TV and Autism

[jennifer_r]

Cornell just published a study which shows a strong correlation between TV viewing and the rise in autism. From various articles I read on the Internet, it seems there's some issues with the study but I think that it is worthwhile to do more research into this relationship.

http://www.slate.com/id/2151538/

Jennifer

Mom to:
Christopher 12/29/89
Adelaide 8/23/04
Bronwyn 11/9/05

http://www.gynosaur.com/assets/ribbons/ribbon_garnet_6m.gif[/img][/url]

[kcandz]

Thanks for sharing this link.

[shilo]

wow. that's a head turner. will sesame street and baby einstein and signing times be the thimerisol of this next decade? i've cut way back, at least when DS is awake, but does make me give more thought to the zero tv concept - sigh - i'm not sure _i'm_ ready to go there...

lori
Sam 5/19/05 How lucky I am that you chose me.

[denna]

Thanks for sharing this link. Autism is such a broad spectrum, and ADD/ ADHD has been linked to watching too much t.v. for a long time. I believe they classify these behavioral disorders as a mild case of autism.

[KHF]

It's funny you should mention this particular study because just two days ago my best friend (PhD candidate at John Jay College in NY) sent me this rant on this exact study:

"If anyone in any of your parenting groups says something to you about television viewing by young children causing autism, tell them your social scientist and statistics professor friend told you that it's cr@p and that correlational relationships between things do not mean one causes the other.

I just found out about a study (which I may use in my classes to demonstrate what bad research looks like) by a bunch of people at Cornell (who should just know better) showing that as cable television access increased in California and Pennsylvania, levels of autism increased as well.

Yes, and according to the Church of the Flying Spaghetti Monster (not a joke - well, it is but not mine), as the number of pirates decreased, average global temperatures began to increase (and yes, this is true) so clearly if we had more pirates we wouldn't have to worry about global warming. There are plenty of reasons to limit the amount of TV children watch, this isn't one of them - at least not based on this kind of irresponsible research."

Just another perspective from someone in the statistics and social sciences field :-)

Kirsten
DD Laurel - 8.22.2005

[MartiesMom2B]

Thanks for posting this. I tend to agree with your friend. :)

-Sonia
Mommy to Martie
& Li'l Girl Bunny to come Feb. 2007
http://bd.lilypie.com/Kchhm4/.png (http://lilypie.com)

[SnuggleBuggles]

I listened to the slate.com podcast on this yesterday. I liked the possible theory that it wasn't the TV- it was the time indoors that might actually be the problem. Air quality issues and other environmental concerns to blame rather than the tv.

Beth

[mommd]

Yup, correlation does not equal cause and effect.

I saw a study that showed a strong correlation between men having beards and later getting lung cancer. So clearly, beards are a risk factor for lung cancer, right? I like the pirate example too. :)

[brittone2]

we're tv-free but personally, I doubt any *major* connection to Autism.

[alexsmommy]

I didn't even open the link - I rolled my eyes and thought - "correlation". It pisses me off because if you haven't been tortured by having to take multiple statistics classes, these studies can really work people up. I'll probably try to read the original study later, but I wish the media would think about this stuff before they promote these kind of studies and terrify parents.
Alaina
Alex Feb '03

[bubbaray]

ITA with your friend. Correlation does not equal causation.



Melissa

DD#1: 04/2004

http://bd.lilypie.com/SasRm7.png

[juliasmom05]

FWIW, I can find the original study, but I'm not so sure that it is a peer-reviewed or published study. Or at least, I haven't been able to find it ;-).

[lizajane]

with sonia again here.

frankly, the tv seems to be the only way to calm my potentially ADD kid. i find it to be an excellent re-focusing tool. (go ahead, flame away. i found something that works for me and my fanily and i will be proud to defend it after warding off all of my own guilt over the past year.)

i don't think you can "make" someone autistic anymore than you can "make" someone gay. you are born with a predisposition, IMO. but what do i know. just my opinion.

ETA: i would think it would be more likely that children who watch way too much television are likely to have be influenced by poor parenting. and that would be the cause this, that and the other. if i kid sits in front of a TV unsupervised ALL DAY, then of course they lack social skills, have no energy release, haven't learned how to focus on other areas of interest...

