pebbles

"I'd also have no problem if they admitted more than 30% of the class early; it seems to me that being accepted to college in December provides much more of an opportunity to enjoy one's senior year, which is apparently a major goal of the MIT admissions office."

Tell me you're kidding. The major goal isn't to PLEASE EVERYONE. Then MIT would accept everyone that ever applied. It's to make it least stressful for the most people. You just don't happen to fall into the "most people" category, and being childish about it (you can wish that admissions ended by Junior year of high school if you want, since that makes for a non-stressful senior year) won't help your point.


And also, "it's much easier to analyze and find fault with their policies than with the murky, cultish practices of the Ivies "

This is kind of an incomprehensible statement. You're saying you'd rather not know anything? It won't stop you from being deferred then waitlisted... you'll still receive two very impersonal thin letters about it. You just won't have anyone to commiserate with. That kind of attitude and you're just ASKING to be abused. You'd rather live under a dictatorship than have to look to closely at the ugly issues of the country in order to vote.

aw5k

mootmom:

Having had an MIT interview myself, I _know_ that it's not of academic nature. You're not posed any math questions, no problems to solve. It's just a get-to-know-each-other sort of thing, with the opportunity to ask questions.
Recommendations are another thing I find highly questionable. Of course, (almost) every teacher will write good things about the students, so that won't stand out anymore. Instead, what stands out are reports in which teachers highlight a certain special aspect of an applicant. In the end, the question whether the teacher knows how to write really exceptional recommendations. It's not whether you're good or not, it's whether your teacher's good or not, in a sense.

mootmom

I didn't say it was academic, did I, aw5k? I said it was incorrect to say it was of "no weight whatsoever". In fact, it may be important because it is NOT "academic", if the intent is to learn more about the person behind the paper application.

And my experience contradicts your assumptions. It was not either the banal glowing words a recommender wrote, nor whether they were great writers. (Although of course having seen so many recommendations, the admissions staff has likely gotten pretty good at sniffing out pro-forma recs, and can tell when a recommender's enthusiasm is genuinely special.) It was what they said when contacted on the phone that seemed to be memorable. (No, not every recommender is contacted on the phone. But the admissions staff do contact recommenders by phone sometimes to learn more. To me, this shows they're going an extra length to try to find out who those "on paper applicants" really are.)

aw5k

mootmom:

Sorry for this misunderstanding. Concerning the interview's weight, I think it is only of considerable importance if the applicant appears either in a extraordinarily negative or extraordinarily positive way.

I am sure the admission staff tries their best to look who that "on paper applicant" really is. I just doubt that this can be done effectively (this is not to say that it cannot be done _at all_, but I think this process sometimes results in unexpected and not seldom "unfair" decisions).

pebbles

What do you recommend they do, aw5k? You know, to eliminate the "unfair" decisions.

aw5k

Well, Oxford and Cambridge, for example, invite (almost) every applicant to an interview where the applicant is (1) assessed academically (he is asked to solve problems during the interview), and (2) assessed personally (there's usually a "general" interview with regards to your passions, etc.). I think it's a much fairer process.

However, I understand that it's impossible for MIT to interview each person individually. Their admission policy right now is probably the best they can do.

sakky

Nice try, Ben. Bikchandhani et. al. and Banerjee are not relevant to the discussion because they are not models of difficult-to-replicate signals , and that is precisely what provides signalling value. Not everybody can get into Harvard (or MIT or Caltech) and that lack of replicability is what makes them valuable as signals.

As you can see from the Bikhchandani model, actors are unrestricted in which particular group they want to join. Hence, there is no a-priori reason why individual actors won't decide to conform to the group. Same thing with the Banerjee model - i.e. in the model of restaurants A & B, there is nothing that 'restricts' patrons from choosing either restaurant.

I would agree that if just anybody could decide to attend any school, that your above references would probably apply. But that's simply not the case. It is precisely because not anybody can go to any school is what gives certain schools signalling value.

If you want a more sociological based view of what I'm talking about, I suggest Podolony 1993 in ASJ. If you want a more business-oriented reputation signalling model, I suggest Fonbrun and Shanley 1990. And of course if you just want a more economic model, Milgrom & Roberts 1986.

