Monday, May 22, 2006

Hey, what happened to the thought experiment?

Remember this little experiment? It's time to look at the results.

First of all, I thank all the participants for 'voting', and the verdict (from an admittedly small, un-representative and self-selecting voters) is clear: being an also-ran (academically speaking) in an elite institution (one of the IITs in that post) is preferable to doing well in a second rung college.

The 'voters' have also provided some good reasons for choosing the elite institutions: infrastructure, exposure, brand image, networking with alumni, etc. Several commenters opined that prior investment and hard work -- and the exhilaration on getting through JEE -- are good enough to choose the IIT option. The discussion there is absolutely great, so go read it.

Here's my own response to the question, which I had written up at the time of writing that post:

In our country, IITs have a special place, a special status (so do AIIMS, NLSUI, and other such elite institutions). If you get into to one of them, you know you have 'arrived'. So what is so special about the IITs? They give you access to some of the best minds: both faculty and students. You are taught by active researchers. Your class mates and you have a high probability of landing great, challenging jobs. Even if your career is not all that great, you will have personal friends with high-flying careers. This network of high flyers is available to you on tap. Anytime you need it. Also, since getting into an IIT is recognized as a great achievement, you benefit from the social 'bhav' you receive from your extended family, neighbours, friends (and this 'social' benefit lasts a long time, indeed!). Finally, the IIT card opens many doors for you.

My response has much in common with those of the commenters: academic excellence, brand equity, network of peers and seniors, open doors and opportunities, and enhanced self-esteem.

* * *

So, what the hell was the point of that experiment? Well, in a terribly written screed by Prof. P.V. Indiresan, ex-Director of IIT-M, I found this paragraph:

As one correspondent has pointed out, reservation is like declaring a boundary scored in a cricket game as a six if hit by a backward caste player. Such artificial boost appears beneficial. It may not be. As one SC student remarked: "I won a degree in the IIT but lost my self-image." How many students would have done better with their lives if they had been exposed to what they can master, instead of being subjected to a difficult drill for which they were not trained, we do not know.

I am now going to ask you to forget the casteist overtones in the paragraph. Forget also the fact that his article is about reservation (two of the commenters -- barbarindian and cipher -- can breathe easy now, and continue to hold on to their absolutist views against reservation ;-). Stripped of all these considerations, Indiresan's algorithm for choosing an institution is:

Since you are less likely to do well in a top tier institution than in a second tier one, you must choose the latter.

To me, this sounded totally wrong! Loony, even. I was going to write a solid rant on how Indiresan was peddling wrong sort of advice, but decided against it. Instead, I decided on the 'experiment' and the accompanying 'mini-poll' that pits my view against that of the mighty Prof. Indiresan. I am glad that I beat him in a near land-slide!

This is what would be called 'Proof by Majority'. In the best traditions of social un-scientific analysis, I will choose to ignore the non-random, self-selecting nature of the voting population!


  1. Anonymous said...


    You are logically challenged. There is no contradiction between what Indiresan "proposes" and your little experiment. Your experiment asks people what they would "choose" given the "choice," not what is optimal in an ex-post sense. And if you read many of the responses you can see that. What Indiresan is saying is what is good for the student from a neutral observer perspective. You of all people not believing in the paramountcy of freedom to choose should empathise with where Indiresan is coming from (I dont).

    What exactly is casteist in what Indiresan says? Explain before you throw such terms. It is of course very common in Tn to say such things against Brahmins. There is a hoary tradition out there. Akin to calling jews names in Nazi Germany. Only that there is nobody to say anything against the casteist dravidianists in TN.

  2. Anonymous said...


    But, for such decisions in life, the real tricky part is knowing (correctly ie) how u will fair in various places.

    on another note, it is interesting that a former iit director says

    ". There is no public information of how the beneficiaries have fared "

    Is this kind of data 'blocked' officially? Does it go to the HRD ministry ? Is it even collected?

    The reasons for good/bad performance of SC/STs could be numerous and complex..but I guess it is useless to argue anything without the basic data. If the HRD ministry is indeed driven by the knowledge/experience gained from such data (not only iits..take TN, Kar, Andhra), Arjun Singh would do well to let us have a glimpse at it.