[mamalou]

My DH sent me a llnk to the actual research a couple days ago also. First of all, it was done by a couple of economists. Second, the study actually looked at precipitation... yes, as in rain. They hypothesized that people who live in rainy climates watch more tv than people who live in less rainy climates. Therefore, there would be higher rates of autism among children who live in areas with more precipitation. Come on... ridiculous!

[writermama]

Ok, your pirates and global warming has just replaced my previous all time favorite example of the correlation doesn't mean causation lesson. (Previously it was the odd but true phenomenon that crime increases as sales of ice cream increase.)

So it is truly as I have long suspected -- Johnny Depp can save the world.

[ppshah]

First of all this guy is an economist. How is he all of a sudden an autism expert. If he is an economist, he should know that just because A and B occur at the same time does not mean A caused B.

Okay if you want to read the actual paper here it is:

http://www.johnson.cornell.edu/faculty/profiles/Waldman/AUTISM-WALDMAN-NICHOLSON-ADILOV.pdf

It's 67 pages long though.

My summary: In counties that have more rain, the kids watch more television. The counties that have more rain, had higher rates of autism, so television must be the cause of the higher rates of autism. Well, maybe it's the high precipitation that causes the autism?

This is total crap! They think the amount of rain impacts how much a tv a 6-12 month old watches? Oh yeah, because so many 6 month olds run around the yard. Also, the rates of autism are rising in developing countries as well.

Okay sorry for the rant, but this is a touchy subject for me. I'm a pediatrician so I generally preach the less screen time the better. DS watches about 3 hours of DVDs a week. He watched even less when he was younger. I do think too much tv may increase the hyperactivity or self-stim behavior of a child with autism.

The media doesn't seem to think it's necessary to check it's sources anymore, they'll just report any crap out there. That's what parents of autistic kids really need-people to think, oh they must have let their kids watch too much tv. We can add it all the "can't you control your child" or "he just needs discipline".

[sarahsthreads]

Thanks for the other perspective! Since I reportedly watched Sesame Street every weekday until kindergarten, I was having a hard time believing that the same habit was going to affect DD. Well, except for the fact that she somehow already recognizes all of her letters, upper *and* lowercase, and I sure didn't teach her those...

But the pirates thing? That's true. If we had more pirates global warming wouldn't be a problem at all. ;)

Sarah :)

[Melanie]

I wonder if it isn't more that it causes Autism-spectrum-like behaviors in children who would not have them, rather than causing *actual* Autism. KWIM?

[cchavez]

my DS has Sensory Processing Disorder (related to ADHD and AU) and he does not even like watching TV!

[KHF]

For your further reading, I found the original link to the Church of the Flying Spaghetti Monster and the pirate/global warming example that my friend referred to in her rant...

http://www.venganza.org/about/open-letter/


Kirsten
DD Laurel - 8.22.2005

[egfmba]

Okay I read this a few days ago and my first thought was, "I call bull-sh*t!"

Is it just possible that the parents of children with autism tend to move to places that enable them to use their skills and that they just so happen to concentrate in that part of the country? This study does nothing to take into account the free will and movement of U.S. citizens.

My DS is on the spectrum, and another mother at his school mentioned that she moved here from California (Silicon Valley) and they had a ton of autistic kids there. Well, duh. The Asperger's parent is likely to be engaged in a job that requires a high level of intelligence, so s/he moved to California to take such a job and just happened to have kids there (who may also be on the spectrum). Thus, a high level of spectrum kids in that part of the country. Does Silicon Valley cause autism? No. Correlation, one having nothing to do with causing the other. I don't imagine it rains much there, either.

I would have been diagnosed on the spectrum as a child, but I spent more time outside than on t.v. Don't get me wrong; I loved me some commercials, but it didn't cause me to behave in a certain manner. Looking down my family line, you can see it (once you know what it is).

Sorry so long; I just happen to agree that this is total correlation and there are much better reasons to restrict a kid's t.v. than this. To throw this out there is to encourage (discourage?) those socially disadvantaged parents who are not as educated (and therefore can't appreciate the difference between correlation and causation) to further give up on themselves as parents and let go of their kids just a bit more ('cause after all, they're such crappy parents they let their kids watch t.v. and now they're gonna catch this autism thing! And hey, doc, is there a pill their kid can take for that, by the way?). Sarcasm!!

