Aedar

I'm sympathetic to the idea that the market and mass opinion can make mistakes, but the herding papers cited (a) don't deliver the result that herding is "often" on the wrong choice and (b) don't apply well to the stiuation at issue.
re (a), the papers deliver a probability of herding on the wrong choice as a function of the informativeness of the signals and of the accuracy of the prior beliefs: if the priors are right, inefficient herding is unlikely. So to make a statement about the probability that we would see ineffficient herding requires a specification of the accuracy of the priors. Re (b), the models describe a stiuation in which one agent chooses at a time, and all agents would have the same preferences if they were fully informed. In that model it is not possibelt o have a steady state where 60% choose A and 40% choose B. With a lot of agents making each choice, there will be more information available, and I'd guess that inefficient herding should occur for fewer parameter values. Which isn't to say that Harvard's prestige is warranted; this seems to be an empirical question.

sakky

Look, I'm not telling anybody what to do. Obviously everybody is free to live their lives the way they wish.

What I am saying is that turning down a school like Harvard has potential future career costs, in terms of networking and politicking. If you are willing to assume those costs, more power to you. But nobody should fool themselves into thinking that there aren't any costs at all.

Sometimes you have to do things you don't really like and don't really agree with in order to get to accomplish the things you want more easily. That's life. Now, if you decide that you're not going to do those things, that's fine, but just realize that your goals may be harder to achieve.

Ben Golub

Aedar -- thus it goes with the theoretical literature, as you know better than I. You're unlikely to find a model which fits exactly, but I think many of the herding analyses would extend (with work) if you increased the size of the groups who make the successive choices. If the t=0 group gets tricked by an signal that's far off the mark (just due to a stochastic misfortune), it might be quite rational for all successive groups to follow the wrong herd. Depending on the parameters, this might happen rarely or often.

As for correct priors, I think it would take a lot of twisting to assume anything other than that priors are very diffuse. The whole point of the discussion is to see whether prestige can usefully inform naifs. If almost everyone has pretty good prior information to begin with, then the point is kind of moot.

I agree completely that whether Harvard's prestige is warranted is an empirical question, but how to test it, I don't know. The Dale/Krueger is a start.

Ben Golub

sakky -- now we're getting somewhere. Caltech/MIT offer a more serious technical education (good #1), on average, and Harvard offers more political connections and skill in politics (good #2), on average. Surely the ordering on these goods isn't lexicographic. For different people, it's optimal to load up on different quantities of these goods. And it might be optimal for certain people, given their tastes and goals, to turn down the most prestigious place.

Now that the discussion isn't monofactorial, I'm very liberal about what is best for most people or for a given person. That's an empirical (and very personal!) question.

But I would like to add that my emphasis on the arbitrariness of the herding location isn't totally abstract. It's probably too much to hope for with big groups like incoming undergraduate classes, but in smaller groups of people can dump the "prestigious" place and make it something else. In the 1990's, evidently, the best econ grad students stopped coordinating on MIT and started coordinating somewhat on Harvard. (See Michael Steinberger, "Harvard Is Seen Gaining In Economics Game; Top Graduates No Longer Favor M.I.T.", New York Times, December 27, 2001. http://select.nytimes.com/search/res...AB0994D9404482. Free with a university email address!) Since the only good thing about the most prestigious place is that most of the best people are there, if they just choose to go elsewhere, suddenly the prestige slowly slips away. It's things like this that limit my reverence for this all-important (at least on CC!) concept.


P.S. I realize the final example I give is kind of funny, given the overall discussion. But in the 70's and 80's, it apparently wasn't even a question -- virtually nobody picked Harvard over MIT.

pebbles

aw5k-

actually thats kind of interesting. I applied to internships for the first time and sometimes thats how they pick their people- interviews with case questions and brain teasers and stuff. Sometimes I have a real problem with that method. Especially the brain teasers, the questions designed to trick you somehow and because clearly, if you're not "with it" enough to realize that they changed one word on you from the last question, you're not cut out to be a good computer programmer. People come in with different levels of preparation, this method strongly favors those who have a lot of experience with interviews and I don't think those people are necessarily the best thinkers or the best doers.

But all that aside you know they are private companies they can do whatever they want and also they want people who from the moment they are hired will do the best job they can so it is understandable. Colleges, though, I think have a different goal. They want students who will be good learners, not necessarily the kids who went to the most rigorous high school and had high levels of academic preparation before they even applied to college. You say this will eliminate the "unfairness" but I feel it exacerbates it tremendously. It will just increase the achievement gap between rich kids and poor kids. Those who can afford it will hire tutors for that kind of interview training and those who are relegated to poor publics will naturally perform poorly by comparison. It eliminates the whole concept of "context" from the admissions process.