Ben Golub

Your criticism is not on point. Usually on these boards people (including you!) discuss prestige in terms of cross-admit battles. If we restrict attention to those people who are admitted to their top choices -- who really are sending whatever signal they want to send -- both models I cited apply, and you can still easily get herding on an inferior option.

Note that the existence of people who are constrained in their choice (e.g. only got into Harvard because their daddy gave a building) only adds noise to the prestige ranking and makes it even less reliable. Even in the purest case when all colleges compete head to head for the exact same group of best students, prestige confounds evaluation of true quality. When you add more noise, things are even worse.

In general your response seems very confused. The two models I gave are quite general in that you have a sequence of people with choices to make. There are no assumptions made about how they came to be in that situation, and whether it was easy or hard for them. So the model applies quite well regardless of the difficulty of getting admitted (in our case).

sakky

Yeah, but that doesn't exactly answer the question of why you also applied to Harvard for undergrad too. In fact, if anything, it actually raises even more questions. In fact, I now have the same question that pebbles has - that if you really didn't think that Harvard undergrad was that good, why did you even apply in the first place?

First off, by that logic, should we be expecting you to matriculate at MIT then?

Secondly, let me provide a defense for Harvard. The Harvard doctoral programs are quite serious places for study. I might agree with you that the professional degree programs are status-conscious. But I'm fairly certain that few people enter a Harvard doctoral program solely for purposes of status. If all you want to do is get a Harvard degree, there are easier ways to do it. Get an master's degree at KSG or the GSE or one of the other professional schools. That's a far more straightforward way to tag yourself with the Harvard name, if that's all you're really after.

sakky

Cross-admit battles are only one aspect of prestige, and for the purposes of this topic, an irrelevant one. Again, you are presuming that everybody who gets into Harvard can get into MIT or Caltech (or wherever is there other top choice). Come on, we both know that's not true. Plenty of people who get into MIT and Caltech didn't get into Harvard. My brother, for example, didn't get into Harvard. {And yes, there are some people that get into Harvard that don't get into MIT, but almost certainly less.}.

Hence, we are STILL in the world of costly signals. Again, I would agree with you that if everybody who could get into, say, one of HYPSMC could get into all of the others, then your cited papers are applicable. But that's not the case, so my cited papers are applicable.

Ah, but that is entirely endogenous. Think of it this way. You say that people give buildings so their kids can get into Harvard. But that just begs the question why Harvard? . Why aren't they instead donating buildings to get their kids into MIT or Caltech? Why Harvard specifically? Again, it gets back to the fact that Harvard is the signal that is the most difficult to replace, hence contributing to its status ranking. As Podolny would argue, the way you preserve status is by being selective in your social network - if everybody could become a Harvard graduate, then Harvard would lose its status. {The same thing applies with Caltech - we both know that if everybody could get a Caltech degree, then Caltech would lose its status.}

Don't get me wrong, Ben. I sympathize with your position. I think MIT and Caltech deserve more recognition and status relative to Harvard. I think that society should place greater value on rigorous technical education and less on social networking, persuasion, and politics. But what can I say? That's just not the way the world is. We have to deal with the world the way it is, not the way we might like it to be.

Ben Golub

Look, I agree many models may work. But it's easy to see that in each one, prestige is meaningless when you're trying to decide the underlying quality of the education (as opposed to the value of the signal).

Take the Podolny story. Prestige there is just this membership which is hard to get and valuable by virtue of that alone. The whole point of these models is that they assume that there is nothing intrinsically better about the prestigious place and try to explain it purely as a coordination phenomenon. And all of them succeed from different angles.

So the market is only measuring a coin flip -- where people decided to coordinate. It is quite possible in each of these models that another place would give you a better education, and people are still herding on the inferior prestigious place because of the signaling value.

Obviously the signal must be very valuable -- nobody's arguing against that. I'm just saying, the prestige ranking here is not measuring quality of the education. In the most idealized free market nut world, it's measuring the net present value of the signal PLUS the education, and if you don't care about status, it will often be true that the better education (in an intellectual sense) is to be found at the less prestigious place.

Ben Golub

We have a lot of common ground, I'm just a little more optimistic. Naive people might think that the herding on Harvard really does reflect some underlying quality. By explaining that it is a coordination effect probably quite disjoint from real merit (defined in the technical improving-your-mind sense), people might make better choices for themselves.