Do we really need to throw such uninformed propaganda out there (as a society - ahem, news media) in an area where there is already so little correct information to be had?

We need to be educating people correctly on autism, because it's not a rare disorder that couldn't happen to you. It's common, folks. 1 in 166. What is that? 6 degrees of separation? In the past year, I've seen more done in the area of autism education (to the general public) than I remember seeing in the 10 years prior to that. Or at least since 1994 when the DSM starting recognizing that autism was a spectrum.

Keep up the good work, mamas! Keep putting the info out there so we can all learn, because that's the only way our kids will benefit!! I love that this board talks about these articles, 'cause good or bad, the only way to know is to discuss.

e

[daniele_ut]

I completely agree with your statistician friend. This study puts forth a perfect example of a spurious correlation. The two things might be related, but NOT causally.

[jamsmu]

Actually, a recent report found there to be no link between TV and ADD/ADHD. can't remember where the report was, but it was a very respected report.

[JulieL]

Also, the rates of autism are rising in developing countries as well.

************************************************** *******************

Do you know if the rates of autism are rising do to correct diagnosis or if the rates are increasing because the amount of children with autism has actully increased. I'd be very interested in your thoughts on this.

Thanks :)

[kijip]

It is rather funny it was done by economists because fallacy of post hoc ergo propter hoc is basically the first lesson of any basic macro course.

[caheinz]

And, well, here's at least one economist saying what many have here:

<http://www.freakonomics.com/blog/2006/10/17/tv-causes-autism-i-doubt-it/>

(Gist: Interesting study, but it is correlation! BTW, author is Steven Levitt, University of Chicago economist.)

I'm actually curious about the original paper, as the media do tend to overstate results in such a way to make better news. (I'd be surprised if the original authors made such a mistake, but it's certainly possible...)

OK, nosed around the Cornell website. Found this press release: <http://www.news.cornell.edu/pressoffice1/Oct06/tv_autism.shtml> The study is a paper being presented at a conference today, so nothing in print, and not peer-reviewed. And the author quoted is much more reserved than the news story: "The analysis shows that early childhood television viewing could be an environmental trigger for the onset of autism and strongly points to the need for more research by experts in the field of autism." (co-author M. Waldman). "Could" and "are" are very very different words.

[Ceepa]

There are some interesting points, but at the end of the day I don't lean toward the cause/effect relationship that's being argued in the study.

-Ceepa

[jennifer_r]

In my OP I said that I thought there were issues with the study from what I heard. I don't necessarily believe that TV CAUSES autism just that they are correlated and thus maybe there is something to look into. I didn't expect to upset so many people and for people to think this is bull-sh*t, crap, etc.

I know it's more complicated than TV => Autism, but there has to be some "environmental" factors involved that Autism has increased so much since in the last decade or so and it can't be merely genetics. I believe part of it is that they are better at diagnosing it but I can't believe that would account for the 1 in 167 rate we have now. They're now showing that there is a correlation between TV and ADHD; why is too much TV that ridiculous? It may be one of many factors, such as bad air, something in the water, maybe some preservatives/dyes in food (I'm just throwing these out here). Or it may have nothing to do with TV.

From now on, I won't post anything like this anymore. Sorry if I upset anyone. I was just trying to provide some info.

Jennifer

Mom to:
Christopher 12/29/89
Adelaide 8/23/04
Bronwyn 11/9/05

http://www.gynosaur.com/assets/ribbons/ribbon_garnet_6m.gif[/img][/url]

[mommy111]

Hey Jennifer, I didn't post to the above, but have been reading the discussion with interest and it doesn't seem to me like anyone was trying to personally attack you, I think they were just reading and discussing the study and its weaknesses etc. You didn't do the study...just posted it to be read and commented on. For me, this is the best kind of discussion b/c I learn so much from it so thank you for posting this study and provoking the very intelligent and stimulating discussion that followed.

[drsweetie]

Hey Jennifer, try not to take it personally -- no one's criticizing *you*, just pointing out -- as you had originally acknowledged -- that the study has some flaws. You didn't do the study, so it's the researchers' claims that are being disputed, not yours. I agree with your original post that the study is interesting; whether or not it conclusively demonstrates a causal link between TV viewing and autism is a different issue, and one that's worth discussing. So I appreciate your posting the link to the article because it led to an exploration of a topic that's relevant to lots of people here.

