Oh, I meant to say: I'm pretty sure there's a section in the Phenomenology about this, though I can't remember precisely where.

At the time I originally applied to college I had 3 close, living relations with degrees from Ivies I might attend. Unfortunately for me, they were all variously poor or stingy, appelations which -- also unfortunately -- could fairly be applied to the many (perhaps dozens?) dead lineal ancestors with degrees from the same institutions. Between that and reasonably crappy grades I successfully showed the lie of your first two proposed justifications.

Under the old order, when one attended college made a big difference. It is less so now. I know "everyone" gets all worked up about getting in to Harvard, but really, who cares? It's not like even after getting in one will be invited to join Porcellein, is it? Quality education can be had many places, so if someone wants to buy their way in, or great grandpapa donated the library, why get worked up?

I don't see why "bribe" is a word that people are utterly unwillingly to say in polite company. It's like "lie". But there are synonyms for "lie" that it is logical to use in order to not have dumb people think you are shrill. Whereas something is either a bribe or it isn't. Political donations by people who are trying to gain the attention of the candidate are not donations, they're bribes.

Alternatively, they could just publish a price list, prices varying with SAT and class rank. Say, SAT in the range 1400-1450, class rank 70th percentile: $500,000 guaranteed admission.

This is a Simpsons episode, when Mr Burns tries to get his newly-discovered (but idiot) son into Yale. The admissions officers whip out a poster board showing the required donation given the SAT score of the applicant, down to "New Dorm" for someone with very low scores. Then Burns hands them his son's test scores, they go into a huddle muttering "Never seen anything like this" and emerge to answer, "Sir, with these scores we'd be looking at the Yale International Airport."

Hah! I had forgotten that, Gonerill. I imagine a lowly blimp hanger would have gotten me in -- especially if I'd just thought to join the crew team in high school.

Anti-Semite. (The fruit, it hangs so low!)

Under the ancien regime one needed to attend an Ivy League school to be granted access to the levers of power in America (and to be honest, not all Ivies are created equal, and there is room for both small liberal arts and regional powers). To the extent that has been swept away, the selective schools now admit only those "others" who will in the new regime control those levers of power, and the ancien regime benefits from this new association. To the extent that America matters, this is important, but in the future will it still?

Political donations by people who are trying to gain the attention of the candidate are not donations, they're bribes.

An amusing set of contrasts might be drawn between the standards of the Foreign Corrupt Practices Act and domestic campaign finance law.

I might be giving the guy too much credit, but I believe that Larry Summers would have ended the legacy preference at Harvard within a few years if he had not put his foot in his mouth and gotten himself thrown out.

To the extent that America matters, this is important, but in the future will it still?

Given the preponderance of children of stratospheric foreign children at Harvard -- and probably many others -- and the overseas outreach being done by these schools, I'd say yes.

Children of stratospheric children?

Intriguingly, the one prestigious school where I had serious undergraduate legacy cachet (grandfather and great-grandfather, both well-ranked in their classes) was the one school that flat out rejected me.

Is it too obvious to suggest that such a sliding scale as proposed in the post might eventually be self-defeating, for if access to Harvard becomes sufficiently decoupled from merit, the value of Harvard would eventually decrease if it leads to the campus becoming filled with ever more spoiled ne'er-do-wells, and hence to employer satisfaction with its graduates decreasing?

12: yeah, that was weird, wasn't it?

I wonder what I meant?

"foreign children of stratospheric privilege"

14: you're forgetting that Harvard earned it's reputation as a campus of spoiled ne'er-do-wells.

My that sentence didn't work.

12: I dunno. Sounds like a prog record.

14: that's why you have to limit the number of slots for sale, and keep a healthy proportion of conventionally-admitted students (CAS). As I acknowledged.

After all, access to Harvard is already decoupled from merit to a certain extent, and everyone knows it, and everyone knows one knows, etc—it's just that no one likes to come out and say it.

First of all, I believe that U of C is significantly less competitive--in terms of percentage of applicants accepted--than other top-tier schools. Whether this is because the applicant pool is self-selecting or the U of C is running a scam, I leave to you to ponder. Second, and of course wholly unrelated, you wrote this:

expressing their gratitude to the university for the education they received their

Once we had kids we stopped ignoring the donation requests. Only tens of dollars per year, but enough so that we're marked "active" on their donor lists just in case. Only undergrad, though- I refuse to give to Harvard on the principle that in the time it took to call me they made more on their endowment than I could possibly donate.

As a Harvard alum, the linked article from 02138 made my freaking skin crawl. The legacies I knew at Harvard in the late 80s/early 90s were VERY smart, and I always just assumed that they got in on their merits with maybe a slight nudge for being a legacy. Now, I wonder just how much money their parents donated to the school before they were accepted. (I suppose that I should state for the record that neither of my parents attended Harvard. Or any other college for that matter.)

I decided a while ago that I would not suggest to my (future) kids that they think about attending Harvard -- this article from the NYT last year (which I can't recommend highly enough) pretty much settled it for me, and this article from 02138 just reinforces it.

And yes, Knecht Ruprecht, you're giving Larry Summers WAAAAAAAY too much credit.

Something wrong with prog records, oudemia?

When I finally start my neo-symphonic prog band, I intend to have a side-long song called "A Cock for Asklepios" just filled with lyrical insight and organ solos.

Yeah, there'll be lots of organ solos on the cock song.

First of all, I believe that U of C is significantly less competitive--in terms of percentage of applicants accepted--than other top-tier schools. Whether this is because the applicant pool is self-selecting or the U of C is running a scam, I leave to you to ponder.

I believe the first assertion's right, and that it harms them in the USNews rankings, and I choose to believe the former explanation.

I believe that U of C is significantly less competitive--in terms of percentage of applicants accepted--than other top-tier schools.

This is true. It's grind reputation precedes it and it is thus very, very self-selecting. They also have a terrible time attracting east coast kids. Northwestern is "harder" to get into than the UofC, but, having taught at both places, I think the UofC kids are much, much brighter. (Or maybe it's just that the NU students' tendency to be well-rounded and happy made them less interesting to me.)

I think that we are all in agreement. The future leaders of the world will still be taught at Harvard, just changing nationalities of those leaders thanks to sellout on the part of the admissions office. There have always been foreign princes in attendance, the new one's just don't have titles. And the middle class will have scant representation therein.

Its not It's=obvs.

First of all, I believe that U of C is significantly less competitive--in terms of percentage of applicants accepted--than other top-tier schools.

I believe the first assertion's right, and that it harms them in the USNews rankings

Wouldn't that make them not a top-tier school?

You know how hard it is to fail someone at Harvard? There are forms involved.

Northwestern is "harder" to get into than the UofC, but, having taught at both places, I think the UofC kids are much, much brighter.

As the facebook group says, Northwestern : Sophists :: Chicago : Socrates.

19: Maybe if you acknowledged things with a few less words we hard science majors would have less trouble understanding you.

28: only if you think that the percentage of applicants a school admits, or the US News rankings, have anything to do with that.

(Or maybe it's just that the NU students' tendency to be well-rounded and happy made them less interesting to me.)

No, they're mostly idiots.

Under the old order, when one attended college made a big difference. It is less so now. I know "everyone" gets all worked up about getting in to Harvard, but really, who cares?

Admission to a "name" college matters for a very different reason that it used to. It's not that the top Ivies and their ilk are the only path to wealth, power, or prominence, because there are plenty of examples to disprove that.

No, the value of the elite degree today is the head start into the upper decile of earnings, and insulation against downside risk. Our economy is characterized by extreme stratification and by lack of social protections. So the value of the elite degree--both in terms of signal value and social capital--is in getting your foot in the door to the best opportunities, and being sure that you can bounce back more easily if disaster strikes. Sure, there are probably graduates of Ohio Wesleyan or Colorado College working at Goldman Sachs, but Goldman Sachs isn't going there to recruit them. And how is it that Mathew Yglesias jumped into a series of high profile (though fairly unremunerative) media jobs after college, while the equally talented Kevin Drum had to be discovered serendipitously?

There is a forthcoming NBER study of a couple of Harvard cohorts that puts the facts in stunning relief:

The members of the H&B cohorts earn exceptionally large amounts, in general, and their distribution of earnings has a long right tail, particularly for the men. Median earnings in 2005 were $90K for women but $162.5K for men. Among full-time full-year workers, median earnings were $112.5K for women and $187.5K for men. Almost 8 percent of the men and 2 percent of the women had labor market earnings in excess of $1 million.

Put another way, the 50th percentile male Harvard graduate (from these classes) earns as much as the 98th percentile American.

So yeah, if you are exceptionally talented and hardworking, and you get a couple of lucky breaks, you can attend State U. and slowly claw your way to the top. But you can also be a run-of-the-mill Harvard graduate and just sort of coast there. I know what I'd choose.*

*subjunctive voice is technically not called for.

That's "a few fewer words", Otto.

pumps there a heart so invidious as to deny that this has intrinsic value

*holds fingers against wrist*

Yes.

32: considering the school I attend I would be unlikely to think that.

Personally I think the Newsweek Top Science School and Washington Monthly rankings are vastly more important to determining a school's quality.

I went to what really might be the most self-selecting school of all with a pretty esoteric and essay-heavy application. We're far too odd to be a "top-tier" school, but almost everyone I went to school with was super smart.

Even rob?!

38: Wrongshore has you beat on the "most self-selecting" thing.

Would I be guessing your school correctly if I said it technically is an Ivy?

I probably shouldn't bother linking, since you all already saw it anyway (as we're all Atlantic-reading whitefolk), but I'll state just in case that "The Old College Buy" in the 9/07 Primary Sources (scroll down about 80%) was highly amusing and is apropos to this discussion.

To 40, based on 39, the answer is probably no.

40: In re: Wrongshore. Sleep Rings?

In re: oudemia. No. There is red brick. There is ivy. But definitely not Ivy.

One has to also factor in the use of a form of legacy preferences in graduate school admissions. Once you get admitted to the undergraduate institution, you're at a significant advantage for, say, law school admissions. One guy I knew was told by his tutor that he was a sure thing as long as he didn't totally screw up his LSAT (for the Harvard value of totally, of course). I presume that grads with no divided loyalties give more.

pumps there a heart so invidious as to deny that this has intrinsic value? Surely not.

::Raises hand.::

The legacies I knew at Harvard in the late 80s/early 90s were VERY smart, and I always just assumed that they got in on their merits with maybe a slight nudge for being a legacy.

That's the argument that Harvard always made: our alumni children come from families where education is valued, so of course they are going to be admitted at higher rates than the general population; they're just smarter. And indeed, you have any number of legacies who fit that mold: children of professors, that kind of thing.

But on average--and a Department of Justice study of Harvard's admission records* proved this--admitted legacies score statistically significantly worse than other admitted applicants on every dimension Harvard measures (academic ranking, test scores, extracurriculars, personal qualities) with the exception of athletics.

*Why was the DOJ peeking around in there? Because Asian American groups charged that they were being discriminated against in admissions, and Harvard defended the claim by arguing that the discrepancy was explained by the underrepresentation of Asians among legacies.

favoring a legacy applicant over someone who is otherwise more qualified

Why do many liberals have no problem whatsoever making statements like this, but go absolutely apeshit over the very idea that there could be such a thing as objectively measurable qualifications whenever someone objects to affirmative action on parallel grounds?

This is a geniune question that is not meant to be troll-y.

the one prestigious school where I had serious undergraduate legacy cachet was Sac State. And I don't give money to any of my alma maters. PK's just gonna have to use his own bootstraps, dammit.

(Not counting the fact that his mother has a PhD, his father makes enough money that he's got access to any "supplementary educational program" we want to bother with, he's in a "good" school district, and his parents are hyper over-involved in his education. But other than that, bootstraps!)

47: I don't know exactly who you're talking about, but the difference in finding legacies objectively worse is that you can pretty much guarantee they had all the educational opportunities and access that can be had, whereas with minority students you can almost guarantee that they didn't, so comparing a brilliant non-legacy, but white and upper-middle-class student to a non-brilliant legacy student is an apples to slightly moldier apples comparison, whereas comparing the brilliant, white, upper-middle-class student to somebody who went to crappy ghetto schools their whole life is much more fraught.

46: More evidence that my unborn daughter will be attending UNC Chapel Hill. That, and the fact that I can't afford to donate $1 million to ensure her acceptance.

47: Three letter answer: GWB.

Apropos speech.

47: Quick and sloppy answer? Because diversity of race and class, etc., benefits the university community as a whole, while the attendance of legacies does not?

52: some would argue otherwise. (About both points, I suppose, but I'm focused on the latter and thinking about both networking value and on donations).

But 49 seems right.

But 49 seems right.

Despite being terribly written, that is.

Serious answer to 46: this liberal, at least, kind of flinched at the phrase "otherwise more qualified," actually. That said, and as my parenthetical in 48 implied, a kid whose parent(s) have been educated at Harvard--or whose parents have advanced degrees from non-ivies, for that matter--has a pretty significant leg up over a kid whose parents went to oh, say, Sac State, generally speaking. So assuming that the two kids appear to be of roughly equal intelligence and academic accomplishments, the child of Sac State alums--or of parents who didn't go to college at all--has probably worked harder to get there.

Largely because the people arguing against affirmative action are totally okay with little EmmaJacob getting in because he was a legacy admit, but not okay with ConsuelaTyrone being admitted based (in a way) on who their parents were. In other words, it's not so much with the test scores, as it is thinking that test scores show objective merit only when it comes to excluding black kids.

I have an irrationally deep hatred of legacy admission policies, mostly because I knew a lot of legacied idiots in college. And KR has said what I would have: if your family is all Harvard alums, you're probably well-connected enough in life that you'll be fine. It's the non-legacy kid for whom there's a big difference in social capital/higher-ed options between Penn State and UPenn.

whereas with minority students you can almost guarantee that they didn't

Actually, this part of 49 seems not right, and does go back to the issue of whether we shouldn't be structuring affirmative action programs on income triggers rather than racial categories. But I don't want to divert the discussion too much.

55: I flinched too, which was what prompted my comment.

I was a multiple legacy and an athlete. That is how I got into college.

They must not have looked closely at the legacy part though. Although there is a dorm with my last name on it, nobody close to me had donated much money.

Plus, I was not that good of an athlete.

Hey! Maybe I got in based on my blazing intellect! Or maybe not.

Also, is no one discussing Ben's proposal? Selling 5 or 10% of each class at auction sounds crazy, but actually seems like a decent approach. (Probably would need to require some minimum qualifications for admittance even in light of the auction, but the basic idea is okay.)

What I was really thinking when I wrote that thing about "otherwise more qualified" wasn't that there are objective measures, but that you could imagine structuring the applicant pool into equivalence classes, such that you can rank the classes relative to each other, but not the members of one class relative to other members of the same class (because, you know, otherwise they wouldn't be equivalence classes). Then, assuming that legacyhood hadn't already entered into the formation of the classes—which needn't have been formed based on quantitative measures, mind—you could prefer legacies within a class to nonlegacies in the same class. The "upholding the tradition" argument seems to offer a justification for at least that much. It is another thing entirely whether the value of preserving traditions (the tradition of attending Harvard? let's be honest: that's kind of a lame tradition, but, whatever, sense of history blah blah) justifies lifting someone from one equivalence class into the next-higher-ranked—i.e., "more qualified"—class.

Maybe some will still find this objectionable, I don't know.

More evidence that my unborn daughter will be attending UNC Chapel Hill.

You're going to be crushed when she ends up at NCSU, aren't you?

Some will no doubt want a justification for excluding legacyhood from the formation of the equivalence classes if other things, like geographic diversity, or poverty, or being the first one to attend college in one's family, etc., are going to get to count, but, eh. If you're a legacy, you probably aren't hurting too much.

Not nearly as objectionable as your use of "shekels."

Especially if you're a legacy whose parents can make megadonations.

62: Go wash your mouth out with soap RIGHT NOW!!!!!

51: I thought Michelle was pretty good on Colbert.

Selling 5 or 10% of each class at auction sounds crazy, but actually seems like a decent approach.

I disagree with Ben's approach for a variety of reasons, but I must concede that auctioning 5%, or even 10% of the class would actually reduce the dilution of academic standards vis-a-vis the current legacy approach, under which something over half of admitted legacies (or north of 10% of each class) would not have been admitted but for their ancestry.

Perhaps I am a bitter legacy nonadmit, but I still say who cares? I am not trying to deny that Chet Saltonstall IV, Exeter, Harvard, HBS, Goldman Sachs has had the way smoothed for him, but what I am saying is that it is folly to pretend that it matters. These are private colleges, with enough money to tell the legacies to fuck off, which they frequently do. No one has a "right" to go there, even super smart brianiacs.

My sister got into NCSU for grad school. I don't think she's going there, though.

I have an irrationally deep hatred of legacy admission policies, mostly because I knew a lot of legacied idiots in college.

I have an irrationally deep jealousyhatred of legacy admission policies because no one in my family had gone to college in my parents' generation on either side and in their parents' generation many of the colleges attended no longer existed. My family suffered a downward spiral of education as a priority over the course of the 20th century. Legacies are one of those things that trips the class-warfare switch in me with absolutely no intervening consideration; it's a purely defensive overreaction.

