Thursday, March 01, 2007

Children and watching TV

There are many ways how an economist come out with a certain topic. Amongst them are the personal experience.

Motivated by experience handling his autism son (now permanently cured due to policy to bandon TV for his child) , a cornel economist , M Waldman, and his other two fellows S Nicholson (cornell), N Adilov (purdue) did a research on autism and watching TV (read here). They found that watching TV could cause autism. In their research, they used rain prescription (thus the wheater report) as the instrument variable.

This results then become the heated debate among economist and geneticist. The geneticist argue that the research needs an experiment in a controlled/steriled area. Otherwise there would be a misconclusion (inconsistency in econometrics terms). (read wsj). Pat Levitt, geneticist from Vanderbilt University, gives some possible misconclusions. He argue that there are possible several factors such as the quility of the house, the cleaness of the house, the area where the house is located and the air quality inside the house, that affect the autism rather than watching TV (read time).

In economics terms, Levitt argue that the instrument variable used by Waldman et al is not valid (the rain prescription is still not orthogonal with the error term on the main regression). Thus, the regression would be inconsistent. (Since from the variables he proposed -the quility of the house, the cleaness of the house, the are where the house is located and the air quality inside the house -, at least one of them can not be observed by econometrician, thus will be included in the error term of the first regression. Since these variable could have some correlation with the rain prescription in one area, thus the instrument variable is correlated with the error term of the first regresssion)

However, despite the debate on the inconsistency, there is one thing that could be learnt. Before further results and decision from the referee in academic journal (this paper is not published yet. So it could receive many comment from referees thus need to be revised), TV could be not good for children. Why do you take a risk, if you could avoid them. Parents should be risk averse after all (regarding the decision related to their children)

In Indonesia, we are all still remember the "smack down" case.(for debate about this, you can read cafe salemba). Taking into account the results of Waldman et al, now there is one additional reason for parents to limit the time of their children to watch TV.

Now, after all this possible bad effect of TV, will you still let your children watch too much TV?

3 comments:

pelantjong maja said...

what personal experience with TV do you have, Mr Dreamer? hehehhehe. Does TV make you lost in concentration? indeed to much TV will hurt you, especially if you are still single..

rizal said...

"They found that watching TV could cause autism. In their research, they used rain prescription (thus the wheater report) as the instrument variable."

Somehow, this reminds me of famous Anwar's Nasution warning: spurious regression.

My take on the methodological issue is here

berly said...

Hmm.. must also check for self selection bias since child prone to autism tend to be less active thus higher probability to spend more time in front of TV.

I think it will be more the case of correlation than cause-affect.