Friday, October 12, 2012

The normaliens

In an earlier post I mentioned that Serge Harouche (2012 Nobel Prize in physics) is a normalien: a graduate of France's Ecole Normale Superieure. Admission to ENS is strictly meritocratic, based on a competitive exam. The result: 12 Nobel Prize laureates and 10 Fields Medalists from a school with fewer than a thousand undergraduates. (The school is similar in size to Caltech; smaller than most high schools.)

See Defining Merit for the story of Harvard's internal debates of the 1950's, during which a realistic and shrewd admissions dean faced down idealistic faculty committees that wanted to make Harvard more meritocratic.
"Do we want an Ecole Normale Superieure, a 'cerebral school' ... ?"

"What's wrong with Harvard being regarded as an egghead college? Isn't it right that a country the size of the United States should be able to afford one university in which intellectual achievement is the most important consideration?"


as said...

Doesnt USA already have three universities where intellectual achievement is the most important consideration: Caltech, MIT and University of Chicago?

Kyrilluk said...

I wonder why there are so many jews like Serge that won the Nobel Prize.

Azkanny said...

Does this count include literature Nobel winners, e.g. Sartre, also a normalien? (Not sure if there are others, but the number of great literary and philosophical normaliens is very high.)

yulva said...

We all know who won the 1950's debates. Just look at the latest cheating scandal at Harvard. Approximately 125 students in an “Introduction to Congress” course, are being investigated for plagiarizing, or more nicely stated, “inappropriately collaborating” on the course’s open-book, open-note, open-Internet, take-home final exam.
Most could find no wrong in what they did.

My sister, in her first year of med school, met a student who had transferred in from some Scandinavian country. He was shocked when, at orientation, the students were told, point blank, to look around the room because by the end of the first year almost half of you will be gone due to the demanding course work. The transferred student was shocked because in his country the first year students were told not to worry and that "all of you" will graduate to be doctors not matter how long it takes.

Richard Seiter said...

Interesting point about the contrast observed by the Scandinavian student. Any thoughts on how the admissions processes compared? I see at least three possible views on admissions/weeding out and would be interested in other opinions/ideas:
1. Admit everyone (or not everyone, but include those with a lesser chance of success) and let the education process weed out those who aren't capable. (intentionally or not)
2. Admit only those you think can succeed and do your utmost to help them do so.
3. Admit everyone. Do your utmost to help them succeed. View anyone not succeeding as system (not individual) failure.
I think 3 is unrealistic for disciplines with non-trivial (intelligence, personality, persistence, other abilities) requirements (like medicine IMHO) so I assume the Scandinavians were doing 2. If so, I wonder what the admission process was.

I think 1 tends to occur frequently in America because of the conflict between the idea that everyone should be given the opportunity (coupled with the profitability of having more students) and the reality that not everyone has the ability/drive/etc to succeed in a given field. I think the signaling theory of education is also important here in that cheating to get a credential allows the marginally qualified "elite" to "succeed" (I think this is especially applicable at a place like Harvard with a substantial quantity of "non-academic" admittees and a very valuable credential).

LaurentMelchiorTellier said...

There's definitely none of that kind of rhetoric that I've ever noticed. People smile and speak quietly a lot.

I happen to know a med student who has failed his tests and is now changing career, I think the results for actual doctor performance show that selection is taking place. But when it comes to a "craft" like being a doctor, I'm more inclined to worry about a "tough", but bad character who gets by via cheating, than a thickhead who gets by via not giving up because they weren't given enough touch orientations telling them they weren't smart enough. Being such a gentle soul that you're likely to be dissuaded by a stern orientation talk is IMHO not a bad thing for a med.

Warm encouragement + tough, cheatproof tests FTW.

Emil Ole William Kirkegaard said...