57: Actually, this part of 49 seems not right, and does go back to the issue of whether we shouldn't be structuring affirmative action programs on income triggers rather than racial categories.

It's increasingly less right, and as such would eventually constitute an argument that the system has worked, and is no longer necessary on those grounds. On the other hand, income triggers aren't a particularly accurate way of measuring educational opportunity either, and it is absolutely true that racial and ethnic diversity on college campuses is a net benefit. Further, it's not like things have reached a state of full post-racial equality.

I had to leave this comment, and without previewing estimate I will have been pwned 7 times.

One issue about Ben's proposal is that while it would squeeze more money out of the eventual winners, the schools would lose out on some "donations" from a bunch of losing parents, who otherwise would have given some money to the school in hopes of pushing EmmaJacob (nice, Cala) over the top. It seems unclear to me which system would be better for the school overall.

who cares?

Well, for one thing, there's Sifu's point that with minority students you can almost guarantee that they didn't. Brock doesn't buy it, but I mean, think about it: a couple of generations ago, most colleges actively discriminated against brown-skinned applicants, and knowing this, most brown-skinned smart people didn't bother to apply in the first place. So one of the reasons that legacy admits matter is that they perpetuate classism and racism, both of which violate the American ethic. (In theory, at least.)

That and frankly the "yeah, well, what about the fucking legacies?" argument is largely backlash to the anti-affirmative-action argument. It's irritating to have people bitch and moan about the "unfairness" of admitting minorities (and sometimes of admitting white women), when those same people are silent about legacy admissions, veterans preferences, and preferences for the disabled--all forms of a.a. that just happen to include white men.

Sorry! Humorless again. I should eat more lunch....

brianiacs

They are a small but devoted group.

If we tried to auction off 5% of the spots of our incoming class, we'd soon find that no one came close to our minimum bid on E-Bay, and we'd be forced to wrap it up with a spa gift certificate, put it in a basket, and donate it to charity auctions. We're having recruitment and retainment problems.

it is folly to pretend that it matters. These are private colleges, with enough money to tell the legacies to fuck off, which they frequently do. No one has a "right" to go there, even super smart brianiacs.

TLL, need I remind you that you made a Faustian bargain earlier today to join our side in exchange for a mess of pottage? Bwwaaahhhaaahaaahaa!

Seriously, no one is arguing that the colleges don't have a "right" to admit whom they want (although I think the disparate impact angle *is* underlitigated), or that anyone has an entitlement to be admitted.

There is a different between "rights" and "right", as in "right and wrong". And it is not right for a society that derives no small part of its legitimacy from its purported equality of opportunity to accept the existence of hereditary sinecures in an ostensibly merit-based system. Thoughtful college Presidents understand the extent to which their admissions committees are the gatekeepers to prosperity in age of "tournament style" economic competition. I think it is only a matter of time before the legacy preference goes the way of the Jewish quota.

It seems unclear to me which system would be better for the school['s finances] overall.

They aren't the same.

If you removed the three most corrupt aspects of higher education--legacy admissions, college sports, and fraternities--most elite institutions would collapse.

The surviving institutions, the ones that would actually do the work of giving young people the knowledge, skills and values needed to keep human society functioning, would most resemble community colleges. They would cover all levels of education, but their mission, values and institutional culture would match that of community colleges.

Fraternities?

If you removed the three most corrupt aspects of higher education--legacy admissions, college sports, and fraternities--most elite institutions would collapse.

MIT says you're wrong.

69: Right, it won't matter for Dexter Winthrop III, but I think a lot of the "oh, Harvard just doesn't matter these days" is more people being annoyed that Harvard isn't as exclusive and is letting in the wrong type.

72: Plus, as I understand it, the myth of the ghetto kid getting in over some Andover kid because Harvard wants a diverse student body is just that, a myth. Affirmative action's worth what, 100-200pts, max, SAT points?

Racial diversity is already thrown in there with geographic and economic diversity, athletics, and still 90% of them are too rich to benefit from the tuition waiver.

I was going to say "nonsense" to 79, but I've already been multiply pwned.

whether we shouldn't be structuring affirmative action programs on income triggers rather than racial categories.

My experience is that a.a. programs attend to racial categories *and* income *and* whether or not one's parents went to college. And sex, and disability status, and veteran status, and region, and which high school people went to, and whether or not English is an applicant's first language. Most of the time those programs are pretty darn complicated.

I think it is only a matter of time before the legacy preference goes the way of the Jewish quota.

I doubt it, somehow. I'd love to know if legacy admits perform worse because good legacy admits get in and go elsewhere. Has anyone examined that?

81 says that Sifu doesn't know how to assess claims involving "most".

MIT isn't exactly post-fraternity. They may wish that they were, and there have been steps in that direction, but they're still a major force on campus. I don't know if they're a corrupting influence on admissions, and any information I would have had on that is a decade out of date.

I'd love to know if legacy admits perform worse because good legacy admits get in and go elsewhere. Has anyone examined that?

See here:

How do those kids fare once they're enrolled? No one's telling. Harvard, for one, refuses to keep any records of how alumni children stack up academically against their nonlegacy classmates--perhaps because the last such study, in 1956, showed Harvard sons hogging the bottom of the grade curve.

86: there's a contrary model, is all I'm saying. The ratio of academically rigorous, competitive universities to schools fitting those criteria who rely on legacies, sports, and fraternities to maintain their reputations is some very large number to one. Perhaps the elite institutions would have to adjust, but it's not like there's no model for a well funded elite school which doesn't rely on those three crutches.

87: can't you let my self-serving elisions be?

I put MIT frats in a different category, which may be a mistake.

86-87: OK, take Harvard, Yale, and Princeton: No fraternities, lousy sports teams (in part because of comparatively high academic standards and no athletic scholarships), and they would all survive the elimination of legacy preferences with a barely noticeable scratch on their finances.

82: Plus, as I understand it, the myth of the ghetto kid getting in over some Andover kid because Harvard wants a diverse student body is just that, a myth.

Signs point to yes.

91: it's a little bit of a stretch claim that Yale has no fraternities. Yes, I know they're called secret societies, but come now.

More evidence that my unborn daughter will be attending UNC Chapel Hill.

You're going to be crushed when she ends up at NCSU, aren't you?

To complete some obscure trifecta, I'm going to be visiting Duke next week.

The question of whether Harvard would actually make more money by this auction system is an interesting one, but I would bet pretty good money that the answer is no.

There are a few main factors to consider:

First off, what percent of donations are actually given to the institution as gifts qua gifts versus as a finger on the admission scales for offspring? It is certainly non-zero, as I myself have donated to my alma mater despite a fairly low chance my child would go there (though I myself was a legacy of sorts, so who knows). But if it were nearly 100%, chances are that the reduction in those generous gifts due to disgust at the open selling of places would easily swamp any effects of higher earnings on the bribery side. Thus, let us assume that a significant portion of donations are currently given to help a child in admissions (an extracurricular, if you will).

The current system has many people paying smaller amounts for an uncertain shot at getting into Harvard, while the proposed system has far fewer people paying much larger amounts for a guarantee. As people are known to be generally risk-averse, that would suggest Harvard would make more by selling guaranteed slots for lots of money than uncertain probability increases for less money. Indeed, this would be the case if people had complete information on how likely their child is to get into Harvard and how much a donation of any given size will help those odds.

However, the other well-established bias is overconfidence. Those donors probably far overestimate their legacy's chances of getting into Harvard with a sizable donation, and thus are more willing to part with their hard-earned cash than they would be if they truly knew their 1400 scoring kid had a 2% chance even with their $10,000 kicker. Pull in a large pool of donors making this systematic error, and you keep pulling in far more money than if the process had full information and you instead sold certain seats for their proper market price to a (relatively) small number of individuals.

I bet the overconfidence dominates the risk-averseness overall (it always seems to), and thus Harvard (and similar top schools) makes more from bribe-style donations under the current system than it would with more certainty and transparency. Plus, it gets to keep a measure of plausible deniability and thus a better reputation.

91: it's a little bit of a stretch claim that Yale has no fraternities.

Princeton, meanwhile, has its selective eating clubs.

Fraternities are corrupt for largely the same reasons that legacy admissions are corrupt. They exist to promote cronyism. In fact, cronyism is the *explicit* justification for fraternities.* The important impacts of fraternities are (in order): binge drinking, hazing, date rape, and plagiarism. (The ranking here is a combination of frequency and heinousness.)

Given that this is all fraternities accomplish, I count them as among the most corrupt aspects of higher education.

* Am I being sexist for excluding sororities? No. They have a different original justification and a different impact. The original justification was to keep the *good girls* busy so they wouldn't have sex before they got their mrs degrees. Their current impact is negligible, since most places sorority girls tend to be good university citizens, just as female students in general do.

I'm not in a good position to comment on this, since I'm sure that Harvard connections/ being a legacy got me in there. I think that being interested in a less popular degree program helped too, but that's fraught with class stuff too. Having preparation in Latin and especially in Greek means that you probably went to a private or very academically challenging public school.

If I hadn't gone to Harvard, I'd have gone to Yale or Wellesley.

I'm told that Swarthmore is similar to the University of Chicago in having a high percentage of applicants accepted for its caliber of school. It's known as being a place for grinds too.

97: then there's Apo's fraternity, which seems to have existed to promote (in order): use of hallucinogens, sexual experimentation, procreation, and hilarious anecdotes.

it is not right for a society that derives no small part of its legitimacy from its purported equality of opportunity

Oh, mission statement stuff. I get it.

Seriously, like in the walk on article, pulling a Captain Renault on the fact that the rich have advantages others do not doesn't particularly move me, but then again I am not in higher education. There are what 1000 freshman class spots at Harvard? So maybe 10,000 slots for all "elite" schools? Not large numbers, no matter who is admitted.

MIT says you're wrong.

Since when does MIT not have college sports? Their legacy policies weren't broad enough to get me in, but I was a crappy student, so I don't know how much that says. And my buddy who went there was in a fraternity (I believe his position was "worthy keeper of pornographic literature") that seemed to do typical fraternity-like things.

Sororities certainly had no shortage of of binge drinking, hazing, or plagiarism when I was an undergrad in Chapel Hill.

93, 96: Yes, and Harvard has its Final Clubs. But none of them dominate campus life to the extent that fraternities do at your average landgrant U.

PMP's point in 95 is probably correct, and one of the many reasons to doubt Ben's scheme. The current system works for much the same reason that the Dubai airport can profitably raffle off Ferrari's at $100 per ticket; for the relevant clientele, the wager is small enough, the payoff attractive enough, and the odds of winning high enough that people will take a punt, even though it's statistically a losing bet.

91: Allow me to introduce you to final clubs, secret societies, and (to a much lesser extent) eating clubs.

then there's Apo's fraternity

With chapters at several of the post-fraternity Ivies mentioned above.

"The important impacts of fraternities" s/b "the *other* important impacts of fraternities"

104: So pwned.

101: huh, I had thought MIT didn't confer advantage to legacies, but I guess I was wrong.

Okay, uh, Berkeley! Berkeley says rob's wrong. Except about the sports. And the fraternities.

My school doesn't consider legacies, and fraternities are a negligible presence, but except for certain very specific programs it's not top tier, so never mind, then.

"The important impacts of fraternities" s/b "the *other* important impacts of fraternities"

109 to 106.

Berkeley! Berkeley says rob's wrong.

I don't know if this is still the case, but Berkeley used to give legacy applicants the courtesy of considering them with the in-state California pool rather than the much more competitive out-of-state pool, irrespective of whether they resided in California.

91 and 104 et al.: For a little more than 10 years there's been an unrecognized fraternity at Harvard. There was also a sorority, but it was not at all high status.

Recent conversation between me and my mom (a graduate of rob's and oudemia's alma mater) upon my seeing yet another letter from my own alma mater asking me to donate:

Me: They keep trying to get me to join the alumni association, but I'm not going to. All you get is a subscription to the alumni magazine and some discounts at places in [college town].
Mom: Well, when I give money to the alumni association I don't think about it in terms of what I get in exchange. It's more about supporting the school.
Me: Yeah, but [my alma mater] has lots of money; they don't need mine. For you it's a little different.
Mom: Yeah, I guess so.

Yes, and Harvard has its Final Clubs. But none of them dominate campus life to the extent that fraternities do at your average landgrant U.

At Princeton, the Eating Clubs pretty much define the beer-drinking/night-life of the university. Only a relatively small number of students choose not to become members, and plenty of students try really hard to get into ("bicker") the exclusive ones.

In fact when it comes to athletic admissions, Harvard and the other "Ivies" (which are formally bound together only as an athletic conference) do something very close to the auction that Ben posits. As I understand it there is basically an agreed to formula which allows a certain number of "recruited athlete" slots to be filled within ranges of SAT scores expressed as a percentage of the typical score for the school.

Hint: this was not a detailed policy proposal.

I donate in years where I want to exercise my football lottery ticket privileges.

Yale has actual fraternities too, though the secret societies are probably a more prominent force. Though, they're much more ignore-able than eating clubs.

The basic reason legacies persist is pretty obviously financial. I mean, there are legacies & legacies. I might have gotten a boost because my older sister went to the same school, but honestly I think I would've gotten in anyway. Then there was my college roommate whose family had donated scads of money, whose father used to play squash with the dean, & who called the school & got her in w/o going through the formal application process after she got unexpectedly got rejected from a much lower ranked school. Very nice girl, but it was pretty blatant.

The geographic thing was real too, & I'm not sure why it's more objectionable to think that your campus shouldn't be all white than to think that it shouldn't be all from the tri-state area.

They'd probably prefer you donate in dollars.

Apo's and my (and KJ's and dob's and froz's) fraternity also served as a pretty important aid and encouragement for me to get back into and finish school.

We are really a piss-poor stand-in for "typical fraternity," though.

120: Rah wasn't in the frat? And here I thought the incestuousness was total.

"Friend of the Hall," not formal member. He knows as many people at the alumni weekend as I do.

Apo's and my (and KJ's and dob's and froz's) fraternity

The thing is, you guys almost make being in a frat sound like fun.

Sororities would be a fun institution if I didn't loathe the girls who composed them. It's sort of what McManus was saying about personalities coming to dominate the field. If I could pick my own sorority, we'd have the time of our lives.

The thing is, you guys almost make being in a frat sound like fun.

It is fun. Best years of my life.

Apo's and my (and KJ's and dob's and froz's) fraternity also served as a pretty important aid and encouragement for me to get back into and finish school.

I think that a fair number of fraternities made sure that their members studied and stuff in the 19th century. I'm pretty sure that some of the fraternities at Hobart College in Geneva, NY had lending libraries before the college did.

I presume that grads with no divided loyalties give more.

anti-Semite.

It is fun.

Yes. Why might this not generalize?

The geographic thing was real too, & I'm not sure why it's more objectionable to think that your campus shouldn't be all white than to think that it shouldn't be all from the tri-state area.

There is some ugly history around the sudden enthusiasm for geographic diversity preferences at precisely the historical moment when Northeastern Jews started winning admission in alarminly large numbers. Which is not to say that the policy is all bad: A. Lawrence Lowell understood very well that if Harvard was going to continue to be the proving ground of the American elite, then it would have to assimilate the elites of the booming regions outside the Northeast; James Bryant Conant expanded the scope to seek out the non-elite "diamonds in the rough" from obscure corners of the country. Neil Rudenstine was taking a page out of Lowell's book when he started heavily recruiting from the sons and daughters of the international elite, and Larry Summers was imitating Conant when he emphasized finding the brightest talents from the global masses.

Which is not to say that the policy is all bad: A. Lawrence Lowell understood very well that if Harvard was going to continue to be the proving ground of the American elite, then it would have to assimilate the elites of the booming regions outside the Northeast

The connection between these two sentences seems to depend rather heavily on looking at things from Harvard's point of view.

The whole fraternity/social club thing is silly. At college people form social groups because they're fun. So any successful school is likely to have them. That doesn't mean they're required for the school to be successful.

CalTech is the sportless, fraternity-less version of MIT. And it's a worse school for it. Having social events and a reasonable level of athletic activity are valuable things to have in your life.

116: Hint: this was not a detailed policy proposal.

Do tell.

I actually believe that at the really "top" schools you could model admission chances pretty accurately as the expected change to the net present value of the endowment using a very low discount rate (they are far-thinking at least). This function will be evaluated with all of the nuance you would expect from such an institution, broad enough to satisfy many constitutencies, so that you get* the truly academically superior; the rich, contributing legacies (and very,very rich non-legacies); the children of the very well known and powerful; athletes; a few of the go-getter $100,000 spent on the application type and enough racial and income diversity to mollify the self-image of various stakeholders in the institution. If done well it is self-reinforcing.

*All of these categories evaluated as appropriate for the perceived "desirability" of your university.

What do you mean, sportless?

http://www.athletics.caltech.edu/

130: It's also the girl-less version of MIT, which probably hurts it more. MIT has worked extremely hard to get more girls in the incoming classes for years, and has a roughly 50-50 mix to show for it.

CalTech realized it has a serious problem with its 70-30 split, but it's having a hell of a time changing things. When they have the orientation each spring for accepted pre-frosh, CalTech will fly and accomodate every accepted girl for free while guys have to pay (except for very few extremely desired students). Still, as my high school friends came back from their free weekend in NorCal, they all said "yeah, that was fun, but now I'll accept MIT's offer (or Olin, or Carnegie Mellon, or UofI, or UofC, etc.)."