Ellen

[rlu]

Jennifer,

I'm sorry you feel attacked. I read the responses to be attacking the study, not you. I think you did us a service by posting the link, as the ensuing discussion, once past the anger, has been interesting to me.

[mainepotato]

I agree with the other posters. I'm glad that you posted the link because the discussion has been interesting.

[purpleeyes]

Ok, no flames, please! I am wondering why too much TV *couldn't* be a reason for the rise in spectrum disorders. I mean, this study has flaws, true, but I think it is a fair question. We are talking about brain development, after all. Kids don't just pop out with a fully functioning adult brain. Isn't it fair to consider that the TV may interfere with that brain development?

FWIW, I LOVE TV. I watch crap and I watch the 'good' stuff. But DS didn't watch TV until he was 2, and even now its 2 hours a week at most.

Beth

ETA: this post is not in ANY way meant as a condemnation of parents with children on the spectrum! I wanted to discuss the *idea* that there may be a connection, not blame anyone. So sorry for any offense!

[dules]

>That's what parents of autistic kids really need-people to
>think, oh they must have let their kids watch too much tv. We
>can add it all the "can't you control your child" or "he just
>needs discipline".


Thank you for sharing your perspective on this. I hadn't thought of it that way because I have not walked in your shoes. To add more public ignorance and judgement to the challenges you already face, well, that goes beyond irresponsible research/media. Grrr. I hope you write a letter - or send Slate a link to your post.

Mary

[ppshah]

My comments were directed to the study and the media, not to you. In fact I'm glad you posted it- there was a discussion and now more people are aware of the flaws of the study.

Can too much tv contribute to autism? I guess it's possible, I don't think it was the case for my DS. But if that's your hypothesis then try to test it in a real substanial way. I don't think we are any closer to answering that question than before this paper came out. I think this study was so poorly done, it doesn't deserve any media attention. I wonder why an economist is trying to study the correlation between tv and autism. I can't help but wonder if he was hoping to get some free publicity b/c autism is such a hot topic.

There is real information out there coming out but it gets overshadowed by stuff like this.

http://www.cnn.com/2006/HEALTH/10/16/autism.genes.reut/

[alleyoop]

"fallacy of post hoc ergo propter hoc"

Poop! Now I feel dumb and the kids can't watch Backyardagans!

[Elilly]

As the mom of a child with autism, I am very interested in all topics related to autism. I do believe that there is a connection to TV watching and autism, but think that the researchers have it backwards. I think that autistic children, in general, are drawn more to TV becuase it is non-confrontational and does not involve social interaction. So, I think that children who are autistic are more drawn to TV. TV does not cause autism. IMO, autism is a multifactoral issue.. genetic, environmental, and immune. And last but not least, IMO, autism is treatable.

[ppshah]

The Combatin Autism Act proposes to increase reseach, early detection and intervention for autism. It was passed unanamously in the Senate.

Unfortunately, Joe Barton (TX), Chariman of the House Energy and
Commerce Committee, refuses to release the Combating Autism Act from his
committee to the House floor for consideration. He will not do so until
the Senate passes his NIH reform Bill.

If you disagree with this course of action here are some phone numbers:

1-Chairman Joe Barton @ 1-202-225-2002
2-Speak of the house Dennis Hastert 202-225-0600.
3-Majority Leader John Boehner can put S.843 on the House suspension
calendar. Call 202-225-4000.

For more information:
http://www.combatautism.org/site/c.adJCKONvFoG/b.1159505/k.BDF6/Home.htm

Thanks!

[michellep]

Another issue about this article to be pondered:

I don't know if I'd call this author an economist, although he does write about the economy among other things.

He writes a football column for the NFL. This one's titled
"Playmate, aging fighter jock almost predict exact final score!"
http://www.nfl.com/news/story/9232134
It includes his choice of cheerleader of the week.

Not completely knocking him, but he's not really who I'd want interpreting the original study.

-M

[Fairy]

Oh, Puh-LEEZE.

[brittone2]

ITA that it may be more likely to draw a child w/ Autism in...that kind of "stim" can be appealing for Austistic kids depending on what sort of sensory issues they present with.