Admission in Denmark works by grade point average of "high school" (gymnasiet) and required classes. By default everybody who wants to gets admitted if there is room. When there isn't enough room, highest grade point average in gymnasiet gets the spot. There are some more details, but this is the how it works in general. This system favors women who work more diligently in gymnasiet and thus have higher GPA's.
In medicine this is actually a problem since the sex-ratio is very skewed now. This by itself is of course not a problem, but it becomes one due to sex differences in specialty selection. Women tend to choose patient-focused work, and thus do not specialize. Instead they become general practitioners. This results in a lack of doctors with necessary specialties.

Your (3) is a very common sentiment in Danish media. It is also true since the humanities degrees have a very poor quality as well, and also attracts the least smart students.

I study linguistics at Aarhus University, the second largest university in Denmark.

The student mentioned is normal for a Danish student. I have told many other students similar things based on official drop-out rates - and they were very surprised.

Emil Ole William Kirkegaard said...

Because jews are very smart. Cf. G. Cochran, J. Hardy, H. Harpending. "Natural History of Ashkenazi Intelligence", Journal of Biosocial Science 38 (5), pp. 659–693 (2006).

Richard Seiter said...

Thanks for the details about how it works in Denmark. Maybe the US should find a way of pooling doctors with Denmark--here we have an overabundance of specialists compared to general practitioners (commonly ascribed to higher compensation for specialists).

MtMoru said...

Is the gpa an average of subject specific national exams or is the grading internal to the gymnasiet and done for 4x2 semesters like the US? Do you graduate with 6 or 8 grades or do you graduate with 4x2x5= 40 grades?

MtMoru said...

Sorry to break it to you Hamlet, but the authors of that paper know very little about Judaism (a behavior), especially in its European variety.

If it is true that specifically Ashkenazic genetic diseases are much more likely to affect the nervous system than those of other interbreeding populations then it seems this might just as likely be the result of the high social status given to the best Yeshiva students by traditional, pre-Haskalah, European Jews.

MtMoru said...

There is no uni in the US which determines admissions by competitive exam. College entrance exams are taken into consideration, but in the US it is possible to max these exams and be denied admission to an elite school. It is fairly common actually.

as said...

That is true, but I asked a slightly different question. Caltech and MIT may not admit students based on a competitive exam but do they consider any criteria more important than intellectual achievement?

LaurentMelchiorTellier said...

Internal to the gymnasie, and using an average of exam grades and performance grades, resulting in an average over a large number of grades (a number on the order of your 40).

I believe there has recently been a reform such that GPA is (to a lesser degree) affected by the second, as well as the third and last year of gymnasie. But for some time, it has been only the last of 3 years which contributed to GPA. A high GPA is only really critical for very sought-after fields - when I first started in biophysics, it was a "new and sexy" major, requiring a minimum GPA which was out of proportion, due to the demand for spots. I recall midwifing being in extreme demand, requiring one of the highest GPAs.

MtMoru said...

Sounds just like the Canadian system or the US system without the aptitude tests unless the grading scheme is objective so that grades mean the nearly the same thing across gymnasie. But what is, the Higher Preparatory Examination?

In the US the quality of schools is extremely uneven and grades are assigned in as many ways as there are teachers.

How is the "class" of the uni degree determined? Like first class honors or second class or pass, etc.

I heard a Danish student say, "As long as I pass my exams I can keep the subsidy for students." An American student might have said, "As long as I can keep getting As I can keep my scholarship." An American student would never say "As long as I pass my exams..."

MtMoru said...

Aren't/weren't a lot of the po-mo bullshit artists Normaliens?

Regarding Sartre, "When the French think they think in German."

MtMoru said...

MIT has claimed that the SAT is the best predictor of performance. The "performers" like science contest winners, chess prodigies, or editors of the school newspaper, or IMO participants I believe are still a tiny percentage at even the most elite schools.

It bears repeating that the distinction between performance on objective standardized college entrance exams and psychometric g is nominal only.

MtMoru said...