The connection between these two sentences seems to depend rather heavily on looking at things from Harvard's point of view.

Is there any other point of view worth considering?

131. I thought that's what they did already. The admissions interview is just a smokescreen.

133: Mightn't it be cheaper for Caltech to just offer male students a financial enticement to undergo gender reassignment surgery? Alternatively (taking inspiration from Ben), they could relax the admissions standards a little for males willing to take the deal. Given the reputed sexlessness of the campus, it might even be possible to forego the surgical intervention and make do with estrogen injections and cross-dressing.

Of course, I realize this would never work in practice: Fox News would have a field day with a public university using taxpayer dollars to pay for sex changes.

Po Mo, Cal Tech is in sunny Pasadena, CA. No where near NorCal.

Dammit, knew I should've just gone with California instead of guessing.

Weird though, I didn't think California had any great schools outside the Bay Area.

a public university

It's also private.

138. Very funny.

I didn't think California had any great schools outside the Bay Area.

You're trolling, right?

I'm fairly certain the UC Regents - so it could apply to all of the UC campuses, not just Berkeley - have the privilege of influencing a certain share of the UC applicant pools. Whether this is called "legacy" or not, I don't know, but I remember reading about it when affirmative action was more in the news (and when I paid more attention to higher ed than I do today).

140: Don't worry Teo, Davis will be in the Bay Area by the time you go there.

All I know is that if PK can get into pretty much any of the UCs, I'll be quite a happy mama.

He thinks he's a fan of USC at the moment, though, thanks to his awful first-grade teacher (herself an alum). I'm hoping he'll get smarter as he gets older. Certainly with current UC admissions trends (pointed out to me a while back by Tweety), if he doesn't, he'll never be accepted as an in-state applicant.

I bet he can get into UC-Merced.

USC doesn't charge tuition to children of staff members, I believe (hint). I hear they're also making a huge push to upgrade themselves academically (that is, they're looking to pick up top faculty from other schools with generous offers).

139, 140: Yeah, kinda. I mean, don't get me wrong, my perspective on higher education is utterly screwed up. My safety school was MIT. So, basically, I can't really talk about good or bad schools without trolling.

My "legacy" was of no use, as Yale didn't go co-ed till after I'd gone to college elsewhere. The only benefit it gave my little brother was that he could skip one interview, given that he was a well-rounded, valedictorian National Merit Scholar with 1597 SATs.

He thinks he's a fan of USC at the moment

And you think that they waste their money on that football program? If a hippie kid in Ventura thinks it's the place to go, call it money well spent.

143: Tell him that USC is stands for "University of Stupid Children". That may put him off.

is

I thought it was Spoiled Children.

USC = University of Second Choice.

145. Here's your test case in real time. USC is a big legacy school, as befits a regional university. If it is truly trying to step up its academic game, most current admits won't stand a chance. USC has a huge endowment, well into the billions, so it doesn't depend on current donations to run. Can it overcome its reputation?

148: LOL. Goddamn their education majors, more like it.

I'd try the "spoiled children" line, but it would only solidify his belief that it's the school for him.

I bet he can get into UC-Merced.

Actually, I really like that campus's mission. I'd be very happy with that. Though getting him away from the beach might be kind of tough.

I thought it was Spoiled Children.

That's how I heard it, too. Pretty unfair, from those I know.

My dad went to a high school in New York that fed a lot of people into Ivies and liberal arts colleges. He wanted to get out of the east and went to USC for a year, then transferred to UCLA.

132: Meh. Things have probably changed at CalTech over the last 10 years, but I would imagine that there still are far fewer athletic opportunities at the school and fewer students participating. But that's not the point. The point is that the sports program at MIT is what you should want to see in a school- a sports program designed for the participation and benefit of serious full-time students. Not unpaid pros masquerading as students.

I'm just arguing that 79 is not correct/missing the point.

OTOH, if PK goes to USC, my grandma will be smiling down from heaven. For god only knows what reason, she was a fan (I think she actually went to UCLA, herself). Stupid football. At one point their goddamn car horn actually played the Trojan fight theme, I shit you not.

More offensively:

UCLA: University of Caucasians Lost among Asians.

I would stand up for my school, but that would involve revealing which school it was. Suffice to say it's better than MIT (or anyone, really) at the few things it does best, and it's not in the bay area. So nyah.

A lot of the UCs are better schools than I think people out-of-state necessarily realize. Even UC-Merced.

A lot of the UCs are better schools than I think people out-of-state necessarily realize. Even UC-Merced.

Well, of course. Even UC-Merced has more Asians than white people.

"few things" in 159 probably should have been "one thing", and even that is probably mindless boosterism on my part. It's up there, though.

USC, film school?

UCLA: University of Caucasians Lost among Asians.

I heard it as Caucasians Living among Asians. There's also University of Chinese Immigrants and U C uS Booze.

The UCs are excellent schools. Despite cost-cutting. The Cal States are pretty good as well.

161: If that one thing is its women's bowling team, I think I know which college you mean.

All the California schools try to be as offensive to each other when they can.

162: Huh, good point. I can see PK doing that. I'd rather he went to UCLA, though, b/c it's cheaper.

165: looking at the rankings, no California schools even rate.

166: That's because we're Californians, and shit-giving comes with the territory.

"... Of course, none of this goes so far as to justify favoring a legacy applicant over someone who is otherwise more qualified, or regarding whom there are other familial or historical reasons to want to extend some generosity to. But we know that there are lots of cases in which you can't really judge from the substantive application material which of two candidates is better, and you might want to look to legacyhood there. Fine."

This is silly, you can always rank the applicants. A small difference is not the same as no difference. You are really saying it is ok to give legacies a small advantage but not a large advantage.

167.Then get a job there, B. Free tuiton

All the California schools try to be as offensive to each other when they can.

My wife still has her "my maid went to UCLA" button. Very few occasions to wear it.

I love the NCAA rankings for things.

If asked, "What do Oregon, Wisconsin, Alabama, Colorado, Minnesota, Louisville, UTEP, Oklahoma State, Northern Arizona, and Iona have in common?" You'd probably say "Wait...Iona? Probably some weird sport or other, but...Iona?"

And yes, that's the top 10 ranked men's cross-country teams.

Why Iona?

171: Eh, it's too far to drive. If/when I get my act together and start applying for teaching gigs again--any day now!--it'll be at CSUCI.

my maid went to UCLA

Holeeee crap.

This is silly, you can always rank the applicants.

Objectively. By looking at their IQ scores.

A small difference is not the same as no difference

But an unmeasurable difference is effectively the same as no difference, so giving sufficiently poor measuring tools, they may end up being the same after all.

61

"What I was really thinking when I wrote that thing about "otherwise more qualified" wasn't that there are objective measures, but that you could imagine structuring the applicant pool into equivalence classes, such that you can rank the classes relative to each other, but not the members of one class relative to other members of the same class (because, you know, otherwise they wouldn't be equivalence classes). ..."

Assuming you aren't talking about a bunch of equivalence classes with one person in them this is wrong. Applicant merit varies continuously, there aren't big jumps. Similarly for grades, the difference between the top B student and the worst B student will generally be greater than the difference between the top B student and the worst A student.

Back when football was played by amateurs, we did ok against the big guys

Throughout their long football history, the Sagehens hold an impressive record of 6-5-1 versus UCLA from the 1920's. From 1897 until 1925, Pomona College played USC on a yearly basis, including the 1923 Dedication Game of the Los Angeles Memorial Coliseum.

In case you were wondering whether there is a school ranked in the top 5 for both their women's bowling team and their rifle team, the answer is yes. University of Nebraska. They'll beat out those West Coast schools for that trophy for best overall athletic program yet!

Here we see the unofficial rankings for top rifle programs as of a year ago. At the top you see UAF and the US Military Academy. No, UAF isn't the Air Force Academy, it's the University of Alaska-Fairbanks. Of course!

I actually went to the U Nebraska for my MA, but all I ever did was swim. Had I but known they had an award-winning rifle team....

Applicant merit varies continuously, there aren't big jumps.

Whaa? Do you give grades as a part of your job? Have you never seen a pool of individuals with clusters in it by quality?

Similarly for grades, the difference between the top B student and the worst B student will generally be greater than the difference between the top B student and the worst A student.

Whaa^2?

If you graded on a hundred point scale and then converted into the cruder ABCDF scale, then you can assert this. But given no other prior knowledge of the pool you are looking at, I see no basis for this.

179: My grandfather played football for Nebraska in 1902 or so. Small world, ain't it?

That was your wholesome period, I take it.

How dare you imply that her wholesome period comprised 1902.

157

"... Things have probably changed at CalTech over the last 10 years, but I would imagine that there still are far fewer athletic opportunities at the school and fewer students participating. ..."

Caltech is quite small (900 undergrads vrs 4000 at MIT) which makes it difficult to attract enough participants to make some activities viable.

Caltech is quite small (900 undergrads vrs 4000 at MIT) which makes it difficult to attract enough participants to make some activities viable.

Full-scale medieval siege warfare, for instance.

I read something by an admissions director from Harvard or a comparable school (one with 10 applicants per slot) saying that he could fill about 10% of his class with automatic must-have admits and eliminate 30% of the applicants as unqualified students who shouldn't have applied. His job was selecting 90% of the class from 69% of the applicants -- there still more than 7 qualified applicants per slot. A lot of what he had to do was putting together a mix. He didn't seem to think that the differences within the 70% group were very striking.

185: yeah that's what I've heard (about 20 minutes ago, from Blume). Most of the kids who don't get into Harvard are just as talented by any measure as those who do get in; it's all deciding what's important around the margins.

Choosing between any two candidates, an admissions officer would be at least 50.00001% sure he/she made the right call - they're not totally identical. But calling them equivalence classes isn't far off.

180

"Whaa? Do you give grades as a part of your job? Have you never seen a pool of individuals with clusters in it by quality?"

I graded papers as a grad student and saw continuous variations in quality. If you have a small class and look at the score distribution you may see clusters but this mostly reflects the human tendency to see patterns in random data. With something like the SAT where you have a large number of scores there are no intermediate gaps in the score distribution.

So if the pool we are talking about is Emerson's qualified, but not automatically in why all the whingeing about legacy admits? Other than Knecht's mission statement in 77c.

One assumption, James, is that neither SATs nor GPAs nor any aggregation of the two would be an adequate grounds for choice. It would be possible to write a simple aggregating formula and admit everyone with a big enough number, but no school has done this, for good reason.

187

"... But calling them equivalence classes isn't far off."

It is misleading. Consider credit scores for instant. The difference between 680 and 681 may not be large but a 681 will still be slightly less likely to default. The line between prime and subprime is arbitrary.

*Is it unacceptably antisemitic of me

No, but this--

Jews--who aren't very, very, very, very intelligent

--is. Anti-Semite.

Because we don't think that legacy admits should be one of the favored groups at all, though realistically they always will be. Plus a suspicion that some of the legacy admits are from the 30% reject pile.

James, your begging the question. Whether or not that's true of the credit scores, people are saying that there's no such score for students.

It would be possible to write a simple aggregating formula and admit everyone with a big enough number, but no school has done this, for good reason.

This is, literally, how I got into college: only test scores and GPA. It was part of a program to recruit high school students (as juniors) likely to be accepted later by other colleges through the regular admissions process. The idea was that you could take college courses as a senior in high school and then if you wanted to, you could simply continue on. I eventually completed the rest of the application - essay, letters - to apply for scholarships, but I didn't have to.

OK, in all seriousness, I am very curious about this:

I believe that Larry Summers would have ended the legacy preference at Harvard within a few years

Even if we grant that "ended" is an exaggeration, what grounds do you have for thinking it? I don't know thing one about Larry Summers, so I have no idea whether this is based on some stated values of his, some policy plan he never got to enact, or what. All I know is that Harvard (and every other college I've ever encountered) never seems to think it has enough money.

Given that people generally assume legacy admits add to the college's pot of unrestricted money, and that precious few people seem appropriately ashamed of the practice, why waste a president's limited political capital to change it?

That was your wholesome period, I take it.

No; shortly before I started the program, I was sleeping with an 18-year old guy who worked under me (haha) in a boring-ass job while Mr. B. was overseas fighting the first gulf war. I never had a wholesome period.

Whether or not that's true of the credit scores, people are saying that there's no such score for students.

But James graded papers as a graduate student! And he saw continuous variations in quality! Plus he has a high IQ, so you know he's right.

But B, your Canadian Google picture is just delectably wholesome.

Canada is full of liars.

192: it's called free indirect discourse, or something.

190

"One assumption, James, is that neither SATs nor GPAs nor any aggregation of the two would be an adequate grounds for choice. It would be possible to write a simple aggregating formula and admit everyone with a big enough number, but no school has done this, for good reason."

You can add any other factor to the formula which you prefer. You may be correct that no school has done this (although I think some aren't too far off) but I don't think their reasons are necessarily good.

Two bad reasons are, this would reduce the importance (and perhaps the size) of the admissions department and it would force colleges to quantify the departures from strict merit (like legacies) which they like to indulge in but prefer not to discuss.

200 to 199.

Nothing is necessarily good, James! Knowing that will save you a lot of unnecessary puzzling in the future.

You seem to be of the faction that believes that anything quantified and algorithmized is better than anything informal, regardless of GIGO and unquantified factors.

194

"James, your begging the question. Whether or not that's true of the credit scores, people are saying that there's no such score for students."

There are plenty of such scores which would predict for example college gpa with varying degrees of success. There is of course disagreement about what colleges should be trying to accomplish with their admissions policy.

Now if you are saying that an experienced admissions officer looking at an applicant's file will out perform an objective formula I have my doubts. I think credit scores were developed because loan officers were not in fact very good at judging credit risk and I expect admissions officers are actually not very good at evaluating applicants either although they no doubt like to think otherwise.

158: "This is the only song we know, it's boring and it's slow."

it's called free indirect discourse

Sure, George Allen.

196: Given that people generally assume legacy admits add to the college's pot of unrestricted money, and that precious few people seem appropriately ashamed of the practice, why waste a president's limited political capital to change it?

Good question.

205: I had never heard those lyrics before. Awesome.

She had a dog named Trojan, too. I *believe* she was unaware of the condom company.

204: Surely one of the problems, at least at an elite school, is that once they'd run their hypothetical algorithm based on GPA and SAT scores, they'd only eliminate 25% of the applicants. Which means to decide who is admitted of the remaining 15,000 kids into a class of 1,500 already entails looking at non-easily quantifiable qualities.

And I get the sense that some of these measurements are only rough predictors of future success, in that the person with the 1500 on the SAT probably has better verbal and mathematics skills than the person with a 1300, but that doesn't entail much about overall GPA, choice of major, etc.

non-easily quantifiable qualities.

Cala, you're going to break its processing program.

208: You know, I don't remember USC's actual fight song, come to think of it. But their band (which does wear big latex-looking ponchos in the rain, to the merriment of all) plays this obnoxious twelve-bar theme (dah, dadadada dit dit dah da dahhh dadadada dit dit da da daaaaaaaaahhh) approximately 500 times per quarter at football games.

This is the only song we know...

I think that *is* their fight song. Or part of it, anyway.

203

"Nothing is necessarily good, James! Knowing that will save you a lot of unnecessary puzzling in the future"

So why do you think it is a good idea to allow random drones in the admissions office to indulge their petty prejudices when admitting applicants?

Apparently I'm wrong. This is their stupid fight song.

And no, thank god, Grandma didn't use the horn much.

And oh look, this one has the lyrics.

208. Please, if one must comment at least know the difference between "Conquest" and "Fight On". You probably think "Boola Boola" is the same as "Eli Yale" too.

The video in 214 makes me unaccountably angry.

CalTech realized it has a serious problem with its 70-30 split, but it's having a hell of a time changing things. When they have the orientation each spring for accepted pre-frosh, CalTech will fly and accomodate every accepted girl for free while guys have to pay...

I nearly went apeshit when I saw a current student bragging on his blog about the hugely white and Asian, hugely male student population. I don't think he even meant any harm by it; he's just stupid and 18 and hyped on the idea that engineers are smarter and better than everyone else. (One of MIT's major means of evening out female admissions was to become more selective, and start weighing SAT verbal scores more heavily. Fuck you, CalTech freshman.)

Given that people generally assume legacy admits add to the college's pot of unrestricted money, and that precious few people seem appropriately ashamed of the practice, why waste a president's limited political capital to change it?

It would have been a more productive use of his time than insulting women in the sciences (or the humanities and social sciences; or the School of Public Health; or basically everyone outside the economics department with his defense of Shleifer). He was a man not tempermentally suited to being a university president.

And I get the sense that some of these measurements are only rough predictors of future success, in that the person with the 1500 on the SAT probably has better verbal and mathematics skills than the person with a 1300, but that doesn't entail much about overall GPA, choice of major, etc.

IIRC the only factor that seems to achieve statistical significance as a predictor of college grades when socioeconomic status is controlled for is high school GPA.

CalTech realized it has a serious problem with its 70-30 split, but it's having a hell of a time changing things. When they have the orientation each spring for accepted pre-frosh, CalTech will fly and accomodate every accepted girl for free while guys have to pay...

