’m not sure a management degree is sufficient for this type of analysis. I realize a PhD in economics does probably qualify one to do statistical analysis, just don’t think it’s sufficient to determine cause and effect when dealing with a medical problem.
Statistics is the study of large numbers of things; being a doctor gives no advantage in statistical analysis.
Even if it isn’t cable they have other thing they could point to in the sequel papers.
Rubik’s Cubes - those things can screw up brains. I know they did mine. Imagine a three year old trying to solve one. Instant brain-fry
MTV specifically - it was really popular starting about then and they played some actual music videos and many of them were either wacked out or terrible. Second-hand exposure could easily cause a naive one to retreat into their own little world.
Any others?
Now we both know that has a nugget of truth to it but it isn’t quite right in the context of whole studies. There are people that are pure statisticians and they help researchers doing all kinds of things make sure there statistical techniques are used correctly. However, studies of this type should not be done through this type of data mining. That is, it is a very bad technique to just started running lots of statistics on large data sets to see what you can find. The inevitable problem is that there is going to be lots of false positive results and there is no way to know which are wrong if you go about it that way.
The causes of autism have various theoretical models that different researchers are using to explore the causes. If a body of knowledge suggests that we should look to cable TV and why that could cause autism, that is great. Research can head in that direction with clear goals in mind. If someone like these students just throws out a random correlation out there from nowhere, all you end up with is a factoid like in those trivia books we all love and it contributes very little or even negative amounts to the existing body of knowledge.
My youngest sister was born in 1980. Obviously, it’s all her fault.
That’s not what they intended to do; they’re not merely data mining, inasmuch as their reasons for thinking television might have a causal influence bears some validity. I don’t find them to be compelling; however, it remains that the intent isn’t to data mine, which is obvious from the paper.
It’s certainly true that a real, physiological explanation needs to be brought out before the claim can be considered proven; it is certainly true that such a study points to possible avenues of further research; and, it’s perfectly true that blaming television is at this point far, far too premature to be sensible. Indeed, my first post indicated that the study methods themselves are very flawed, so much so that it can be rejected out of hand, IMO. For example, if a mother is watching television while breast feeding, that will count as a child-watching-television data point.
Besides, there’s surely sufficient individual data to be had to go beyond the crude proxies they used. It’s a shame that it’ll most likely just serve to distract people from important considerations.
Even if there is anything to this, dubious though it sounds, surely the more likely explanation is that kids with autistic tendencies would watch more TV?
Nah, it was my cousins. I got 3 that year and the whole family is nuts.
I have no kids, and I don’t know a lot about what women are educated about when they have children.
However, I’m wondering whether there isn’t some kind of class basis to this claim - that is, I think that more-wealthy women may be more alert to symptoms of autism, and they might also be in the socioeconomic bracket that has cable tv.
(I did not go read the study. Please edify me if I’m off the mark.)
Crap, that sounded bad. To clarify, I think that women who come into the hospital boned-up on all the latest literature may be more aware of symptoms of autism than women who are on the making-ends-meet end of the economic spectrum. Not meant to be a judgement on anyone - just saying one has more time to research than the other.
Sounded fine to me, MM. Like posts above mentioned, a correlation (more autism in regions with more cable) doesn’t necessarily imply causation; the causation could go the other way (autistic kids being more attracted to TV) or both things could be related in the way you posit.
And hey, cable where one of the channels is dedicated to medicine and/or childrearing would be a way for those ‘cabled’ women to get their info on authism.