The US is schizophrenic when it comes to what determines success, innate ability or striving, ignoring that it depends on what the society values. The result is an absurdly competitive society where the winners feel no obligation to the losers. The American ethos: "Play the game. It's the best of all possible games. Because it is a game there will be winners and losers. BUT if you've striven to win and lost it's because you didn't strive hard enough or because you're an idiot. Suck on it." Queue Star Spangled Banner.

This false ideology may still be closer to optimal for economic growth. Maybe.

as said...

Are you suggesting that Caltech and MIT are admitting students whom they know are intellectually inferior to other applicants?

LaurentMelchiorTellier said...

The gymnasium exams are objective, in the sense that everyone takes the same exams, which are graded by censors unaffiliated with the particular gymnasium. Similar to the Chinese 中考 and 高考, but nationally (rather than provincially) standardized, an exam for each subject.

After the 9th grade, some people go straight to gymnasium. And some, usually based on whether they've been "recommended" to go to gymnasium by their main teacher (every kid has one), take the preparatory 10th grade, prior to gymnasium. At the end of 10th, there's HF: the Higher Preparatory Examination.

Really, if you're dead set on gymnasium, the recommendation of the teacher, the 10th grade option, the HF, they all mean nothing. It's your choice. You can just skip 10th grade, as many of the most confident kids do. But a lot of kids don't. Human nature is such that young Danish teenagers generally take the teachers "recommendation", and the HF exam result, to heart, and mostly are not "grinders" who rush ever forth until they find themselves out of their depth/smashed into a hard ceiling. Amy Chua would contemptuously disapprove, I think. ;-) I suspect she's right, ambition is not a bad thing.

Uni gives grades, but generally no "honors". Some grades are high, some grades are passable. Some kids want high grades, some want passing grades, some want beer. Grading standards seem pretty consistent, though of course everyone knows subtextually that a good grade in math speaks to more than a good grade in art history.

Richard Seiter said...

Thanks for the additional details. Two questions if you don't mind. First, is a student's "main teacher" only for 9th grade (or different for each grade) or is this association for more than one grade? Second, do you see the same tendency towards grade inflation that I believe we do in the US (another way of asking this would be "what is the typical grade distribution")? It would be interesting to see a study comparing outcomes for those near the boundaries (to try to understand whether "grinding it out" is worth it). A lot depends on what alternative opportunities are available.

Richard Seiter said...

I'm not @MtMoru:disqus , but I think the answers to your two questions are different. For "do they consider any criteria more important than intellectual achievement?", I would say no. But for your last question I think the answer is yes (I would speculate more so for MIT than Caltech, but would be interested in other opinions). "Intellectually inferior" is hard to define precisely (IMHO), but there are other criteria (e.g. I'm pretty sure MIT takes at least some interest in recruiting athletes) and I suspect it is safe to say the admissions results aren't a strict linear ordering of test scores with a cutoff.

as said...

I agree that the admission results are not a strict linear ordering of test scores. In my opinion, that is a good thing because I doubt that intellectual capabilities can be linearly ordered with complete certainty. As an example here is a quote from the famous article by Richard hamming:"In mathematics, theoretical physics, astrophysics, typically brains correlates to a great extent with the ability to manipulate symbols. And so the typical IQ test is apt to score them fairly high. On the other hand, in other fields it is something different. For example, Bill Pfann, the fellow who did zone melting, came into my office one day. He had this idea dimly in his mind about what he wanted and he had some equations. It was pretty clear to me that this man didn't know much mathematics and he wasn't really articulate. His problem seemed interesting so I took it home and did a little work. I finally showed him how to run computers so he could compute his own answers. I gave him the power to compute. He went ahead, with negligible recognition from his own department, but ultimately he has collected all the prizes in the field."