I nearly went apeshit when I saw a current student bragging on his blog about the hugely white and Asian, hugely male student population. I don't think he even meant any harm by it; he's just stupid and 18 and hyped on the idea that engineers are smarter and better than everyone else. (One of MIT's major means of evening out female admissions was to become more selective, and start weighing SAT verbal scores more heavily. Fuck you, CalTech freshman.)

Given that people generally assume legacy admits add to the college's pot of unrestricted money, and that precious few people seem appropriately ashamed of the practice, why waste a president's limited political capital to change it?

It would have been a more productive use of his time than insulting women in the sciences (or the humanities and social sciences; or the School of Public Health; or basically everyone outside the economics department with his defense of Shleifer). He was a man not tempermentally suited to being a university president.

And I get the sense that some of these measurements are only rough predictors of future success, in that the person with the 1500 on the SAT probably has better verbal and mathematics skills than the person with a 1300, but that doesn't entail much about overall GPA, choice of major, etc.

IIRC the only factor that seems to achieve statistical significance as a predictor of college grades when socioeconomic status is controlled for is high school GPA.

So why do you think it is a good idea to allow random drones in the admissions office to indulge their petty prejudices when admitting applicants?

Same reason we allow you to hang out here and start up this argument every five minutes, James.

Jesus Christ. Something's wrong with my keyboard, people! I don't have tequila to blame this time.

Jesus Christ. Something's wrong with my keyboard, people! I don't have tequila to blame this time.

It's not my fault I can't remember their fight song because it was drummed out of my head by THIS IS THE ONLY SONG WE KNOW, IT'S BORING AND IT'S SLOW.

When we roughouse poor old Harvard
they will holler boola-boo!

The house system at Caltech seems to generate a fair amount of socializing, and the ratio is improving.

Still, I think that I had maybe three women in the classes I TA'd. And two were in a graduate class. Even here, the women seem to cluster in biology.

Snarkout, have some tequila.

209

" Surely one of the problems, at least at an elite school, is that once they'd run their hypothetical algorithm based on GPA and SAT scores, they'd only eliminate 25% of the applicants. Which means to decide who is admitted of the remaining 15,000 kids into a class of 1,500 already entails looking at non-easily quantifiable qualities."

How do you figure 25%? Just set a higher cutoff. It is true that there is a problem with the current SAT in general and the mathematics SAT in particular being too easy so you have saturation at the top of the scale and decreased ability to discriminate among the top applicants. But this doesn't apply to 75% of the applicant pool even at a top school and would probably be fixed if the top schools asked ETS to provide a harder test for them.

213: What you say is not a response to what I said.

Also, your use of the Socratic method is no good. Socrates was smart.

Snarkout, have some tequila.

Your ideas are intriguing to me and I wish to subscribe to your newsletter.

226: I figure 25% based on talking to admissions people. Somewhere like Harvard gets 20,000 applications for 1500 spots. About 5,000 were longshots at best. Of the remaining 15,000, all of them are comparable with a typical incoming class.

I think credit scores were developed because loan officers were not in fact very good at judging credit risk and I expect admissions officers are actually not very good at evaluating applicants either although they no doubt like to think otherwise.

Admission officers at Harvard are very good at evaluating whether applicants will be admitted to Harvard.

Cala 209: What I believe that Shearer is saying is that the algorithm would be perfect, so that a 29.33725 score would abviously be better than a 29.33724 score. (This is old stuff, actually, back 40 years ago Bronx HS of science calculate GPA to 5 significant figures).

James: There is of course disagreement about what colleges should be trying to accomplish with their admissions policy.

This of course is what we've been talking about all along.

they will holler boola-boo!

At the risk of going all Jonah Goldberg on you, I remember a "Gilligan's Island" episode in which a film crew deposited an actor to be a Tarzan type. Said actor ran around screaming "Boola boola". Of course, upon hearing this Thurston Howell III, says "Egads, a Yale man!" Even at ten, I thought it was a long setup for a dumb joke.

(Songs of Yale featured prominently in the Leech upbringing)

But at such fine levels of gradation, there is no meaningful difference in student quality. A point 0.00001 is what, a spelling mistake in one English essay?

219

"... One of MIT's major means of evening out female admissions was to become more selective, and start weighing SAT verbal scores more heavily. ..."

This doesn't make you more selective it just changes the type of student you are selecting for. Note this would also reduce the number of Asians which I suspect MIT was not adverse to.

Admission officers at Harvard are very good at evaluating whether applicants will be admitted to Harvard.

An admissions officer with executive authority doesn't evaluate whether or not someone will get in, s/he makes it the case that someone will or won't get in. One admissions officer might actually be not very good at evaluating whether or not someone whose application is being decided on by another officer will get in, short of knowing that other officer fairly well.

Especially given that the outcome Shearer appears to want to control for -- academic success at Harvard -- does not have nearly so fine a gradation.

If, instead, Shearer thinks he is controlling for something else (life success? earning potential? sheer awesomeness?) then he's being every sillier than I thought.

James, if I'm looking for redheaded left-handers, I'm being more selective than if I'm looking for redheads. Looking for students with stellar math SATs and good verbals is more selective than looking for stellar math SATs alone.

236 to Cala.

This doesn't make you more selective it just changes the type of student you are selecting for

Assuming the only criteria for selectivity is number of admissions per number of applicants, you're exactly right. Inane, but exactly right.

233

"But at such fine levels of gradation, there is no meaningful difference in student quality. A point 0.00001 is what, a spelling mistake in one English essay?"

I believe Emerson was being sarcastic. A legacy bonus of .00001 would in fact make little difference. However I believe typical legacy bonuses are in fact a bit larger than this.

Assuming the only criteria for selectivity is number of admissions per number of applicants, you're exactly right. Inane, but exactly right.

I hadn't even considered that this might be James' objection.

240: you have to learn to think like a robot, snark.

229

"I figure 25% based on talking to admissions people. Somewhere like Harvard gets 20,000 applications for 1500 spots. About 5,000 were longshots at best. Of the remaining 15,000, all of them are comparable with a typical incoming class."

If this means the worst student Harvard admits would typically rank about 15000 (out of an assumed 20000 applicants) on an objective scale perhaps this is correct. However it just indicates how far (at least in a few cases) Harvard is from a pure merit system. I believe the 15000th student (on an objective scale) would in fact be readily distinguishable from the top 1500.

220 allow to hang out?
how strange to say such things about another commenter
i think James B. Shearer is right in this argument
it's fair if things, anything, admissions including, are regulated by something quantifiable and systematic, not by random mercy and caprices of random people having position and power to decide
if all scores are exactly the same then one may want to evaluate other parameters

I believe typical legacy bonuses are in fact a bit larger than this

I believe the 15000th student (on an objective scale) would in fact be readily distinguishable from the top 1500

You got to believe!

Unless somebody has a cite -- peer reviewed, or else! -- James is gonna go on believing what he believes, people. Blah blah "I talked to Admissions officers" blah blah "I read an article on this" blah blah "I've been involved in the Admissions process". Data or faith: that's the Shearer way.

237

"James, if I'm looking for redheaded left-handers, I'm being more selective than if I'm looking for redheads. Looking for students with stellar math SATs and good verbals is more selective than looking for stellar math SATs alone."

But this is not in fact what you are doing unless you are reducing the number admitted. You are comparing stellar math SATs alone to excellent math SATs and good verbals.

He's perfectly right, though. Take my proposed objective scale: the number of hairs on one's head, and, in the event of a collision, the average number of bites one takes during dinner on sunday. You can totally distinguish #1500 from #15000 using that scale.

I saw a whole long argument in this thread with Shearer about the meaning of life or something and I didn't read a word of it

feels pretty damn good

You are being choosier, or let us say, more demanding, regarding which members of the stellar math SAT group you select. Now, admittedly, the admissions office was always using some way of choosing among them, but the standards for admission are now stricter. I don't think it's nuts to call this being "more selective."

it's fair if things, anything, admissions including, are regulated by something quantifiable and systematic

Sure. Only problem is that "academic merit" is not quantifiable or systematic.

how strange to say such things about another commenter

Did you mean how rude? Because my goal in life is to be as rude to James B. Shearer as possible.

Felix, I hate you. HATE.

James, you're an idiot.

The pool of potential admits is greater if you're looking for quality P alone rather than qualities P and Q. Since, in the latter case, you've right off the bat eliminated a chunk of the population, you're being more selective. Whether the total population of people with P and Q is greater than the number of slots you have (in which case, you can still take the same number) has nothing to do with it.

I mean, you are really fantastically stupid.

You are comparing stellar math SATs alone to excellent math SATs and good verbals

Nnnnooo.

Given 15000 applicants, the assumption would be that you have more than enough with stellar math SATs. You therefore have to find some other metric to select people on. Previously, I imagine MIT would do something ad hoc; in an effort to recruit more women they increased the minimum verbal SAT scores without changing the math score required. In this way, judging via the metric of test scores, they have become more selective.

It's like magic!

A purely GPA and ETS system would give high school teachers too much power. They must be replaced by Scantrons.

[ ] Agree
[ ] Disagree
[ ] None of the Above
[ ] All of the Above, Excepting the Box Immediately Above, and the Boxes that Contradict Each Other

James, people are saying that there is no objective scale that would select the top 1500. Once you've thrown out the weak candidates, you're dealing with people with A / A- GPAs and high SATs. And I believe that schools have found that just aggregating SATs and GPAs does not work.

Also (204), Harvard grades is not what Harvard selects for. Some of the students that give Harvard its reputation didn't have good grades at Harvard. Harvard's product is graduates, not GPAs.

Pwned by the rest of the let-us-now-slowly-recite-extremely-obvious-things posse.

Good work, guys!

Oh my god, I think I just fell in love with Ben w-lfs-n.

Someone, help.

But this is not in fact what you are doing unless you are reducing the number admitted. You are comparing stellar math SATs alone to excellent math SATs and good verbals.

My understanding is that the cutoff point for acceptance was not adjusted downward; it's that by raising the bar for verbal, the pool of admitable students became more gender balanced as you were eliminating a bunch of people with all the quantitative evaluation skills in the world but a certain lack of humanitas. Our experience shows that this group is disproportionately male, nu? I'm looking for the interview I read with the MIT admissions director and having trouble finding it, so anyone who can turn up a link should post it. (It may have been in a hardcopy magazine, sadly.)

(IIRC MIT already did gender weighing in SAT score evaluation, as their actual results have showed that female students in the hard sciences outperform male students with equivalent math SAT scores; this data predates the re-weighting of the math SATs in the late '90s and may not still be the case.)

I mean, you are really fantastically stupid.

So.. I'm guessing legacy admit, or maybe athletic scholarship?

Fuck, pwned by Howlin' Arctic Jetpack.

i would feel terrible after saying this rude things about another person
James seems cool to endure this much rudeness though
why James can't have his opinion however it is controversial, why everybody should agree here?
i would enjoy more reading different povs

(Did you guys know that there are people who actually grade the SATs? CLEARLY THIS IS SUBJECTIVE AND INVOLVES PEOPLE'S PETTY PREJUDICES AND MUST BE STOPPED.)

why James can't have his opinion however it is controversial, why everybody should agree here?

We're not breaking into Shearer's house to take him to the reeducation camps, Read. When in doubt, this explains everything.

260: don't worry, read, I don't think there's any danger of everybody here agreeing all of a sudden. I also don't think there's any danger of Shearer quitting trolling first grade math. So you should get what you want.

James's opinions are not "controversial." They are ignorant. He can believe whatever he wants, as long as he is willing to put up with people telling him he's an idiot.

I don't feel terrible about saying rude things to people who are ignorant, arrogant, and elitist. Nope. Sorry.

This explains nothing, but is cracking me up.

I don't see how that talk could really be a giant warmup to that pun.

ignorant is not absolutely bad, one can always learn, but one's innate rudeness tactlessness agressiveness can't be erased it seems

266: actually writing the paper, including a complete proof, putting together a powerpoint, arranging to give the talk, handing out proofs of the paper, publishing it in a peer-reviewed journal? I can see it.

267: being rude to people on the internet is actually a friendly and sociable thing to do, as it turns out.

Believe me, I was as surprised as you are when I found out.

My understanding is that the cutoff point for acceptance was not adjusted downward; it's that by raising the bar for verbal, the pool of admitable students became more gender balanced as you were eliminating a bunch of people with all the quantitative evaluation skills in the world but a certain lack of humanitas.

This is what I've heard, too. They didn't 'dumb down' MIT to get more women. They made it harder.

The whole process maybe. But not the talk itself. The pun is, after all, telegraphed in the first sentence.

read's calling you out, B.

271: so he screwed up the delivery.

It's also somewhat ironic that the "work ethic" of Jews--who aren't very, very, very, very intelligent, but will grind through everything to get high marks--was originally one of the motivations for seeking tests that would measure raw intelligence, so that the Jew could be excluded.

Huh?

One can learn if one is willing. Repeated arrogant assertions of one's ignorance without showing any sign of intellectual curiosity demonstrate unwillingness to learn. Which is reprehensible and should be roundly condemned.

Look, I'm sorry that you're offended by my rude, tactless, aggressive nature. But surely it is just as rude to presume to chastise me as it is for me to chastise James. And in any case, I wasn't being rude to you. James is a big boy. He can probably defend himself if he wants to.

This is what I've heard, too. They didn't 'dumb down' MIT to get more women. They made it harder.

LOLWUT

how is this even possible. With the same number of places, and the same number of applicants, you can't make it harder to get in.

What they did was make it harder for people who kant rite good to get in, and easier for people who write really really well to get in.

GB: Admissions officers were concerned that there would be too many Jews if admissions were based on grades, essays, etc., because your average Jew, though not innately intelligent, was willing to put in a lot of time grinding and mastering the material and whatnot. If you had a test that measured intelligence, and not knowledge of particular facts, though, you could foil the Jewish threat, because they wouldn't be able to do well on them.

No lie.

272: Indeed. See 275. If read really wants to rumble, she can throw down.

275: James, however, isn't being rude.

how is this even possible. With the same number of places, and the same number of applicants, you can't make it harder to get in.

You're an idiot too. Banned.

273: then, since I was talking about that talk, we're agreed.

281: You can make it harder to become one of the people qualified to get in. But you can't make it harder to actually get in. Unless you ask for more applicants, or reduce the number of spaces.

unless by "get in" you mean "enter a lottery by which people are randomly chosen to get in". which is really what happens anyway.

i would feel terrible after saying this rude things about another person

We're glad you didn't say it, then.

This is what I've heard, too. They didn't 'dumb down' MIT to get more women. They made it harder.

It might be qualitatively different, though. Higher verbal scores ⇒ more well-rounded student body that isn't locked in the basement of the Infinite Corridor planning mystery hunts, polishing their Brass Rats, and ruining fingerd for the rest of us ⇒ nobody ends up building SkyNet and the world is saved hurray!

Ben, if you keep doing this I'm going to offer to have your babies. And none of us want that.

279: No, he's not. He's playing the asshole trick of saying asshole things in a polite way, thereby pretending to occupy the moral high ground. And I don't like that game, and I'm direct enough to just fucking say so, flat out.

Don't like it? Tough titties.

If you had a test that measured intelligence, and not knowledge of particular facts,

My favorite SAT question: Yacht is to regatta as ___ is to____

Nope, no bias there.

281: You can make it harder to become one of the people qualified to get in. But you can't make it harder to actually get in.

You are continuing to be an idiot. Scram.

284: Also, if you actually do start suddenly admitting women, more women might actually bother to apply. Voila! More applicants for the same number of positions.

Duh.

I'll just quote ben:
"The pool of potential admits is greater if you're looking for quality P alone rather than qualities P and Q. Since, in the latter case, you've right off the bat eliminated a chunk of the population, you're being more selective. Whether the total population of people with P and Q is greater than the number of slots you have (in which case, you can still take the same number) has nothing to do with it."

281: You can make it harder to become one of the people qualified to get in. But you can't make it harder to actually get in.

Similarly, you can lead a horse to water, but you can't actually get a horse to go to the place the water is.

284: I completely don't buy your conclusion. The previous things are (a) awesome and (b) still completely prevalent. There's just more girls doing them now.

Oops, I've slurred MIT -- rtm was a Cornell student when he wrote the Morris worm.

The previous things are (a) awesome and (b) still completely prevalent. There's just more girls doing them now.

Making them, perhaps, awesomer?

284: Or you have girls down in the basement, too. I have no doubt that the culture changed, but that doesn't make it dumber (as I think you agree.)

284: I completely don't buy your conclusion. The previous things are (a) awesome and (b) still completely prevalent. There's just more girls doing them now.

I was preempting the next objection by pointing out that people in favor of a less gender-balanced campus are objectively pro-Terminator and would wish global annihilation upon us all.

291: yeah! Punk! MIT students are honest, simple folk, given to clambering around steam tunnels and playing xtrek and making novel phenethylamines. I won't have you slandering them.

292: I think it would be fair to say that, yes.

Also I don't think a corridor can actually have a basement.

Also I don't think a corridor can actually have a basement.

Why not?

Any corridor, if sufficiently wide, is indistinguishable from a room.

And you could have another chamber beneath it.

Why not call that a basement?