I maen, for the record, I don't think I would have ever gotten into MIT under that kind of policy. I just wasn't trained that well by my high school in that regard. After all, thats why I wanted to go to college here.

sakky

First off, what I would argue is that more people would probably be more optimally served by attending one type of school, followed by another. For example, if you've gone to MIT or Caltech for undergrad, then you should go to HYPS for grad school, and vice versa. That way, you would get the best of both worlds.

{But of course most people will never have that option. So if that's the case, then the safe thing for you to do is to choose whichever good is more valuable at the time you get to make the choice, because you won't know if you will get another such choice later on down the road. In fact, this gets back to something I've been saying to you before. You turned down Harvard for Caltech for undergrad, and you did well. Good for you. But what if you had not done well at Caltech? In that case, I would strongly suspect that you would have liked to have gone back in time and chosen Harvard, as it ist the safe choice.}

Secondly, you raise the issue of geting a serious technical education at Caltech/MIT. However, speaking of MIT, like I said before, MIT is changing its focus. Consonant with spirit of the Steinberger NYTimes article, MIT has been raiding scholars from numerous other top universities in building up its own political science department. That goes a long way towards explaining why that department has been able to attain a #10 ranking in just 40 years of existence, a rather impressive accomplishment when you consider that MIT has to compete against many far older poli-sci departments. I believe that it is the youngest elite (as in top 15-20) poli-sci department in the country. Similarly, you know see many of the top poli-sci graduate candidates turning down far older schools to come to MIT. MIT still does not compete toe-to-toe with Harvard or Stanford for the top poli-sci grad students, but the gap has been significantly narrowed.

Similarly, you can see the same sorts of changes at the MIT Sloan School. It wasn't that long ago when basically the only thing that Sloan did well was industrial/technical/operations management - or what was generally deemed 'management science'. The Sloan School then went through a transformation in adding broader aspects of business administration, especially in culture, leadership, organizational behavior, social psychology, and the like such that now Sloan is a fully-fledged and broadly based business school that now attracts numerous students who are more interested in those 'softer', non-technical aspects of management. For example, some of the recent Sloan PhD's who do not specialize in technical subjects but rather have specialized in sociology, leadership, international business, or psychology have been placed at top B-schools like Wharton and HBS.

But this gets back to what I was saying before - do you want MIT to change? Specifically, do you want MIT to enter those less technical fields, like political science or management culture/sociology/psychology? If the answer is 'yes', then I think that inevitably means changing your admissions process to place less emphasis on technical ability and more emphasis on social skills and social knowledge. After all, I would argue that most of the best political scientists are not geniuses at math. Many of the world's best business students, either future practitioners or fuure business academics, are not geniuses at math. MIT is never going to change to become a broader-based university if it solely emphasizes technical acumen in its admittees.

Now, of course, one could argue that MIT shouldn't even be trying to become a broader-based university at all. But if that's the case, then you have to believe that many of the initiatives that MIT has started in the last few decades were ill-advised. For example, it would mean tha MIT should not have started a fully-fledged poli-sci program. It should not have built a strong linguistics and philosophy department. The Sloan School should have remained a technical/operations management boutique school. It should not have launched the Media Lab. In other words, most of the new initiatives that have happened East of Ames Street in the last 50 years should not have happened.

Ben Golub

I think there are lots of places basically similar to Harvard (e.g. Princeton, Stanford, Yale, and countless others) in trying to cover everything and give a liberal arts education to good students who fall in all areas in terms of interests.

There are very few places who have historically set as their goal to educate the science and math geniuses of tomorrow in a focused environment with very high core standards for everyone. I definitely will not argue with you about whether that's good-- but my personal belief is that it's valuable to have at least a few. The fact that MIT is flaking out and leaving Caltech to be the only such place makes me sad.

pebbles

FLAKE FLAKE FLAKE I don't think MIT will ever be BAD at Science and Math. Or even NOT-THE-BEST at science + math. Like, ever.

Ben Golub

rawrrrrr. i would attack more but your whole school would crumbe into thousands of little flakes and i am too nice.

Sklog_W

Don't worry if you work hard you can avoid the liberal arts education at MIT. Remember economics counts as a HASS course, it is a social science. I've written a total of three papers in 2 years at MIT.

pebbles

I've written 7 papers this semester!