Btw, I did not know that MIT took athletic capability in mind, but now that I think about it, it makes sense. I also like to think of admission selection as a problem of the Precision Recall curve. If one were to use linearly order test scores witha cutoff, then one will likely get a high precision, but lose out on the recall, but a little more flexibility can help you give you more recall, but you will lose some of your precision.

LaurentMelchiorTellier said...

Same main teacher for the last few years, I don't really recall if I had the same main teacher before 7th grade tho.

Re: grade inflation, I do recall two teachers, smartest I encountered in the gymnasium (high biology and high chemistry), who took comedic pleasure in generally giving grades lower than average, and making comedic displays of ignoring dramatic female ploys for sympathy, much to everyones amusement. Caustic bastards. Conversely, my physics teacher was eager to pat on the head with high grades, kindhearted and sharp guy. I don't recall seeing a distribution, but I'd guess it was slightly to the right of middle, with most of the worst grades given to no-shows. I have to imagine that the average grades were monitored, though Inever considered it at the time.

Richard Seiter said...

Thanks. I like the idea of having an extended relationship with a main teacher (in rare cases I have heard of US schools keeping teacher/class together for multiple years). I would worry that the influence of those teachers on the students future could be very high and hope they are selected carefully and that an effort is made to ensure teacher/student compatibility.

Richard Seiter said...

It sounds like we see things similarly. Thanks for the Bill Pfann anecdote (I hope we become better able to identify people like that. I wonder if work like Steve's BGI study can help.). The Precision Recall curve analogy is intriguing. I have to think about it more, but it sounds like a useful way to think about this.

Regarding MIT and athletes I should be clear that I have no inside information about admissions. I am basing my conclusions on my experiences as a (fairly mediocre) high school wrestler and people I saw/talked to at MIT (25+ years ago). I don't know how much preference there was, but I think of two otherwise equal candidates one who is believed able to play on a college team would have an edge (there are a number of other types of achievement/ability like this IMHO). The coach and some members of the wrestling team played a role (in a small but appreciated way, like providing a place to stay where I could interact with students) in my visit to MIT before making my decision where to attend, so there may be a difference in effort made to actively recruit as well.

Iamexpert said...

Probably a combination of high IQ, high IQ family and networks, and perhaps neurological conditions like aspergers which might facilitate creativity in high IQ individuals . I don't think their mean IQ of 110 alone is sufficient to explain it,otherwise where are are all the Hong Kongoid Nobel prize winners?

Emil Ole William Kirkegaard said...

HF which you mentioned earlier is one of the four types of student-exams (studentereksamen) aka. those that count as gymnasium level. HF is a 2-year version. The others are stx (normal), htx (technical focused), hhx (business focused). They are all 3-years.

I don't think we see as much grade inflation as you do. I used to follow the statistics (they are publicly available), and didn't see any major increases after the recent reform (in 2004).

There is no or almost no accelerated learning in Denmark. Couple this with the fact that people don't like other people who feel smarter than other people, this tends to make people less career and school minded. For this reason, many students take the 10th grade even though it isn't necessary. I did too. 10th grade is usually taken at an efterskole (means "after school", a kind of boarding school with a more free spirited setting), which was very nice, and an almost complete waste of time academically.

As for the money that the state pays you. As long as you take some classes and pass them at university, the state will continue to pay for a max of 6 years. Since this is 1 year more than the nominated time for completion (=5 years, for a master's degree), most students feel that they are in no rush and so purposefully delay their graduation so that they can work less hard and enjoy student life for longer - after all, why not?

Who knows how many money is wasted on this ineffective system. And how much potential is unused due to a lack of accelerated learning opportunities.

MtMoru said...

So the "performance" part is totally up to the teacher? If you teacher dislikes you too bad? But the exam is standardized and graded by "censors"?

The grades at uni are the same or are determined in as many ways as there are professors?

Is it like this:

As many as three midterms and one final in a semester, some may be take home, all are written by the professor and graded by him or an assistant, and homework, "research" papers, and lab reports all are unique to the uni and the class and graded by the instructor?