"The pool of potential admits is greater if you're looking for quality P alone rather than qualities P and Q. Since, in the latter case, you've right off the bat eliminated a chunk of the population, you're being more selective. Whether the total population of people with P and Q is greater than the number of slots you have (in which case, you can still take the same number) has nothing to do with it."

But you could have also limited the pool to people with qualities P and R. Choosing to instead limit it to people with P and Q makes it easier for P-people who also specialize in Q to get in, but makes it harder for P-people who also specialize in R.

This is obviously an issue of having different definitions for the same word, that word probably being "harder", so the discussion should probably turn instead to the insanely pointless and trivial "debate" that is now going on.

252

"Given 15000 applicants, the assumption would be that you have more than enough with stellar math SATs. You therefore have to find some other metric to select people on. Previously, I imagine MIT would do something ad hoc; in an effort to recruit more women they increased the minimum verbal SAT scores without changing the math score required. In this way, judging via the metric of test scores, they have become more selective"

I don't think this is how SAT scores are typically used. There may be a cutoff but scoring above the cutoff is an additional plus factor. I expect MIT went from some formula like 2*Math + Verbal to Math + Verbal which would increase average verbal scores (and women) among admitted applicants but at the cost of a reduction in average math. If in fact MIT managed to raise average verbal ability without reducing average math ability among admitted applicants than I would agree it is reasonable to say they became more selective. However I doubt that's what happened.

I don't think this is how SAT scores are typically used.

Do you have a link to a peer-reviewed article on the use of SAT scores?

But you could have also limited the pool to people with qualities P and R. Choosing to instead limit it to people with P and Q makes it easier for P-people who also specialize in Q to get in, but makes it harder for P-people who also specialize in R.

Yeah, so? No one's denying that there's more than one way to become more selective.

I mean, no one with a lick of sense is denying that.

xtrek came from either Cal or CMU, but MIT spinoff BBN gave us Adventure and thus the grue.

If in fact MIT managed to raise average verbal ability without reducing average math ability among admitted applicants than I would agree it is reasonable to say they became more selective. However I doubt that's what happened.

James, your doubts aside, that's my assertion, backed up by Cala's recollection. Why is it so weird to think that one of the most selective schools in the country would have a surfeit of smart applicants to choose from? Doesn't Harvard brag that it could fill its entire class with students who aced the SATs if it wanted to?

I expect MIT went from some formula like 2*Math + Verbal to Math + Verbal

That's what you expect, eh?

I don't think this is how SAT scores are typically used.

I don't believe this is how SAT scores are typically used.

Consistency is important!

this is who I think of when Cal Tech/ MIT girls are discussed
http://i209.photobucket.com/albums/bb208/EdwardCopeland/genius/jordan.jpg

I expect the comely lasses of unfogged to start making use of g3nitalia@unfogged.com (now accepting secondary sexual characteristics!) and that peter snees will be along shortly to mock me.

298: but the basement would be a feature of the building, not the corridor. Corridors only exist in terms of the rooms and buildings they serve! I! I!

Oh, fine. It totally doesn't have one, though.

308 to...?

I expect James B. Shearer to go away now.

308 to 305, A-b.

309: So you have a corridor-like building.

(Ok, wise guy, but noway does it sing.)

Oh, fine. It totally doesn't have one, though.

Or so the mullahs would have you believe.

I also expect a pony. And I expect dinner to magically appear any minute now.

The problem with expecting things in the past is that you can be disappointed immediately, or even earlier than immediately.

Okay Ben, you have goaded me into imagining an enclosed passageway, open on both ends, with another room beneath it. This entirely useless structure now occupies imaginative space in my consciousness that could be more profitably occupied by Kazakh swear words or the etymology of the word "Cola".

317:

useless? That would be perfect for sleeping on the roof with the fresh air and bird noises, without being exposed to things falling on your head!

304

"James, your doubts aside, that's my assertion, backed up by Cala's recollection. Why is it so weird to think that one of the most selective schools in the country would have a surfeit of smart applicants to choose from? Doesn't Harvard brag that it could fill its entire class with students who aced the SATs if it wanted to."

I was a grad student at MIT. Even among MIT students some students are better at math than others. If you select for math ability alone you will get more math talent than if you select for math and verbal ability.

the etymology of the word "Cola"

It's from the kola nut, dude. You can get high off it.

How could it be an enclosed passageway if it's open on both ends? (if I say I have locked the man up fast in the room - there is only one door left open - then I simply haven't locked him in at all; his being locked in is a sham.) So we need a fully closed passageway.

But then how is it a passageway?

(A whole cloud of philosophy condensed into a drop of grammar.)

Hey thread, what's up? I was just sitting here listening to some music, reading a book, and I wondered how things were with you these days.

I don't agree with B. that Shearer is a classic troll. He has an annoying way of arguing and I usually disagree, but he says some good things too. He doesn't have the troll's usual habit of arguing miscellaneous rightwing lines over and over again, and he doesn't seem to be trying to be deliberately annoying.

Oh man, thread, you're still full of a quasi-mathematical argument with Shearer? That sucks. I'm so sorry.

Look, thread, I'm sorry, I hope it gets better but I just can't read you right now. . .

I just can't, OK. . .

Thread, I'm sorry, I have to go now. Call me if things change, OK? Call me. Good luck, thread.

So James, what you're saying is that the math SAT is a perfect predictor of the math skills that will be required at MIT? That, for instance, you skill in group theory will be predicted by your ability to solve quadratic equations by hand? That, given a population who all got 800 on the math SAT and 5 (or whatever crazy scores they have now) on the AP calculus test, they would all be equivalently good at network theory?

Or are you saying that there is a completely comprehensive math skills test, unknown to anybody here, that MIT now cannot use because they have to take verbal skills into account?

I just want to be clear, here.

Don't let the door hit your ass on the way out, felix.

Hey, thread. It's such a relief finally to be alone. You know that felix guy was no good for you, right? He never really got you.

I have an idea. Put doors at the two ends of the passageway!

Are doors open things or closed things?

Both!

Suck on that, Aristotle!

Math is hard

321: okay, a partially enclosed passageway, open at two ends, with a subterranean room of the same dimension beneath it.

Philosophy: solved!

Actually, John, doors are only closed things, because an open door isn't a door.

It's a jar.

Walked into that one

Actually, John, doors are only closed things, because an open door isn't a door.

It's a jar.

A whole cloud of philosophy dried, ground into powder, cut into lines, and snorted by clowns.

On the MIT thing... Does anyone have a better source than me, telling the story here, for that? I think it's true, but my source is an MIT faculty member I ran into at a party once, and that what he said about their admissions process in the late 80s made sense in light of differences I'd observed between MIT classes before and after that point.

Myself, I believe it -- the guy I was talking to seemed credible. But I hate seeing it treated as solid fact unless there's another source, and if there is another source, I'm interested in it.

334: I'll ask around.

(A whole cloud of philosophy condensed into a drop of grammar.)

This, in conjunction with 298's Any corridor, if sufficiently wide, is indistinguishable from a room

is making me stare at you. Ben.

Indeed, though, rowhouses sometimes feature "rooms" that are essentially wide corridors, or hallways, if you will. Terrible idea, really.

336: a railroad apartment either contains no rooms or no corridors, or all rooms and all corridors, depending on how you define things.

I was a grad student at MIT

Oh dear. Well, hopefully they've raised their admissions standards since then.

323: You're on crack, John.

336: I was actually thinking about the line about technology and magic when writing the line in 298, but now that you point it out I can see the connection.

337: I've never heard of a railroad apartment, but I believe I've been in them, and have lived in one that might as well have been. Terrible idea, as I said.

If you select for math ability alone you will get more math talent than if you select for math and verbal ability.

But Sifu's dead right that this is ridiculous. I pulled a perfect score on the Math II Achievement, (or maybe it was a perfect score on the Physics Achievement, and a 780 on the Math) and a pair of 5s on the Physics and BC Calc APs. (I do apologize, I'm saying that only to make the following point.) And I'm not particularly mathy at all; proofs throw me once you get past the basic geometry kind, and when I ran into tensors in upper level physics classes, I got all tense and sad. The difference between good and stellar standardized test scores is much more about the format of the test than about the chance that you'll have the aptitude to do serious work in the field.

It's not just that you can have good math scores without good math skillz. It's that it's possible that there are populations equally matched as to math talent, but unequally matched as to verbal ability. Selecting for both simply doesn't mean compromising, it just means no longer ignoring one. The procedure could easily have been: first get all the super math folks, then discard the sucky verbal folks. If you discard mostly men, then you're left with a better balance, without compromise.

It's astonishing that someone who went to MIT—as a grad student!—would be incapable of figuring this out.

Maybe he was in the Architecture program.

the line about technology and magic

I can't think which line you're referring to, babe, but I'm tired. Let me know if you like. I've got to get off this machine shortly. Long day.

340: what I'm talking about.

341: ooh I want to learn tensors so bad. I'm a version* of the opposite of you; I did generally terrible in math (including on the math SAT), and generally had little use for it until I met linear algebra, which I loved and found remarkably easy.

* A version of the opposite because I haven't really taken much math past differential equations, so you may well be much more skilled than me generally, but opposite in terms of what I thrived on and what I hated.

344: sufficiently advanced magic is indistinguishable from an iPhone, or something. That Stanley Clarke guy said it.

346: You're thinking of Tori Clarke's dictum that any sufficiently advanced demagoguery is indistinguishable from jacket.

Shearer would be right if standardized tests were an accurate way of ranking at the top end, but they're not. I'm right at the top end of every standardized test I've ever taken, but that's the result of fairly high, but not astonishingly impressive, general academic aptitudes, and a very strong knack for being accurate under timed pressure -- I'm more, rather than less, likely to do arithmetic right if there's a stopwatch running. (Which isn't useless, but it's not what you do research with.)

Once you assume that admissions people know that, and that they treat everyone above a percentage cutoff as tied for "Knows all the material on the test" rather than accurately ranked against each other, then you can be selective within that unranked group by taking the ones who are also top performers on another test.

347: no, wait, I think Edmund Clarke said it, something about any sufficiently robust transition tree being indistinguishable from logic.

Far more interesting than the question of whether a passageway can have a basement, is: how does Princeton have "a basement"?

I shouldn't have used quotation marks in 350.

345: Linear algebra was fun, but the version I took wasn't terribly proofy.

Tensors may not actually be that bad; I got spooked by them in a physics class that didn't formally have any prereqs I hadn't taken, but was conventionally taken a couple years after I took it and assumed that people had already run into them.

It's funny talking about this stuff, because I've forgotten it all so completely: the e to the pi i equals negative one thing came up in conversation recently, and I was babbling about how neat it was that it's not just a weird coincidence, but makes real logical sense once you're looking at the complex plane, and realized halfway through the sentence that I couldn't remember how it actually worked any more.

Or maybe it was Richard Clarke who said any sufficiently alarmed GOP counterrorism strategy is indistinguishable from panic?

326

"So James, what you're saying is that the math SAT is a perfect predictor of the math skills that will be required at MIT? That, for instance, you skill in group theory will be predicted by your ability to solve quadratic equations by hand? That, given a population who all got 800 on the math SAT and 5 (or whatever crazy scores they have now) on the AP calculus test, they would all be equivalently good at network theory"

I am not saying the math SAT is a perfect predictor just that people who score 800 will on average have more math ability than people who score 780 who in turn will have more math ability on average than people who score 760. And as shown here it isn't true that everyone at MIT scored 800 on the math SAT. It appears that about 25% scored 800 which as I said above means the test is too easy. The problem is not just with the 800 scores the reliability of the test near 800 is also reduced because most of the questions are too easy for the top scorers meaning the scores depend effectively on just a few questions increasing the role of chance and reducing the ability of the test to discriminate between differences in ability among the best students. However this could be easily fixed if MIT cared.

"Or are you saying that there is a completely comprehensive math skills test, unknown to anybody here, that MIT now cannot use because they have to take verbal skills into account"

No, I am saying if you give more weight to verbal you are likely giving less weight to math. They can still use whatever math tests they prefer.

A large number of MIT applicants who got 800 or close to it probably also took the AHSME or the AIME, which should separate them a bit more (although not just by raw ability but by course availability at their high schools.)

352:

This formula can be interpreted as saying that the function eix traces out the unit circle in the complex number plane as x ranges through the real numbers. Here, x is the angle that a line connecting the origin with a point on the unit circle makes with the positive real axis, measured counter clockwise and in radians.

Thank goodness for wikipedia!

just that people who score 800 will on average have more math ability than people who score 780

As I said above, I doubt this is true; I should expect that above some cutoff point, people with 740 (or whatever) would be distinguishable from people with higher scores primarily on their generalized aptitude for standardized tests, rather than their mathematical knowledge or aptitude.

You can say "No problem, write a harder test". But that's going to find you the kids who have been exposed to more advanced mathematics, and will miss the kids with a strong aptitude who would catch up if given the exposure. It really is a genuinely difficult problem.

Also, I'm glad I didn't apply to MIT. There was nearly a 200 point difference between my math and my verbal scores the first time I took the SAT.

No, I am saying if you give more weight to verbal you are likely giving less weight to math.

Based, I feel once again compelled to point out, on no evidence other than your gut feeling that MIT must be letting its standards drop if they're letting more women in.

357: indeed, of my friends who went to MIT, one of them (who was perfectly successful) didn't even take calculus in high school.

And he's a white male. What were they thinking?

356, yeah, that much I remembered, but I couldn't come up with how complex numbers translated into polar coordinates properly. I hate that there are things I used to know that I don't any more.

361: oy, wouldn't that involve remembering all those trigonometric identities? They have charts for that shit.

359, 360: In Shearer's defense, everything he says would be sensible if there were good tests to distinguish math aptitude at the top end of the scale, that weren't confounded by exposure to different levels of math before testing. He's wrong, but in a comprehensible kind of way.

361: I hate that there are things I used to know that I don't any more.

This really smacked me in the face when my mother gave me some of my old blue books she had found. In some of them completely nonsensical questions were asked and then some impostor using my name answered them. Now I'm doubting everything from back then.

363: Well, not really, per Emerson, if "Math Aptitude" were some kind of fucking Holy Grail maybe, but it is not. Even at MIT.

I hate that there are things I used to know that I don't any more.

I'll leave some flowers on Algernon's grave for you, LB.

363: well, right. Given a perfect set of objectively differentiable characteristics, it should be possible to sort people according to those characteristics. Where he then decides that those characteristics must exist, if colleges only thought to use them, is where things get silly.

...but makes real logical sense once you're looking at the complex plane, and realized halfway through the sentence that I couldn't remember how it actually worked any more.

It shakes out of the Taylor series of e^ix.

I've been poking around looking for some sources on this, and so far haven't found anything, but I did come across this MIT newspaper article from 1990:

The report said many faculty did not perceive a decline in students' academic ability, but rather a lessening of the "intensely focused interest in engineering and science which once characterized nearly all MIT undergraduates." It said that while students were still majoring in engineering and science, they were more likely to distribute their intellectual energies more evenly between technical and non-technical subjects. One mathematics instructor quoted in the report said his students were just as bright but not as interested in the subject as students were five years ago.

Or you can solve the differential equation y''+y=0 and get Euler's formula! That was the bonus question on the exam I've been grading for the past week.

Heebie is right! Heebie is right! Heebie is right! Heebie is right! Heebie is right! Heebie is right! Heebie is right! Heebie is right!

Ahhhhhh, that feels so good. Keep it coming.

365, 367: I have a certain amount of sympathy for the idea that, at least within a particular domain, there are people who really are smarter in some objective sense; that's what it feels like when you interact with them. All of my fussing on the topic comes down to (a) disbelief that any tests we've got are anything more than a very crude way of measuring that, and crude to the point of uselessness on the ends of the scale, and (b) complaining about using such tests to establish differences between demographic groups with systematically different environments.

370: whoah. That's really neat.

Put "solving ordinary differential equations" on my list of things it now occurs to me I've completely forgotten how to do.

I actually ran into the e to the pi i thing in conversation some time ago and I realized that I probably never paid enough attention in that class to know why it worked out that way.

||
Is there a political thread around where I can bitch about one of the debate moderators being basically a part of the extended Clinton team (George S)? I won't bring it up here, since I'd hate to derail the search for objective perfection as related to MIT undergraduate selection, seeing as our future so clearly depends on its resolution.
|>

Given a perfect set of objectively differentiable characteristics, it should be possible to sort people according to those characteristics. Where he then decides that those characteristics must exist, if colleges only thought to use them, is where things get silly.

Quite so. There are universities that attempt to use such an algorithm, but they tend to have the characteristics of..
- needing to sort through vast numbers of applicants, far more than could be treated individually with a labor intensive process
- needing to sort those that exceed a threshold of minimal competence for higher education from those who fall below (as opposed to choosing a rarefied few from a large pool of highly qualified applicants)
- having unlimited, or nearly unlimited class size

In such cases, the roughness of the objective measure is acknowledged, but thought to be within a reasonable error tolerance. After all, the stakes of the decision are low, and the consequences for the institition of making a mistake in any individual case are negligible.

a lessening of the "intensely focused interest in engineering and science which once characterized nearly all MIT undergraduates."

Students were less likely to confuse Sacagawea with Socatoa.

being basically a part of the extended Clinton team (George S)?

I thought he was officially disgruntled. Does he count as a loyalist still?