Now, if in the end all people want is prestige, then you get nowhere. But I'm a little more hopeful about people.

sakky

See, I'm not even entirely sure about that. One of my greatest criticisms of MIT (and by extension, Caltech) is that it often times doesn't prepare students properly for the way the world really works - namely teaching students the communications, social networking skills, leadership, and politicking that go a very long way towards determining success in the real world. Like I said before, like it or not, we live in a world where the best ideas don't always win out - rather the ideas that win out tend to be the ones that are properly packaged and marketed. Or, put more bluntly, you can be by far the most technically qualified candidate, and still lose out on a job offer or a promotion anyway to somebody who is less technically strong but more socially polished.

I have heard it numerous times that the philosophy of MIT is that it doesn't need to market itself and therefore the students don't need to learn how to market themselves, because quality wil inevitably "out" itself. But this is simply false, or alternatively, the definition of quality needs to be carefully delineated. Quality means not just having the best technical ability, but also means having the ability to remain consonant with the organizational forces at play. For example, in your previous case of VHS vs. Beta, it has been asserted (rather questionably in my opinion) that Beta has better "quality" as measured by higher resolution and less crosstalk, what really mattered was ultimately the easier manufacturing/engineering design of VHS which gave large-scale manufacturers, especially powerhouses such as Matsu****a, a strong incentive to pick VHS. I would argue that simpler manufacturability(which ultimately led to lower prices and greater product variety and availability due to the large number of vendors) is at least as important a part of overall 'quality' of a technical design as is picture quality.


(I say 'questionably' because Cusumano and others have shown that whatever picture quality advantages Beta had were quickly matched by VHS improvements i.e. VHS-HQ and SuperVHS, and that whatever fidelity advantages Beta had were swamped by the fact that recorded over-the-air broadcasts were going to be relatively low quality anyway due to simple transmission interference - hence in practice, beta's picture quality advantage was small at best. It should also be noted that Betamax did in fact actually "win" within a particular market - as the related Betacam technology dominated the professional studio industry itself and actually eliminated the VHS-derived professional standards - because that market is one where picture quality really truly mattered and where customers were less sensitive to price and availability issues, because these customers would seek out the products wherever they are, and pay whatever price, in order to get high picture quality. Hence, you have a case of dual markets where each market's customers prioritize the product features that they consider important, VHS dominating the consumer markets, and Betamax derivatives dominating the professional markets}

But again, like I said, I don't think the problem is really with univeristies. The problem is that the world does not value rigorous technical knowledge as much as it probably should, but instead places great value on social skills and politicking. The fact that Harvard is considered to be more prestigious than MIT, or that any of HYPS is considered more prestigious than Caltech is a simple reflection of this fact of life. The valuation of Harvard as a signal and as an educational institution is just a simple reflection of what society as a whole chooses to value.

Olo

I would say there's a good reason the world values social skills/politicking more than technical skill. Even in fields of technical skill, social skill often has far more impact than it should. Graduate school and academia in general is horrifically political*. In a perfect world, funding and grants and space and resources would be allotted purely on intellectual merit, but they're not. I can see why having students capable of politicking would definitely be beneficial for an institution in the long run. That is to say, it makes perfect sense why an institute would value such skills...

This is not to imply that this is the way things *should* be, but every institute does have it's own interests in mind. Every student is an investment...

*I am not a graduate student, and this is simply filtered from my friends who are.

Ben Golub

sakky -- okay. All I'm saying is that if someone, for her own reasons, wants to care only about technical ability and not about politics, then it may well be that Harvard is not the best place. Even if that hurts this person in the long run with "the world". For this purpose, understanding the way prestige influences matriculation decisions -- and obscures true judgment of technical education quality -- is crucial.

I hope you don't want to tell individuals what they "should" and "shouldn't" care about in planning their lives, based on what "the world" values. That would be a very unscientific paternalism. Simply put, people have the right to value many important things that aren't measured in "prestige" units, and you're getting dangerously close to telling people there's only one right thing to care about.

Anyway, I think your long post is kind of off topic. You admit that the most prestigious place doesn't have the best technical or intellectual education and then argue that this is not what "really matters" because politics can matter more for gettng your ideas implemented. But I'm not really interested in arguing about what "matters", just the facts. People can make their own value judgments.