Many kids I've worked with on the spectrum do a lot of rote memorization of TV shows/movies.

[jennifer_r]

I hope that you (and everyone else) knows that I was in no way trying to put down parents of autistic children. I give you and other parents who have to deal with children who have disabilities alot of credit. I just hope "they" find the cause soon and (even better) find a cure. I really worry about the sudden increase in Autism and have to believe, that while genetics definitely plays a role, that there is some "environmental factor" in play now that wasn't there 15-20 years ago.

Jennifer

Mom to:
Christopher 12/29/89
Adelaide 8/23/04
Bronwyn 11/9/05

http://www.gynosaur.com/assets/ribbons/ribbon_garnet_6m.gif[/img][/url]

[egfmba]

I also hope you don't think I attacked you (and I'm sorry to have come off that way).

Yes, I thought the study was suspect, but as I mentioned at the bottom of my post, thank you for posting it. It gets people to talk about it and it helps us all educate each other (like the poster with the statistician friend and the pediatrician).

Please understand I disagree with the study, not your posting of the topic. The topic is good; it's the study that's poorly done.

Don't know what causes autism. Just know that you can trace it back a few generations in my family. I agree that there may be *some* type of environmental factor at play; I just don't know if that factor is t.v. But that's not to say I disagree with you personally. I just think the study was poorly done.

e

[egfmba]

Bingo here! I totally agree with this assessment.

e

[Judegirl]

I'm not going to get involved in a big debate here, but Kirsten, your friend is not correct. I appreciate his zeal, but as a statistician and a social scientist, here is where he is off the mark.

The Cornell team (they do, in fact, know better) did not assert causality. Your friend - and anyone who has taken even the most basic graduate stats class - is correct that correlation is not causality. However, multivariate regression can control for all intervening variables to eliminate other possibilities. The study, like any qunatitative study, is limited by the imagination and logical thinking skills of the researchers who conducted it.

Example: A study finds that overweight people watch an inordinate amount of tv. Channel 2 news (or Cosmo or whatever) reports that tv causes fatness. No one checks the original study, which never claimed that tv causes fatness, but that excessive tv watching might be a precipitating factor in obesity. The original study actually reports:

- that they controlled for eating by holding tv-watching constant and comparing weight gain across caloric intakes - and found that people who watch tv gained more even if they ate less than people who read books

- that they controlled for high-carb and low-carb diets since high-carb diets ultimately decrease energy levels and could make people more inclined to watch tv in the first place

- that they controlled for class and found out that this only applies to people in the working and lower-classes. Upper class people who watch a lot of tv are LESS likely to be overweight than working class people who watch a lot of tv (or who don't.) Authors hypothesize that upper class people eat a higher variety of healthier better food and are more physically active and thus there is no relationship with tv.

Etc, etc. Guys, the Slate article says this - the authors give about four possible explanations other than causality - and that's not even the original study.

There is a relationship between tv watching and autism. Confirmed relationship, according to this study. No causality. But multivariate analysis suggests that of all the other things that happened during this period (when tv became common in places that have high tv) none of them have a correlation to autism.

The hypothesis above that the relationship that autistic kids like tv does not speak to the overall autism rates climbing. The Cornell study is looking at the change in autism rates post-television use by region and by season, not the frequency of tv watching by infants who are later diagnosed with autism.

Scientific studies never, ever claim to have proven anything. They only claim to have found evidence to the contrary of something else. Cornell has found evidence contrary to the assertion that there is no link between tv and autism. There appears to be a link. No scientist - at Cornell or at John Jay - would tell you that that means causality.

I'm not an anti-tv actvist. I'm just a social scientist fed up with mainstream misrepresentations and misunderstandings. But for what it's worth, I keep my kid away from tv as often as I can - without losing my mind. And sometimes I even lose a little of it just to do so. And when I get lax, I'll be reading the Cornell study.

Best,
Jude

[Fairy]

Damn! Nice.

[kaylinsmommy2]

I know many people who believe the first - that rates of autism are raising due to knowledge about correct diagnosis. It's also likely that the awareness of autism allows parents and doctors to be more likely to look for symptoms.

The other big argument (not mine) right now is the fact that children diagnosed with autism are given much more support (education wise, therapy wise, etc) then children who are diagnosed with something such as mental retardation. So there are parents who are fighting for the autism diagnosis, even after being told their children do not have it.