374: It's not too hard. Verify that y=cos x, y=sin x, and y=e^ix are all solutions to the differential equation. 2nd order diff eqs have exactly two linearly independent solutions. Since we have three solutions, you must have a linear dependence relationship. And since sin x and cos x are linearly independent, you must be able to write e^ix as a linear combination of sin and cosine. Then you choose values of x which allow you to solve for your constants.

378: you forgot the hypotenuse, kemosabe.

373: Well, sure, but there are so many other things that might be important., such as creativity, enthusiasm for the subject, diligence, large genitalia, etc, etc.

To me this need to rank and quantify is one of the fatal hobgoblins of the conservative mind. Yes, there are thresholds of aptitude that must be met (and standardized tests help here). And yes, you do not want to miss the next prodigy, but most tests do not help with that.

381: I left the "h" off intentionally, so both names would end in "a."

379: I thought he was officially disgruntled. Does he count as a loyalist still?

Not sure. I was simply going by the tenor of some of his questions to Obama, which verged on being Russertesque in some case, like with the frigging Weatherman dude.

383: "Sohcahtoa"?

Actually I'm sorry to keep pedanticising you eb; I was just really excited to use "hypotenuse" and "kemosabe" in the same sentence.

Oh, right. I knew there was a reason to leave the h off. It just looks ugly with it in.

Anyway, I thought I remembered reading something by Louis Menand on college admissions, but when I search for Menand, both google and yahoo return results for both his name and the phrase "men and" - not a helpful addition on their part. The programmers probably had low verbal scores.

eb you goofball.

Oops. You goofball.

Actually. (I can no longer access that.)

Mainly I wanted to complain about "men and" - I've had a few other searches thrown off by overly helpful search engines. There should be a way to search only the character string you want, if you're sure that it's correct and is all you want.

I have a certain amount of sympathy for the idea that, at least within a particular domain, there are people who really are smarter in some objective sense; that's what it feels like when you interact with them

Sympathy with the feeling, if not the idea, yes. No sympathy with the need to insist that this is somehow a fact, against all the accumulated evidence that shows that tests are subject to bias and whateverit'scalled, that thing where if you draw attention to women's sex, or to black people's race, or to asian people's race, before you give them a math test, it affects their results.

And active antipathy towards the refusal to acknowledge one's own biases while simultaneously insisting that one has some kind of privileged access to objective truth.

whateverit'scalled, that thing where if you draw attention to women's sex, or to black people's race, or to asian people's race, before you give them a math test, it affects their results

Stereotype threat.

392: the "no really I mean it" flag?

I don't know thing one about Larry Summers, so I have no idea whether this is based on some stated values of his, some policy plan he never got to enact, or what. All I know is that Harvard (and every other college I've ever encountered) never seems to think it has enough money. Given that people generally assume legacy admits add to the college's pot of unrestricted money, and that precious few people seem appropriately ashamed of the practice, why waste a president's limited political capital to change it?

My conjecture was not based on any inside knowledge, but on inference from four data points:

1. Summers spoke out passionately about the fact that tertiary education was becoming more and more segregated by class at the very moment that the role of elite universities as gatekeepers to the highest levels of economic opportunity was becoming more pronounced.

2. He got rid of the "early decision" admission program because it demonstrably favored the children of the well-off. (Initially it was converted to non-binding "early action", and later abandoned entirely in favor of a single annual application deadline.)

3. He dramatically expanded financial aid for families with modest incomes, reducing the expected "parental contribution" to zero.

4. He altered the focus of Harvard's internationalization program, in which his predecessor set a goal of taking 11% of the student body from abroad. Summers emphasized attracting the best academic talents from abroad rather than the children of prime ministers and moguls.

Taken together, these point to a man who took seriously the democratization of access to higher education, against which the legacy preference stands in jarring contradiction.

394: Right, thank you! Sorry. Three days of sick kid = tired me.

395: Apparently, if you put quotation marks around individual words it treats them like phrases. I just realized this.

Also, I wish I had access to the NYT paid archive so I could read the article headlined thus:

THE CHINESE: A WAY WITH MATHEMATICS

363

"359, 360: In Shearer's defense, everything he says would be sensible if there were good tests to distinguish math aptitude at the top end of the scale, that weren't confounded by exposure to different levels of math before testing. He's wrong, but in a comprehensible kind of way."

Why would this confounding just be a problem at the top end? Of course the tests measure exposure to mathematics as well as ability. So what, higher scorers still on average have more ability. And I am not convinced differing exposure on average within the US will matter all that much for a test designed to pick up ability. Anyway I think it is legitimate and sensible for a school to be testing for existing knowledge as well as ability. Why should MIT be offering remedial math courses?

When did you take the SAT? The tests were recentered in 1995 which considerably increased the compression at the top end. See here .

More on MIT:

If someone can read this, it looks interesting. This is on topic, except doesn't mention test scores.

And I am not convinced differing exposure on average within the US will matter all that much for a test designed to pick up ability.

Well then, if James has yet to encounter the specific inconvenient data at issue here, then we must stipulate that his prejudices are correct.

Anyway I think it is legitimate and sensible for a school to be testing for existing knowledge as well as ability.

As indeed they do! They look at GPA, and AP test scores, and SAT II (formerly achievement test scores), as well as many other types of evidence of successful acquisition of knowledge.

If anyone's still paying attention:

The Massachusetts Institute of Technology, for example, admits women with slightly lower scores, though not by any formula. Michael Behnke, director of admissions, said women's average math score was about 20 points lower than men's, while verbal scores were about the same. (''But they're both well over 700,'' he added, a range placing everyone in the top percentiles.) He said women took about the same courses as men, earned virtually identical grades and had a slightly higher graduation rate.

That's from 1989.

393

"And active antipathy towards the refusal to acknowledge one's own biases while simultaneously insisting that one has some kind of privileged access to objective truth."

What are you talking about? I think everyone has biases certainly including myself. I generally try to be objective but acknowledge I am unlikely to be completely successful.

As for privileged access to objective truth this is some sort of strawman. I have opinions as do we all and naturally I believe my opinions are correct and other people's are wrong. I don't expect other people to share this bias.

396: Hm, thanks. Interesting suggestive evidence. It would be interesting (read: excellent) to see someone in a position of power make the argument that legacy should be a minimal/nonexistent factor. Always nice to see the pretzel-like contortions that folks get into defending irrational privileges.*

*Yes, I'm aware that legacy admissions is perfectly rational in some respects.

404: Larry Summers was quite ahead of his time outside of those areas in which he was a complete neanderthal.

Needs a little more balance in his life, Larry does.

Also to not be a sexist dickwad, maybe.

The best thing I've learned from the article in 398 is that there is such a thing as the Study of Mathematically Precocious Youth.

The article itself is not quite as horrible as its title suggests, although it was published in 1983.

What's so bad about 1983?

Just that it seems awfully late in the day for such blunt stereotyping in the liberal NYT.

for a test designed to pick up ability

BTW, James, not even ETS makes this claim for the SAT anymore. They have been backing away from it for decades, as more and more evidence acreted that the elusive "aptitude" being tested was a mirage. Most recently, ETS abandoned the name "Scholastic Aptitude Test", renaming their flagship product the "SAT Reasoning Test".

The boldest claim that ETS will make today is that the SAT assesses how well test-takers analyze and solve problems, and that SAT scores in combination with high school GPA are better correlated with first year college grades than high school GPA alone. They don't claim that high school preparation is irrelevant to success on the test,* nor do they even claim (as they used to) that it is impossible to improve test performance through coaching.

*From the SAT FAQ: "The SAT was designed so you can demonstrate your reasoning and problem-solving abilities, not just the amount of information you've accumulated during school." [emphasis added]

The Massachusetts Institute of Technology, for example, admits women with slightly lower scores, though not by any formula. Michael Behnke, director of admissions, said women's average math score was about 20 points lower than men's, while verbal scores were about the same. (''But they're both well over 700,'' he added, a range placing everyone in the top percentiles.) He said women took about the same courses as men, earned virtually identical grades and had a slightly higher graduation rate.

Right -- there's a study somewhere that indicates that women's GPAs in suggest that they do as well in mathematics, CS, and hard science classes at MIT as men with math SAT scores about 25 points higher. This is a separate issue.

403: I'm talking about your wankitude about the "objectivity" of anything with a number assigned to it, despite the obvious fact that, e.g., grades and test scores are not assigned by machines; your acknowledged disdain for teaching; your clear and unacknowledged belief that intelligence is inherent, easily defined, and measurable in the face of a great deal of evidence and research to the contrary; your unacknowledged sexism and classism; and your refusal to respond to people who point out obvious and glaring errors of logic or fact.

This last, in particular, pretty much casts doubt on your claim that you "try to be objective" as well as on your motivations. Stating and restating beliefs without actually engaging people who respond to those statements to be discussion.

Since you asked.

"to be" s/b/ "isn't"

nor do they even claim (as they used to) that it is impossible to improve test performance through coaching

I totally bought this. Practice tests? Kaplan? My cohort were fools: it wouldn't help.

Lucky I'm pretty good at winging it.

I'd say that an advanced Neanderthal has a balance that the rest of us can only wish for, bogged down as we are in our one-dimensional progressiveness.

B, Shearer seems more trollish to you than to anyone because, as I understand, you value objective quantitative tests at zero.

(Oh, plus ignorance. For all your faith in standardized tests, you seem to know very little about them.)

I expect MIT went from some formula like 2*Math + Verbal to Math + Verbal which would increase average verbal scores (and women) among admitted applicants but at the cost of a reduction in average math.

This just doesn't follow at all. Is it not possible for one person to have an 800 in math and a 750 in verbal? Would they not fare better under a system that weighted math and verbal equally than someone who scored an 800 in math and a 600 in verbal? And that this does not entail that the average math score drops?

The only way I can get your statement to follow to argue that true mathematical geniuses would do unusually poorly with verbal reasoning skills, such that they needed to double their mathematics score in order to compensate for it. I don't see this as reasonable given that the test caps at 800 (so mr. mathematical genius wouldn't have an advantage over mr. merely-good-at-standardized tests, and would have lost out on the 2M+V calculation anyway.)

402: And, I think the SAT scores normally come with something that someone who scores X is likely to score X +/-Y on future tests, and 20 points is well within the range.

I totally bought this. Practice tests? Kaplan? My cohort were fools: it wouldn't help.

If your friends had the misfortune to have me as their Princeton Review instructor, you would have been right: apparently the average scores of my charges declined after the costly instruction I imparted to them (the company tracks this pretty closely because the kids qualify for some kind of warranty if they don't improve their scores). I didn't get asked back, for obvious reasons.

you value objective quantitative tests at zero

I don't value quantitative tests at zero. I don't believe in "objectivity" in testing, and I'm extremely suspicious of people who claim that they, or some test (especially tests that have been shown, repeatedly, to contain bias) are objective.

Especially given the reams of research showing things like stereotype threat (thanks rfts!) and the inescapable cultural norming involved in testing.

Of course practice helps. It's a game. Learn the rules, pwn the game.

(I did not take a course, but I read one of their books.)

And, I think the SAT scores normally come with something that someone who scores X is likely to score X +/-Y on future tests, and 20 points is well within the range.

I remember seeing that too. On the other hand, maybe the concentrating on the verbal can make someone dumber. When I retook the test and improved my verbal score, my math score dropped (overall score improved, but not by very much beyond the +/- variation). And then I went on to become a history major!

Of course practice helps. It's a game. Learn the rules, pwn the game.

My barely informed take on the matter: If your kids aren't very smart, send them to Princeton Review and you can move up from, say, an open admissions public university to a mildly selective private one. If your kids are reasonably smart already, and they're shooting to stretch from a selective private college to a competitive one, send them to Kaplan, where you learn more actual content to know the right answer as opposed to probabilistic strategies for improving your guesswork from a one in five chance to a two or three in five chance.

I tried to comment earlier, on the way home from work, but it's apparently one of those days where my blackberry can't comment. Here it is, somewhat late:

Fucking Pope.

While we're on the topic: How to combat stereotype threat. Apparently just moving the demographic questions to the end of the test helps (!).

Kaplan once employed me, hence they are not to be trusted.

Wow, Witt. Thanks for that link. I didn't know about that blog.

The SAT may be objective, but it isn't an objective measure of intelligence. (Seriously. There's $1000 courses to take on how to beat it.) It's a measure of familiarity with material, ability to do well on timed exams, whether you went to the right schools...and such information may be useful to an institution. It's just not good evidence for much beyond that.

425: Well, they knew about you! ::insert banned cheerful punctuation here::

411

"... your clear and unacknowledged belief that intelligence is inherent, easily defined, and measurable in the face of a great deal of evidence and research to the contrary; ..."

I think the evidence and research to the contrary is mostly garbage.

"... your unacknowledged sexism and classism ..."

Like most people I am biased in favor of people like myself.

"... and your refusal to respond to people who point out obvious and glaring errors of logic or fact."

One person's "obvious and glaring error of logic or fact" is another person's "objection which is too dumb to waste time responding to". I respond to what I consider to be reasonable points. Especially when politely expressed.

I agree with read at 260 that people are being too rude to Shearer. Of course, I would think that.

If you wanted to really improve math selectivity at a top school, you wouldn't use the SAT at all. It's too easy to get a perfect score on that test and be pretty bad at math -- that's true of me, and apparently LB had a similar experience (though I'm sure she's better at math than me). Anyway, you'd just recruit on the crazy math freak circuit, which exists and is way beyond the SAT. Or you'd make up your own test.

The attempt to distinguish "aptitude" from "ability" is hopeless and ideologically motivated to boot.

What are y'all arguing about? I really can't tell.

Oh, I don't give a shit if you respond to me or not. But there are plenty of people on this thread who've pointed out that you've been saying dumb things, and you ignore them.

And if you're going to dismiss peer-reviewed bodies of research that don't suit your agenda as "garbage," well then I think that pretty much demonstrates that you're not arguing in good faith.

PGD: I agree with read at 260 that people are being too rude to Shearer.

JBS: I think the evidence and research to the contrary is mostly garbage.

You talk out your ass, people are gonna be rude.

There should be a way to search only the character string you want

At google, this works: menand -"men and"

"This just doesn't follow at all. Is it not possible for one person to have an 800 in math and a 750 in verbal? Would they not fare better under a system that weighted math and verbal equally than someone who scored an 800 in math and a 600 in verbal? And that this does not entail that the average math score drops?"

With equal math scores the person with the better verbal score will rank higher as long as the verbal score is given any weight at all. Moving from 2*math + verbal to math + verbal would for example reverse the ranking of (700V, 800M) and (800V, 720M).

I think the evidence and research to the contrary is mostly garbage.

Shearer, dude, this is needlessly inflammatory and way stupider than anything I think you actually believe. I'm generally with Emerson's 323, but when you go and throw out statements like this I reconsider.

First of all, why would you start with the presumption that intelligence isinherent, easily defined, and measurable? Those are characteristics of height, weight, and other simple traits. Those are not characteristics of more complicated phenomena, much less anything as complex as intelligence. I'd start by looking for evidence that intelligence is inborn and easy to spot, not the other way around.

But even granted that, how much "evidence and research" have you looked at? Have you read the stuff on lead-paint exposure and youth?

432: Reminds me of this quote from ex-hacker Fravia's Searchlores site.

Time and again, seekers realize that many fellow humans don't even know how to use the exclusion operator on google - this usually leaves us feeling hollow, sick, and ashamed that we inhabit the same planet (let alone that we belong to the same species) as such scum :-) Time to throw some knowledgeballs down the webhills!

I have never seen such a well-fed troll. This is simply epic.

431: it's admittedly kind of ridiculous to flatly deny that IQ is a problematic concept. But Shearer had a great deal of provocation before he snapped. And if nobody could talk out of their ass comment threads would be much shorter than they are.

James, if you're interested in some of the problems with IQ I suggest reading Cosma Shalizi on "IQ: A Statistical Myth":

http://cscs.umich.edu/~crshalizi/weblog/523.html

and also Dickens and Flynn, "The IQ paradox resolved":

http://www.apa.org/journals/features/rev1082346.pdf

Right, but the data we have says an average of 20 points (well within the margin of error), and we don't actually know what their formula was. (It wouldn't surprise me if the verbal score had been completely ignored.) So we're maybe talking 800 math vs. 780 math, or 750 vs. 730 or some such. The test isn't that sensitive.

And we should also consider whether more women applied once they heard of the change in atmosphere/de-geeking, because I can imagine a woman with 800M/750V deciding not to bother applying to MIT because she had so many other options and MIT didn't sound like an appealing place to spend four years.

And if nobody could talk out of their ass comment threads would be much shorter than they are.

I didn't say he wasn't allowed to talk out his ass, just that he shouldn't expect people to be polite while he was flatulating.

435: I don't want to exclude results that have "menand" and "men and", if there are any. Anyway, as I realized and noted above, simply writing ""menand"" instead of "menand" does the trick. I will continue to harbor disgust with search engine designers who made including "men and" the default.

OK, I'm headed for bed. In lieu of refighting the Mismeasure of Man, you guys may want to enjoy reading about knife fighting. (Where's gswift? He ought to enjoy that.)

it's admittedly kind of ridiculous to flatly deny that IQ is a problematic concept. But Shearer had a great deal of provocation before he snapped.