Ben Golub

seeeeeeeeeeeee! flaky.

phuriku

"There are very few places who have historically set as their goal to educate the science and math geniuses of tomorrow in a focused environment with very high core standards for everyone. I definitely will not argue with you about whether that's good-- but my personal belief is that it's valuable to have at least a few. The fact that MIT is flaking out and leaving Caltech to be the only such place makes me sad."

True, but only on the national level. Majors at schools outside the U.S. (the typical example being Cambridge, of course) are usually very focused on a specific area of study, despite the fact that the school may be strong in both humanities and science.

And what's wrong with that? Absolutely nothing. And if MIT wants to expand to be more focused on political science and humanities, so be it, as long as the education of the student of science isn't harmed. But with the way MIT is going, this doesn't seem to be possible, and certainly with the modern admissions process at MIT, students of math and science who are not the best at humanities are certainly discriminated against.

MIT seems to have inherited the poor qualities of our high school public educational system. At an attempt at a more "well-rounded", "generally educated" student body, humanities and social science courses are forced upon high school students. MIT already has established policies that accomplish this task of making its students more educated in the humanities, but it seems that, with the MIT administration, the policies will become more oppressive as time progresses. Hell, I bet you that in the next 20 years, MIT will make community service a graduation requirement, like many of our public schools have done already (mine included).

Caltech has similar problems, though, although certainly not to the extent of MIT. Chem 3a, for example, sounds like a pretty oppressive and unnecessary course, and it seems to be stuck into the core curriculum for no reason at all. Core science curriculum should be dealt with very carefully, and I think the Caltech administration has not done a good job in this. Instead of forcing students to be more "well-rounded" in the humanities, the Caltech administration is imposing "well-roundedness" pertaining to general science, which is just as silly. Say you're studying to be a theoretical physicist. Now, core math courses and core physics courses are common sense, and probably a core chemistry course or two wouldn't hurt. But why biology? Why Chem 3a? Just to have more useless information that you'll never use so that you'll be more well-rounded in the sciences? Bullcrap.

The good thing about Caltech, though, is that there's hardly any chance that the conditions will get worse. Caltech's not going to implement a policy that requires students to take Geology just because Caltech is a very good school for Geology. MIT, on the other hand, would without a doubt create a (probably small) political sciences humanity requirement if MIT suddenly became the #1 school in the nation in politics, using the excuse that "we want our students to have an education that extends as far as possible". A mile wide, an inch thick. Completely useless.

"That government is best which governs least." MIT could improve a bunch and Caltech could improve a little in this aspect. Now, why don't we have more schools in the United States like Cambridge where students can study whatever they want without all this "well-roundedness" bullcrap?

Rister_Chutophs

The difference between American univs and Cambridge (and all the other European universities like Oxford, Sorbonne, ETH Zurich, etc) is that students at Cambridge have a liberal arts background prior to matriculation: Everyone who wants to enroll has to sit and pass the national examinations (A-Levels, Abitur, Matura, etc.).
In order to pass these examinations students have to pursue a high school with a broad curriculum. Even the prospective Cambridge Physics major has had 4 years of history and at least one foreign language in high school, even the prospective Oxford "Comparative Literature" major has had Calculus in high school.

Therefore, I think it's reasonable that American colleges try to give teach their students a liberal arts background, because education is not an apprenticeship for a specific profession; education is, too, about teaching students how to think critically and how to use their skills in a variety of fields.

phuriku

"Even the prospective Cambridge Physics major has had 4 years of history and at least one foreign language in high school, even the prospective Oxford "Comparative Literature" major has had Calculus in high school."

How is this any different from American high schools? The typical English major (who ends up going to a "prestigious" university) takes AP Calculus AB, at least, by their senior year in high school. Also, most of the people majoring in science that I know have taken AP classes in history and English and received 5s on the examinations. Most American public school systems require 8 semesters of English, 4 semesters of social sciences, 6 semesters of science, and 6 or 8 semesters of math to graduate. Also, I think they're making a policy in some high schools where you have to have at least 2 years of foreign language as well.

"Therefore, I think it's reasonable that American colleges try to give teach their students a liberal arts background, because education is not an apprenticeship for a specific profession; education is, too, about teaching students how to think critically and how to use their skills in a variety of fields."

The scientist who is able to get into MIT or Caltech should be able to think critically and argue logically in any field. It shouldn't take training. This is particularly because the logic behind science and humanities is very similar.