Caroline
mommy to Kaylin 6/5/04

http://tickers.baby-gaga.com/t/bunbunadb20040605_-9_Kaylin+is+now.png[/img][/url]

and one on the way, due 2/26/07

[mommy111]

Thank you!

[KHF]

I was not trying to get involved in a statistical debate, I was merely passing along information on a topic that came from someone far smarter than me in that area. I'm a computer programmer and a mom, not a statistician or a social scientist.

I have, unfortunately, not been able to open the original study...my computer has issues with it's PDF format. However, I have seen several quotes from it, and their conclusions did use the word "cause" not just relationship. A quick Google search comes up with a rather large listing of people who also disagree with their conclusions. Not just my friend (a she, not a he).

My main concern is that parents of autistic children will see this study (and all articles relating to it...which may be the main problem, not the study itself...perhaps the articles are misstating the researchers conclusions), and believe their children are autistic because of something they did (or did not do)...i.e. keep them away from TV.

Kirsten

[mapg]

Isn't he they guy that said booster seats didn't matter for car safety?
MA

[ppshah]

Thank you! I'm not sure I still understand it, but that makes a little more sense.

"The Cornell study is looking at the change in autism rates post-television use by region and by season, not the frequency of tv watching by infants who are later diagnosed with autism."

I still don't understand something. If you're interested in determining what is causing an increase in autism rates, and you hypothesize tv may be related, then wouldn't you record the frequency of tv watching of infants and see if there is a difference between those that later got diagnosed with autism and those that didn't?

If I had to study this, I would ask 100 pediatriacian to participate in a study. I would design a questioneer about screen time to be filled out at 6, 12 and 18 month check-up. I would then track who later was diagonosed with autism and look back and see if screen time seems to correlate. That would not however correct for the fact that children with autiam may watch more tv so which came first?

You could also do a study on younger siblings of children with autism. For 1/2 the children, you can ask the parents to virtually eliminate tv and other 1/2 to do as they would normally do and see if there is a difference in recurrence.

[mamalou]

>The Cornell team (they do, in fact, know better) did not
>assert causality.

>There is a relationship between tv watching and autism.
>Confirmed relationship, according to this study. No
>causality.
>
>Scientific studies never, ever claim to have proven anything.
>They only claim to have found evidence to the contrary of
>something else. Cornell has found evidence contrary to the
>assertion that there is no link between tv and autism. There
>appears to be a link. No scientist - at Cornell or at John
>Jay - would tell you that that means causality.
>
>I'm not an anti-tv actvist. I'm just a social scientist fed
>up with mainstream misrepresentations and misunderstandings.


Jude,

Are we talking about the same study?

The name of the study linked above is "Does Television Cause Autism?"

From the original research abstract...

"Our precipitation tests indicate that just under forty percent of autism diagnoses in the three states studied is the result of television watching due to precipitation."

[ppshah]

Well, there are definately kids being diagnosed today that never would have been years ago, especially at the milder end of the spectrum. And apparently the children at the most severe end of the spectrum were previosuly misdiagnosed as mentally retarded.

But I also think the acutal numbers are rising. I have nothing to back this up with, just my gestalt feeling.

[Saartje]

Thanks for this post, Jude. Well said.

[Judegirl]

Lou,

Yes. You quoted out of context, for one thing, and conflated "autism diagnosis is the result of televsion watching due to precipitation" (three variables, none of which is autism itself) with "televison watching causes autism" (two variables, one of which is autism).

Here's the last paragraph of the study:

"As a final point, although as discussed our results do not definitively prove that childhood television watching is an important trigger for autism, we believe our results sufficient support for the possibility that until further research can be conducted it might
prudent to act as if it were. In other words, maybe there should be additional emphasis placed on the recommendation of the American Academy of Pediatricians that early childhood television watching should be eliminated or at the very least quite limited (as discussed in footnote 3, current recommendation is that there should be no television watching before the age of no more than one to two hours per day for older children). We see little downside in taking
step and a very large upside if it turns out that television indeed causes autism." (Waldman et. al. 2006)

Jude

[Judegirl]

Thanks. And nice to see ya. :)

[mamalou]

I don't think I conflated anything. I merely stated the name of the study and a sentence from the abstract which might lead one to believe that the authors were trying to make a case for cause.