Shearer's first appearance on this blog was all about bragging on his goddamn IQ, how stupid teachers are, and so on. We've also had why women are bad at math, why trying to educate poor children is a waste of time, and a host of other crap. He snapped a long time before showing up here, and a long time before I started being rude to poor, poor him.

RTFA, people.

Fucking Cheney.

you guys may want to enjoy reading about knife fighting. (Where's gswift? He ought to enjoy that.)

Awesome. I just got online, saw this thread, and though, "jesus I'm glad I missed that...hey, knife fighting!"

442: a highly tendentious restatement of his commenting history. You seem to have made an executive decision that his views aren't simply incorrect, but so offensive that they can't be permitted. So you're apparently going to actively try and drive him off the blog now. I don't think that's a good thing.

What could Cheney have done that would call forth more cursing at this late date?

434

"Shearer, dude, this is needlessly inflammatory and way stupider than anything I think you actually believe. I'm generally with Emerson's 323, but when you go and throw out statements like this I reconsider.

First of all, why would you start with the presumption that intelligence isinherent, easily defined, and measurable? Those are characteristics of height, weight, and other simple traits. Those are not characteristics of more complicated phenomena, much less anything as complex as intelligence. I'd start by looking for evidence that intelligence is inborn and easy to spot, not the other way around.

But even granted that, how much "evidence and research" have you looked at? Have you read the stuff on lead-paint exposure and youth?"

I chose not to quibble with B's strawman characterization of my views since it wasn't too far far from my actual beliefs. I see this was a mistake. Obviously I don't believe intelligence is entirely inherent just substantially inherent under current conditions in the US. No one denies a sufficiently bad environment will reduce your intelligence possibly severely. And by measurable I don't mean measurable without error.

Oh please, PGD. He's commented on other shit, but he's consistently an ass about education, and not just because I disagree with him but because he's intellectually dishonest, what with that "I think that research is garbage" thing.

Anyway, if I could drive him of the blog, he'd be long gone. I'm sure he's not such a pussy that a little bitching from me is going to scare poor widdle him away. Especially since he seems to hold me in much the same contempt that I hold him.

But fine. I'm offending your delicate sensibilities, everyone knows he's an ass anyway, there's very little point in me babbling on about it at this point, regardless.

I'm just parachuting in, but why did my smart friends, who went to the same school and took the same classes as my not quite as smart friends, do better on standardized tests? It's all very complicated, yes, but there is such a thing as aptitude or intelligence (you all know this) and the SAT does measure some aspect of it. It gets denied because any way of measuring it that has real-world implications is bound to be unfair and discriminatory, but you should say that instead of pretending that some people aren't smarter than others and that we can't get some sense of who's who by using tests.

Is that even what y'all are arguing about? If not, ignore.

Cheney?

448: nobody's saying that. The con-Shearer argument is that when you're dealing with a large pool of very talented and high-achieving people, making objective judgments about exactly how they should be ranked is impossible.

Contra-Shearer, that is. Up down up down left right left righ b a b a select start.

It gets denied because any way of measuring it that has real-world implications is bound to be unfair and discriminatory, but you should say that instead of pretending that some people aren't smarter than others and that we can't get some sense of who's who by using tests.

Did you ever read Michael Lewis's The Blind Side?

Fucking CIA.

The argument isn't that we can't get "some sense." It's that we can't claim to be able to make incredibly fine distinctions beyond that.

Then there's a further argument about whether MIT would have necessarily had to lower their math aptitude test standards in order to increase verbal aptitude standards, which is just too stupid for words.

On the intelligence test tip, the problem is that an ideal case such as the one you present almost never happens, and while yes, obviously some people are smarter thm others it's fiendishly difficult to get much more then a noisy, occluded sense of how much smarter if all you have to go on is standardized quantifications.

Ok, I buy that. I think Shearer argued that 760 and 740 measured distinguishable differences in aptitude. I don't buy that. I buy that, on average, 800 is materially different from 740 and and 740 is materially different from 640. I'd be surprised if admissions people didn't know all this.

B-pwned, amusingly.

A friend of mine who does phil bio and phil race presented a really interesting paper last year on stereotype threat and justice.

THE MORE YOU KNOW

I don't think doing well on the SAT necessarily helps in a knife fight, but if you bring a whole stack of ETS booklets, you'll probably be able to use them defensively. Flash cards could help too.

456: right. He was arguing in essence that you should come up with an accurate numerical ranking for "smartness" which you could use as an optimal college admissions criterion.

(nb unlike some of my comments above I'm trying to non-sarcastically represent his views here)

448: I think some people do say that. But, you know, as a big geek I spent some time once working with some people who were using college basketball performance statistics to come up with a single statistical measure that predicted how good someone would be in pro basketball. It was incredibly difficult, and in the end they couldn't really come up with one that was more than roughly useful. Too many different kinds of skills in play.

Well, there are a zillion more ways to use your brain to be good at the game of life than your body to be good playing basketball. Trying to extract a single common factor that predicts mental performance is problematic for all sorts of reasons, and there's certainly no guarantee that it corresponds to something "real". The reasons it was invented were institutional/bureaucratic as much or more than scientific.

Intelligence is real, and takes many different forms. That doesn't mean IQ is.

The Cosma Shalizi piece I linked in 437 is quite good on this, although long.

459: "the next five minutes:your death::the 5th century AD::the sacking of Rome"

multiply pwned, of course.

Don't bring a long, earnest lecture to a knife fight.

If I had it to do over again, I wouldn't want classmates who are super smart -- what good are they ever going to do me? I'd want classmates who are going to be rich. And well connected. (I didn't say stupid. Normal is fine, though).

456: Yes,"materially different" for what it measures not necessarily for better to admit. Plus 740/800 is marginal, or at least much better signaled through other means.

I buy that, on average, 800 is materially different from 740 and and 740 is materially different from 640.

Sure. Do you think that those material difference are the end-all and be-all of intelligence, or that an 800 score with good grades and a very conventionally narrow upper-middle-class upbringing and all the advantages might actually not be more intelligent than a 740 or even a 640 with decent grades whose parents never went to college? Because that, too, was part of the argument.

For example, my bosses boss is extremely stupid in a conventional analytic sense. To a striking degree. But she's crafty. Disciplined. And successful.

Big problem for me, because I communicate best with people who are analytically facile but loose and undisciplined. Like me!

IQ has nothing to tell me here.

If I had it to do over again, I wouldn't want classmates who are super smart -- what good are they ever going to do me? I'd want classmates who are going to be rich. And well connected.

So fucking true. Sigh.

wouldn't want classmates who are super smart -- what good are they ever going to do me?

Or, indeed, themselves. Everyone who has ever spent any time at all in an elite grad -- or even undergrad -- program can point to at least one much-smarter-than-everyone-else-there character who nevertheless failed miserably to make anything of themselves, because it turned out that they couldn't work reliably, or ever finish anything, or have a creative idea, or whatever. And that's not even touching the ascriptive stuff wherein some people just can't get coded by others as future winners no matter how smart they are.

or ever finish anything

Uh oh.

Get to work, eb.

But I wouldn't want whiny privileged classmates. At least not too many. A tough balance.

(My brother graduated HS a year ahead of me, and we went to college on opposite sides of the Bay. I ragged him mercilessly -- but justifiably -- about the entitled never-got-a-B-ever whitebread whiners at "Junior University.")

Fucking Robber Barons.

My current problem is a lack of projects in need of finishing. Hence the uptick in commenting.

Also, sources indicate that Leland Sr. was not highly regarded by his associates on the railroad they were building.

Leland was crafty, that's how he won out. He ripped everybody else off. Craftiness is one of the many useful mental skills that is not strongly correlated with IQ.

His craftiest move was to join up with some seriously crafty associates. Who often found him exasperating.

I've probably told the story of my professor who was talking to Habermas about a grad student, and said that the student might be a genius. Habermas said, "But in philosophy, that is not such a good thing to be."

Of course, that's partly a sensitivity to philosophers going very very wrong, but still. I don't know who they were talking about, so I don't know how the story ends.

That reminds me of something one of the managers at a place I worked said regarding people overthinking things for our mostly repetitive data-entryish job: "If you can't aim for mediocrity, we probably can't use you."

Ok, I buy that. I think Shearer argued that 760 and 740 measured distinguishable differences in aptitude. I don't buy that. I buy that, on average, 800 is materially different from 740 and and 740 is materially different from 640

Materially different, and probably correlates somewhat with at least a good memory and sense of patterns, but what's meant to come of that, beyond bragging rights? Scholastic success? Scientific aptitude? Anecdote, but hey, if you were to compare my scores and those of one of my sisters, you'd conclude that one of us would be the soft humanities major and one of us would be the ballsy engineer, and you'd get it exactly backwards.

437

"James, if you're interested in some of the problems with IQ I suggest reading Cosma Shalizi on "IQ: A Statistical Myth":"

The link actually seems to be entitled "g, a statistical myth". g and IQ are not the same thing. As Shalizi says in one of his linked pages:

".. Whether IQ means anything or not, it is, unlike general intelligence, unquestionably something we can measure, so we can consider how heritable and malleable it is. ..."

And as he says your link:

"All of this, of course, is completely compatible with IQ having some ability, when plugged into a linear regression, to predict things like college grades or salaries or the odds of being arrested by age 30. (This predictive ability is vastly less than many people would lead you to believe [cf.], but I'm happy to give them that point for the sake of argument.) ..."

The whole issue of whether g is "real" or not is technical and something I don't have a strong opinion on. I don't think it makes much practical difference.

437

"and also Dickens and Flynn, "The IQ paradox resolved":"

I didn't read the whole thing but I note that they concede a high heritability:

"One could challenge existing heritability estimates. However, a committee of highly respected researchers convened by the American
Psychological Association concluded that by late adolescence, heritability is "around .75" (Neisser et al, 1996, p. 85). Future research may change this value, but we do not choose to dispute it. We suspect that when the dust settles, the value for h2 in adults will be high enough to allow Jensen to make his argument. ..."

They go on to argue that this may be a bit misleading because intelligent individuals seek out environments that allow them to further develop their intelligence and this has the effect of amplifying the inherent genetic differences in intelligence. This seems plausible.

456

"Ok, I buy that. I think Shearer argued that 760 and 740 measured distinguishable differences in aptitude. I don't buy that. I buy that, on average, 800 is materially different from 740 and and 740 is materially different from 640. I'd be surprised if admissions people didn't know all this"

Distinguishable when comparing large groups. I don't get this small differences are the same as no differences argument. If 740 is equivalent to 760 and 760 is equivalent to 780 and 780 is equivalent to 800 then if we were really dealing with an equivalence relation 740 would be equivalent to 800. But in fact while there is some chance if you repeated the test a 740 would outscore a 760 this chance is less than 50%. And if you compare a group averaging 740 to a group averaging 760 the chances the 740 group would outscore the 760 group on a retest rapidly drop to zero as the group sizes increase (assuming you are not selecting unrepresentative 740 and 760 scorers to make up your groups).

You see this in public opinion polls where a 2-point lead is called a statistical tie. In fact the person with the 2-point lead is quite likely ahead (assuming an unbiased poll with just sample size error) just not the 95% favorite conventionally required to be statistically significant.

OK, I am totally Becks-style and have read the last few hundred comments too quickly to absorb all of them, but what is being argued here? Shearer says there's a reliable way to order people by -- what, intelligence? "deservingness of ivy-league education"? some other ineffable meritoriousness? -- and... that's being argued, when it's obvious nonsense? What the hell is intelligence, anyway? I'm puzzled. Is there a substantive argument here that I'm missing? Or are people just trying to engage with utterly incoherent bullshit?

How many people annually get a 1600 (err, it's 2400 now? whatevs) on the SAT? Surely enough to populate Harvard, right? Would anyone really think that's the best way for Harvard to choose its entering class?

If 740 is equivalent to 760 and 760 is equivalent to 780 and 780 is equivalent to 800 then if we were really dealing with an equivalence relation 740 would be equivalent to 800.

JBS is Tim Williamson, or possibly Mr Magoo.

Oh, and I claim my convincing refutation given by a program-mate some time ago, which may or may not have been published already?

Yup!

479: I mean, there is some correlation between high SAT scores and later good stuff (higher college grades, wealth, etc.), which shouldn't be a huge surprise. But they are noisy correlations, to say the least.

And PGD, those schools did pull from the crazy math freak circuit. Not too many, it was hardly a guarantee of getting in, but the kids I knew from the USAMO were all going to really really really top schools. Like, the same 5-6 top names in math/science/engineering. So I presume it helped.

430

"And if you're going to dismiss peer-reviewed bodies of research that don't suit your agenda as "garbage," well then I think that pretty much demonstrates that you're not arguing in good faith."

Really? There is also of course a large peer-reviewed body of research on IQ which I find more congenial and which I suspect you would characterize as garbage. Am I wrong? What is your opinion of Arthur Jensen?

And don't people on this group routinely ridicule peer reviewed papers in for example evolutionary psychology?

There is also of course a large peer-reviewed body of research on IQ which I find more congenial and which I suspect you would characterize as garbage. Am I wrong? What is your opinion of Arthur Jensen?

In my experience in my own field (and I think it generalizes), maybe... say... 70% of peer-reviewed research is mediocre (not wrong, but not even mildly interesting), and another... maybe... 20% is utter crap. Peer review filters out things that are egregiously wrong and inept. It still lets through plenty that is, at best, controversial.

The real test is not whether someone like Jensen has publications that pass peer review, but whether his work has led to general agreement and other research that follows his own. The follow-up literature in this case is, to put it mildly, less than fully supportive.

442

"... bragging ..."

Speaking of which, who is so proud of their Phd that they have to remind everybody they have one with every post?

Me!

I walked at my graduation, too. You're goddamn right I'm proud of my PhD.

483

"How many people annually get a 1600 (err, it's 2400 now? whatevs) on the SAT? Surely enough to populate Harvard, right? Would anyone really think that's the best way for Harvard to choose its entering class?"

According to this 269 which I don't believe is enough. And the issue isn't whether that would be the best way but whether that just using the SAT would be better than what they do now. I wouldn't depend solely on the SAT myself (since among other things it is too easy) but I don't think the arbitrary subjective judgement of typical admissions office people is likely to improve things.

489

"The real test is not whether someone like Jensen has publications that pass peer review, but whether his work has led to general agreement and other research that follows his own. The follow-up literature in this case is, to put it mildly, less than fully supportive."

The general idea (which did not originate with Jensen of course) that IQ usefully quantifies something that approximates what people mean by intelligence is widely accepted. As is the evidence that it has a substantial genetic component. Jensen's views on race and IQ are less accepted but it is not as if no one agrees with him. The work B is citing is widely disputed also.

Jensen's views on race and IQ are less accepted but it is not as if no one agrees with him.

James, substitute Behe for Jensen and I think you'll see the flaw here. Have you ever read Mismeasure of Man? It seems to me that it renders Jensen's racist stuff impossible to even talk about as science.

It is true that there is a problem with the current SAT in general and the mathematics SAT in particular being too easy so you have saturation at the top of the scale and decreased ability to discriminate among the top applicants. But this doesn't apply to 75% of the applicant pool even at a top school and would probably be fixed if the top schools asked ETS to provide a harder test for them.

I'm not going to engage in any further troll-feeding in this thread, but I wanted to draw attention to an amusing irony. There actually exists an educational system that reserves all the best educational opportunities (which in turn act as gatekeepers to all the most prestigious positions in economy and society) to the highest scorers on a ridiculously difficult math test. It's called France. Shearer has cast his lot with the cheese-eating surrender monkeys!

441: Holy guacamole, Witt! I just read that knife-fighting article, and that dude is scary. These two quotes really jumped out at me:

"As a bouncer, I moved around a lot because I'd get bored," Kier says. "I remember going to the manager of Barnaby's and saying, 'I am just checking IDs here. I wanna be punching someone every night.'" [This confirms every prejudice I ever had about bouncers. --KR]

"We don't train kids. We won't train someone who's clearly maladjusted. And we teach a lot of philosophy before teaching how to kill," Kier assures. But he can't deny teaching scared, overly violent people. "If you get someone like that in class," he says, "they might do it anyway. Better that they be trained."

Uh, yeah, sure, Dude. So the victim will die quickly and painlessly instead of slowly bleeding to death, I suppose.

Shearer, just because results of a test are consistent over time doesn't mean it's a good test for what you want to test for. It means it's a test for something, but doesn't tell you anything about what. You could give me the SAT over and over until I plotzed, and I'd keep on getting top-end scores. This would not make me any likelier to be a creative mathematician, nor would it make Jane Smith, who makes calculation errors under pressure but has more mathematical aptitude than I, any less.

What, actually, is the ethical argument against individual universities setting their own entrance tests to select for the kind of student they want to attract. I know there must be one because it's a practice that's been in retreat throughout my life, but I've never seen a sensible explanation.

There's obviously a facile point that universities which are interested principally in attracting rich white kids might set questions like "What is the name of the King of Spain's daughter's favourite cat?", but this could surely be easily got round by establishing a standards authority which ensured that the criteria were only academic aptitude, creativity or whatever.

who is so proud of their Phd that they have to remind everybody they have one with every post?

Oooh, me too! I'm really proud of mine.