  3. Anonymous said...

    sorry about hijacking this comment thread, but I just read the transcript of arjun singh's interview.

    sample :

    Karan Thapar: According to the NSSO - which is a government appointed body - 23.5 per cent of the college seats are already with the OBCs.

    Arjun Singh: What do you mean by college seats?

    Karan Thapar: University seats, seats of higher education.

    Arjun Singh: Well, I don't know I have not come across that far.

    NSSO is a govt organisation.

    most of the interview is like "parliament did it...hence it is correct". Is this the only logic that mr.arjun singh has at his disposal?? If so, he should first do his home work with the HRD ministry before deciding on what % to reserve. He can then talk about how "urgently" the whole thing needs to be done.

  4. Abi said...

    Anon: My version of the 'experiment' is stronger, because it had you as the person to be affected by the decision.

    I do not want this thread to be about whether there is casteism in the paragraph I quoted; since you didn't see it, I suggest that you send me an e-mail, so that the discussion can proceed off-line. I will ignore references to certain castes and certain regimes, in the sincere hope that we can overcome Godwin's law.

    Aswin: Thanks for your comments. This post is really not about reservation. Can we move on please? I will try to get back to a different argument used by Indiresan (not this one, though); can we defer talking about reservation until then, please ...

  5. Anonymous said...

    I had already smelled what you would want to prove and had markedly listed what I thought was wrong with the "thought experiment". That still holds true. I would rather have you answer the questions I raised instead of labeling me biased and one sided. I could argue the same about you. After all your take on reservations too is one sided. If there is a take it has to be one sided. The so called middle path is indeed unfeasible at least in the short term.

  6. Anonymous said...

    Since when did 'thought experiments' become a statistical tool

    I would have much rather preferred a rant to ssuch abuse of statistics.

  7. Anonymous said...

    Abi: what your thought experiment has shown me is that people don't like thought experiments used to contradict anything they believe.

  8. Anonymous said...

    please explain why you want to infer that I(or anyone else) believe in anything at all when the methodology is questioned.

    And by the way, my response was not considered - if that does matter to you and stuff.

  9. Anonymous said...

    Hey, what happened to the thought experiment?

    My response is : Hey, what happened to a scientific study of the OBC quota policy ?

    P. Chidambaram's "life tells him" that quotas are reaching the target OBC groups perfectly. Therefore, there is no need for a statistical study. Maybe you should send your thought experiment to him, he can use it on TV channels.

    To others: the current issue is not about quotas any more. Nobody is demanding a rollback of the quota. It is only about a demand for *basic* data about the policy.

    Would you agree to adding caste enumerations to census 2011 for example ?

  10. Abi said...

    Cipher: Since you were sure about what you said, there is nothing there to answer, is there? Thanks for your comments. BTW, I didn't call you biased; I only described your views as absolutist. I believe there is a difference.

    Nilu: I am sorry to have disappointed you. But, this is the way I wanted to do it; it adds a bit of variety to what I do here (this was the first time, by the way). A rant on a different part of Indiresan's article appears here. I believe it's unlikely to thrill you ...

    Dilip: Yes, *some* people don't like such experiments.

    RC: Please stop this attempt to paint all reservation supporters as arguing against 'data' (see my post here). And, yes, if it matters to you so much, I would support the 2011 census collecting data on people's caste.

  11. Anonymous said...

    My simple question is, why is the goal considered a sacred cow? In all your analyses, you have presumed that I would want what you consider to be a just society.

  12. Anonymous said...

    What I understood after looking at the comments was that almost nobody was interested in education, everywhere wanted to see how much money they could get out of it and whether they could get into IIM A, B or whatever.

  13. Abi said...

    Nilu: Nobody is forcing you to agree with others' views, and I believe you have been doing a wonderful job of carving out your own niche in your blog. So, I really haven't 'presumed' anything on your behalf.

    Anirudh: That's precisely the point! There are many non-academic reasons to do one's studies in elite institutions.

  14. Anonymous said...