This is the US system. I think it's the same in Canada.

The standardized exam part already means Denmark is much better at selecting for g than the US. As mentioned above there is no such thing as enough for Americans and if you aren't rich you're a loser.

Amy Chua is a striver. There is still a small percentage of Americans for which "striver" is a term of contempt and "gentleman's C" one of praise.

MtMoru said...

They are admitting students who are intellectually inferior to some students who aren't admitted, yes, but they may not think they are doing this. Even at these schools there are other considerations, and admissions people trust their own subjective judgement over mere test scores.

MtMoru said...

All of my test scores SAT, ACT, CBATs (I left my shit hole high school for the local state u before sitting AP tests) were significantly higher than the mean for admits to MIT and SAT. MIT granted an interview, CalTech didn't, both denied admission.

I suspect my situation is common, and I know that in most countries it would be uncommon or simply wouldn't exist.

Richard Seiter said...

That is unfortunate. I now understand some of your comments better. Did you take the SAT pre or post 1995 recentering? After the recentering I think the mean for schools like MIT and Caltech became less meaningful because of the lack of resolution at the high end (do they make the median numbers available?). I would think the recentering made it harder to reliably identify top individuals (e.g. I have a tendency to make stupid mistakes on tests, this is deadly when there is a low ceiling for the test) and therefore added to the chance element in admissions. You've probably looked at the numbers more carefully/recently than I have, but I would expect the averages at MIT are impacted significantly by admits with significant non-academic strengths (less sure about Caltech) (yes, I realize this helps make your point). By any chance are you a member of an "over represented group" (relative to the MIT/Caltech populations, this can mean many things, e.g. when I was at MIT people from NYC were overrepresented and I think this raised their admissions hurdle)?

I agree about the "obedience" factor in the US, but I think MIT selected less for that than most places do ;-) I think it would be hard to separate out the obedience and drive factors in grades/etc. I wonder if the admissions officers try to do this.

MtMoru said...

It would be a good thing if it were possibefor admissions to do a better job. I don't think they can. There is also the down-side that "corrections" for test performance are subjective and opaque.

MtMoru said...

On the SAT I took both, but I sat the recentered just to compete with my brother. I haven't looked at the numbers.

I was/am a white gentile male with well educated but divorced and not so well
off parents. My dad graduated Harvard in '66, but if there is an
opposite of Amy Chua, my parents were it.

I can't claim to be a "victim". I think I can claim that things would have been easier/different for me had I (my genome, parents' status, etc.) grown up elsewhere. Some of the strivers I knew in high school might have had a harder time elsewhere. I knew two guys definitely smarter than me in hs. One went to G'town (Yale denied) the other to Harvard.

as said...

It does seem like we see things similarly. Occasionally, I wonder how the admissions process can better identify people like that and sometimes I feel that the folks at Cambridge and Oxford that used a combination of entrance exams, recommendation letters and personal interviews used to do a good job in admitting a wide variety of interesting students(along with a lot of legacy types of students).

As to MIT athletics, I did not mean to suggest at all that you had inside information. Since I am an outsider to American undergraduate admissions, I had been fed the stereotype that MIT was populated exclusively by uberubergeeks, I am glad to know that its not so.

as said...

I think that subjectivity implies a certain opaqueness by definition because if it could be done by a transparent algorithm, then it wouldn't be subjective. Subjectivity carries a certain risk, the question is whether the rewards from this subjectivity are worth the risk.

Richard Seiter said...

No problem, I did not take it that you implied anything, I just wanted to be clear with my statements. Don't get me wrong, MIT is pretty ubergeeky (speaking as a geek, many were less geeky than me, some more so). It's just that geeks are a more diverse and interesting group than many realize.

MtMoru said...

You're right. To say "objective and opaque" is like saying "scrofulous and ugly".

Blog Archive