I'm the first one to say that the media will jump to incorrect conclusions over research (I've said it several times on this board actually), but I think the authors really helped along the misrepresentation and misunderstanding *in this case*. If I had turned in a paper to my graduate research methods or stats class with that title... the professor would have put a big "rewrite" next to the title. However, the paper doesn't follow the APA guidelines that is typical of autism research. It's been awhile for me though, so perhahps I'm just rusty :)

[michellep]

Yes, I agree with you that the authors are helping along the inevitable misunderstandings. It seems a little unfair to title a study "Does tv cause autism?" and then be surprised when the general public confuses the correlation disucssed in the paper with the causation in the title. I think the authors know this will get them more attention.

But since there are social scientists here...I was reading this http://www.freakonomics.com/blog/2006/10/17/tv-causes-autism-i-doubt-it/ written by the economist cited in the original article as someone who first looked at the precipitation/tv watching connection.

What does he mean when he says this?:

"The difficulty with the cable TV analsyis is that there is an incredibly strong positive trend in autism. The cable TV data are basically on an upward trend. The regression analysis is going to have a very hard time sorting out between a steady rise in cable TV penetration and the time trend. In the current version they only include a linear time trend, which is an extremely powerful predictor. My guess is that if they generalize their specification to allow for non-linear time trends, the cable TV result will disappear."

Is there a simpler way to say this that non social scientists might understand??? thanks!

-M

[sdbc]

Hi Jennifer,

I'm glad you posted this. Coincidence doesn't necessarily point to cause, but it doesn't negate it, either. It's at least something to think about and keep looking into. I really think people are just arguing the study, not flaming you in any way.

I do have one possible alternative to this statement: "I really worry about the sudden increase in Autism and have to believe, that while genetics definitely plays a role, that there is some "environmental factor" in play now that wasn't there 15-20 years ago." Another possibility is that autism hasn't actually increased to the level it seems, but rather doctors have gotten better at detecting and reporting it. 20 years ago, a kid who was on the spectrum, but not diagnosed with autism would have just been calleed a "bad" kid. It's impossible to know whether the numbers really have gone up. Does anybody here have data on that theory?


Sue, mommy to Aurora (Rory) born 5/13/04

[Fairy]

Honestly, I was bemoaning the study, not you, Jennifer. I apologize if I contributed to you feeling badly. Please don't!

[Fairy]

Dude. I don't even know what conflated means. I did skool and that degree stuff, but I'm not so smart now that I'm old. Maybe it's just me, but while I love your posts, dude, ya need to dumb it down a bit for me. See, it's the childbirth. When you get pregnant you lose brain cells, and it just goes downhill from there.

[Judegirl]

"The difficulty with the cable TV analsyis is that there is an incredibly strong positive trend in autism. The cable TV data are basically on an upward trend. The regression analysis is going to have a very hard time sorting out between a steady rise in cable TV penetration and the time trend. In the current version they only include a linear time trend, which is an extremely powerful predictor. My guess is that if they generalize their specification to allow for non-linear time trends, the cable TV result will disappear."

Sure. :) He's objecting to the variable they used regarding cable tv, on the grounds that the aggregate pentration of cable tv (how many people have cable) linear in time - it came into existence at a certain time and has grown since then. If autism is on the rise, than this correlation may be spurious - that is, there may be other variables at work here. He speculates that if they complicate the analysis sufficiently to examine the trend in a non-linear fashion, the relationship may disappear.

It may be, for example, that there is something else linking people who subscribe to cable tv and people whose children are diagnosed during this same time period - something else that's also time-dependent. (For example, perhaps people during this time period who subscribed to cable tv are overwhelmingly women who began working in jobs in technology during the childbearing age, and exposure to technology caused a mutation which led to an increase in autism.) Because all of these variables increase with time, multivariate analysis would have to be very rigorous in order to extricate them from one another.

I didn't read the entire study, so I can't respond. And if their analysis was indeed that sophisticated, it'd be beyond me anyway. :)

Hope that helps!
Jude

[Judegirl]

LOL - yeah, I know. This may all sound pretty, but I regularly walk into walls these days.

conflate = to merge, to combine distinct things (ideas, thoughts, voices, etc.) into one

Jude