I'll take this opportunity to make a plug for Nick Lemann's The Big Test.

It's a work of history, not a polemic, but it illuminates a lot about the uses and misuses of standardized testing in 20th century America. The tension between the democratic promise of testing ("we'll sift through the masses and find the students of extraordinary ability, whatever their origins or previous schooling") and the real-world tendency of tests to favor the already privileged--a tension that emerged contemporaneously with the tests themselves--is explored in rich detail.

And I never did that well on standardized tests, and my SAT scores were probably low for Unfogged, and I didn't get into any Ivy League schools. (...Um...I didn't apply to any. But I applied to Brown, and was turned down.) So apparently the SAT doesn't measure awesomeness that well.

this could surely be easily got round by establishing a standards authority which ensured that the criteria were only academic aptitude, creativity or whatever.

First, assume a can-opener.

I didn't get into any Ivy League schools.

The quota for Jews from New Jersey fills up very quickly.

I didn't get into any Ivy League schools. (...Um...I didn't apply to any. But I applied to Brown, and was turned down.)

Wait, what?

"If it's Brown, it must not be Ivy."

It's true! I dunno, I didn't set myself apart from the masses, I guess.

Oh right, Brown is Ivy League. They didn't teach us that at gigantic sprawling State U. I GOT SHOT DOWN AT EVERY IVY LEAGUE SCHOOL I APPLIED TO.

What, actually, is the ethical argument against individual universities setting their own entrance tests to select for the kind of student they want to attract.

The argument is that schools that aren't well-connected and don't traditionally send students to e.g. Oxford aren't going to know how to prepare students specifically to take e.g. the Oxford entrance exam -- and in fact lots of schools won't even encourage students to apply, or make resources available to them to take it.

Though AFAICT, doing away with the exam has just made the interview even more important, which seems even more nerve-racking and difficult to prepare for.

Convergence of evil: John Yoo spouts off on affirmative action and college admissions. The style of argumentation that characterizes his legal writing--uncompromisingly extreme conclusions based on the flimsiest possible logical and empirical basis--is on display here.

His view on standardized tests:

I think that's the fairest way. It's the way that removes human bias from the system. A standardized test is something everybody takes. There are a lot of claims that these SATs and LSATs are somehow culturally biased. I don't know why a newspaper just doesn't publish some of the questions. I mean math is math. And math is a substantial portion of the SAT. A lot of the LSAT asks very basic reading comprehension questions and logic games. It's not culturally biased. You don't have to be from a certain culture to understand the questions. But these are arguments that people are making.

re: 509

Plus, colleges still use admissions tests. They are just subject specific, or sometimes college specific, rather than formal entrance exams.

509(3) was the sort of point I had in mind in asking the question. But if all universities had appropriate entrance tests (I suppose some of them would cartelise), then schools would have to get their shit together to prepare students for at least some of them. Which would mean more work for school teachers, but don't confuse me with somebody who would care about that.

I feel amazingly proud of my university now, on the grounds that it doesn't accept bribes for admission and doesn't even ask who your father is or where he went to college. Merit only.

Every now and again someone tries to buy their kid a place with a huge donation, and they get told where to shove it.

Ah, Oxbridge.

508: I GOT SHOT DOWN AT EVERY IVY LEAGUE SCHOOL I APPLIED TO.

RIGHT AGAIN HEEBIE!

I stayed away from the Yoo-tenure arguments in purpose, but in a perfect world anything as asinine as the stuff quoted in 510 would get you barred from universities, and indeed teaching in general.

And to 497 yeah, that second quotation jumped out at me too. It's like driving by something and then doing a doubletake. Superficially plausible, and then you realize that's the same kind of thinking that says armed robbers need a few shooting lessons. Uh, no.

(In related news, I was disturbed yesterday that a colleague and I both had the same reaction to hearing about a shooting. "At least it was only in the leg." What on earth is wrong with us? Talk about the normalization of violence.)

on purpose.

One final tangential comment, and then I need to focus on some other things:

There has been a controversy in recent years about ETS making accommodations for test-takers with learning disabilities. If you have a certified learning disability, you can petition ETS to get more time on the SAT, or even to take the test untimed. Having more time for the test is solidly correlated with higher scores.

The controversy stems from the fact that ETS has allegedly become more forthcoming with these waivers (post-ADA?), and that students and parents are becoming more aggressive in getting a friendly psychologist to provide a diagnosis of a learning disability.

My aunt, a retired special education teacher, knows the president of the College Board, the parent organization of ETS. She saw him reading something once, and blurted out "You're dyslexic, aren't you? You're a context reader." He smiled and confided that he did, in fact, have a learning disability.

No particular point in this comment, I just found it interesting.

"At least it was only in the leg."

By God, sir, I've lost my leg!

By God sir, so you have!

By God, sir, so you have!

Auto-pwned, despite being dead for 150 years. I blame the playing fields of Eton.

Thought you scions of privilege would be interested in this.

Tim Williamson, or possibly Mr Magoo.

As I recall, Tim Williamson's glasses are indeed of Magoo-like proportions.

Indeed.

re: 523

Not really, that I remember. Unless he's gone all high-tech and thin-lensed in the past few years.

stubborn people kept arguing, i like it
BitchPhD, i apologize too, i really don't know people here well enough to make assumptions about their characters and i agree with you sometimes
as about James B. Shearer, sure he can defend himself, i just was curious who he is, he always sounds like that detached, a bit mechanical, talks about the debate points only, revealing nothing personal, interesting
i imagine someone 50 yo, an engineer, a MIT graduate, working for ibm?, a gentlemanly autistic type, arrogant? i don't sense that, lonely enough to comment here
i would google him, but i won't, coz i stalk imaginary ogged only :)

Ajay, I think that a lot of the Oxbridge feeder schools do a pretty good job of screening for class.

Maybe Oxbridge admission is 100% meritocratic now, but isn't access to quality secondary education still class/money-based?

526: lonely enough to comment here

Sometimes I feel like I don't have a partner
Sometimes I feel like my only friend
its the blog that I post on, the comments of strangers
Lonely as I am, together we cry

See? The RHCPs are sensitive, like James Taylor. Not just a bunch of dorks.

527. Yes. Simple question/answer. This is a major problem, but where in the world is it not so?

it doesn't accept bribes for admission and doesn't even ask who your father is or where he went to college. Merit only.

Errr... What? I read maths at Emmanuel, and they certainly knew and cared that my granddad had gone there, and that my cousin had gone to Johns (but I swear they didn't hold that against me).

Also, are you seriously going to deny that STEP exams are skewed toward helping students from better preparatory schools? And since the demise of the grammar school system, there's not even much of a way for the poor or working-class to get into those high-end schools in the first place.

(besides, everyone with four middle names and a landed estate in waiting knows that you just foil the meritocracy by applying for Anglo-Saxon, Norse, & Celtic)

527: well, technically, the "Oxbridge feeder schools" do a moderately good job of screening for money, which is definitely not the same as class.

But there are no literal feeder schools which have a better chance, ceteris paribus, of getting a pupil into Oxbridge than others; it's no longer the case (as it was before the War) that, for example, Jesus College, Oxford has scholarships reserved for pupils from certain schools in Wales.

There are schools which tend to produce students of better academic quality, yes. And many of those are fee-paying schools. But there's no reason - certainly no official reason! - why a clever, articulate, four-A Etonian should have more of a chance of getting in than a clever, articulate, four-A pupil from a state comprehensive school in Bradford; it's just that Eton tends to produce more four-A students.

Also, are you seriously going to deny that STEP exams are skewed toward helping students from better preparatory schools?

The only course that still uses STEP is mathematics at Cambridge. Are you suggesting that the papers somehow focus on a special sort of upper-middle-class mathematics?
I think things may have changed a bit since you went up to Emmanuel.

In 2002?

Of course, the problem with all defences of privilege by the appeal to "excellence" is that discrimination is itself anti-excellent. If you want to pick the top 10% for characteristic X, pick 'em. If you pick some percentage more men, by definition you must be leaving some women in the top 10% out. Look around you; it's not South Africa, Nazi Germany and the Confederacy who won.

But this doesn't just apply to what you might call crude discrimination (no girls, blacks, gays, Jews etc). If your selection process has an intrinsic bias - i.e. it overvalues Steven Den Beste-libertarian mathletes (maths SAT scores) or Awxfud bullshitters (tutorial interviews) compared to some platonic unified measurement* - it's doing the same thing.

The French education system is notorious for exams that require you to pass all the subjects to qualify, so the Kurt Vonnegut thing about having to go back and retake the sports unit is almost a reality.

*This is of course unattainable, but necessary for purposes of argument. Anyone mind if I borrow the spherical cow for a moment?

The French education system is notorious for exams that require you to pass all the subjects to qualify

Yeah, for this reason the most frighteningly impressive people I know are products of the very elite end of the French system. They tend to defy the efforts of Anglo types to classify people as quant or qual or hard science vs humanities, etc.

534, cont.: Though I have to admit that if the STEP exams are only given for math these days, then things have changed surprisingly quickly. When I applied, they were still requested for engineering and most of the hard sciences, at least for the most competitive colleges. And the interviews in the humanities sounded basically like the equivalent of a STEP exam done on site anyway.

The interviews must be doing one hell of a lot of work these days, which is good and bad. I do feel it's probably a better admission system on the whole than that used by US undergrad, but it's certainly way more work-intensive.

537 (2) My understanding, which is probably way out of date, is that the US undergrad admission system is called freshman year. Those who pass go on to do a three year degree.

And it only costs $27,000, and you get to apply to one school per year.

clever, articulate, four-A pupil from a state comprehensive school in Bradford

You think I *wanted* to go to Oxbridge?

538: Well.. some schools are apparently a little like that (I believe Emerson's mentioned Reed in the past, and I've heard CalTech does get a decent number of burnouts). But the Ivies and other elite non-technical schools in the US have pretty staggering retention rates. So it really is kind of a lottery for most to get in, then you're there no matter what. There's no large freshman class + winnowing.

"If it's Brown, it must not be Ivy."

But we beat Harvard and we beat Yale! We tied a knot in the Tigers' tail!

There are a lot of claims that these SATs and LSATs are somehow culturally biased. I don't know why a newspaper just doesn't publish some of the questions.

That's the kind of "it sounds right to me!" professoration that gets you in trouble with the Hague (and tenure at Boalt Hall). Jesus Christ.

I have read that the French system emphasizes math at all levels more than any other. I'm really sorry that the American system allowed me to slack on math. I never would have been very good at it, but a solidly mediocre math background would be better than what I've got.

Different topic: I think that the American system of giving second and third chances to late starters is a good thing from an egalitarian point of view, but tends to produce narrower scholars. If a bright 19 year old with a weak background in everything gets serious and starts working, they're unlikely to have the time to get themself up to speed in anything not necessary for their specialty. Even difficult skills relevant to specialties tend to get piched, such as foreign languages for English majors.

"pinched"

I'm really sorry that the American system allowed me to slack on math. I never would have been very good at it, but a solidly mediocre math background would be better than what I've got.

If you aspire to solid mediocrity, it's never too late to start. As my brain ossifies, I keep meaning to try working my way through some of the MIT Open Courseware math surveys (particularly the undergraduate combinatorics one, which I think I could still handle).

537 - yep, they binned STEP for everything except maths in 2003. But it's difficult to see how even an engineering STEP could be biased to the upper middle class. A capacitor is a capacitor whether you drop the t or not.

543.2: what?

You just have to be a late finisher, too. Don't ask me, though, ask soup.

526: Eh, no worries. I was irritable by the time you stepped in.

I think that the American system of giving second and third chances to late starters is a good thing from an egalitarian point of view, but tends to produce narrower scholars.

Disagreed. IME, most "late starters" are polymaths. I kinda think that having broad general interests tends to lead people to be less academically "successful" during adolescence, when college-prep courses expect you to start "focusing."

B's got my back.

546: Wow, I didn't know that. Weird.

But the bias isn't so much a fancy-pants "upper-class math" as it is a very distinct style of question on the Maths III STEP exam. If a school didn't offer further maths A-level, you're just plain screwed no matter what. But I also remember the questions seeming fairly odd and idiosyncratic, not very similar to the questions I'd seen on stuff like the Putnam and Math Olympiad series, which are designed to be doable using very common elementary maths and to isolate something like higher-level maths ability.

Basically, it struck me as a test that one could really prep for, and that will naturally bias toward those whose schools can afford that out-of-curriculum prep. I'm nearly positive the admission percentages for comprehensive school students versus public and private school students wtih similar similar A-levels will support me on this (I remember them being pretty damning when I was there, and part of the reason so many colleges were setting up outreach programs to bring in more comprehensive students).

480-481:

The link actually seems to be entitled "g, a statistical myth". g and IQ are not the same thing.

ummm, one could say that they are. Or at least that g is so essential to the measurement of IQ that the two stand or fall together. Are you sure you're familiar with psychometrics?

As for Dickens and Flynn, their point is much more fundamental than you perhaps understand. They are saying that IQ can be highly heritable *and* highly environment-dependent at the same time.

The general idea (which did not originate with Jensen of course) that IQ usefully quantifies something that approximates what people mean by intelligence is widely accepted. As is the evidence that it has a substantial genetic component.

Since the two links I posted call into question both of these contentions -- that IQ quantifies intelligence and that it is genetic -- I'm not sure you understood them when you read them. Heritability is not the same as genetic, coming up with a pen-and-paper test score that has a small correlation with later earnings is not the same as quantifying "intelligence".

To the extent that IQ scores approximate popular ideas of intelligence, this is because popular notions have themselves been shaped by IQ. You get things like Mensa, which so far as I can tell is an organization of high-IQ stupid people. IQ scores have something to do with a facility for quick manipulation of abstractions, which in turn has something to do with brainpower, but that's about it. Like vertical leap has something to do with athletic ability but doesn't really get you very far in figuring out how well people will play sports.

526: read, I love you.

549: B, tell me: is a 19 year old late starter with a weak background going to learn several foreign languages? Or even one language well, if not necessary for his major? Is a 19 year old late majoring in the humanities going to learn much math or science? Is such a science major going to learn much humanities?

The elite European systems have the advantage that everyone who comes out of them has a pretty solid general foundation in what are called "distribution requirements" (often already by age 16). The American late starters I'm thinking of breeze their way through school not studying much of anything, even though they're bright. Even American students who work hard in HS often have serious deficiencies.

I think Emerson is absolutely correct. It goes deeper than "late starters". In a sense most Americans are "late starters" because we have mostly lousy high schools but try to make up for it with great post-secondary institutions. But even most of our colleges don't have a required curriculum, so people often ramble around for a couple of years before locking in on a major. You get people who are well specialized in one field for their career, but have major holes in their education otherwise.

It's the flip side of the democratic openness of the American system, as well as the vocational emphasis.

Heritability is not the same as genetic

Could you say a little more about this? I know you're right, but I have a hard time articulating the difference. It's related to genetics, but what is it exactly? There's an example something like having 10 fingers is 0% heritable, because (by hypothesis), anyone having nine fingers lost one in an accident. So differences in fingers among a population aren't due to a genetic tendency to have 9 fingers.

coming up with a pen-and-paper test score that has a small correlation with later earnings is not the same as quantifying "intelligence".

How does IQ correlate with academic success? I always thought there was a pretty decent correlation there.

You get things like Mensa, which so far as I can tell is an organization of high-IQ stupid people.

Hah! I'm not sure I've ever interacted with a Mensa member, but I suspect that Libertarians are over-represented in their ranks.

My late mother in law was a member of Mensa. If being high-IQ stupid was an Olympic event, she'd have been captain of the US team. But she was actually sort of socialist, and had no time for libertarians.

Cala: it's obvious that heritability and genetic basis are different in a very simple sense -- lack of a foreskin is extremely heritable among Jewish families, yet it is not genetically determined.

The IQ argument is more complex, and comes from things like studies where identical twins are separated into different family environments but end up with similar IQs. Genes are constant, environment is different, IQ stays similar, hence it must be genetic.

The Dickens/Flynn point is how easily this conclusion falls apart once you introduce a gene/environment interaction, and how dependent it is on the exact nature of the gene/environment interaction. If relatively small genetic differences in IQ slot you into very different learning environments, for example in school, then IQ can look highly *genetically* inherited but still be highly environmentally dependent. And if those IQ-beneficial learning environments are not available for high potential kids to be slotted into, then people with good genetic endowments will never develop them. The exact relationship between genes and IQ will be very sensitive not just to the family but the environments offered in the wider cultural context. Simple point, but Dickens/Flynn develop and model it in the context of the literature.

I think Cosma makes some similar points and says some other things about genes and inheritance in some other posts of his.

How does IQ correlate with academic success? I always thought there was a pretty decent correlation there.

Yes, there is. It does depend how you define academic success. IQ correlates better with success on (other) tests than with things like years of school completed, although it correlates somewhat with that too. Years of school completed -- an indicator of determination and discipline -- is a very important indicator of life success even when you hold IQ constant. (One of the many dishonesties in "The Bell Curve" is that they didn't control for years of schooling when looking at the effect of IQ on earnings, which biases their results significantly).

Academic success does not have an overwhelming correlation with earnings, though.

IQ really is an indicator of mental ability. It's performance on a mental task, after all. But there are lots and lots of mental and emotional tasks that are important in life. An overemphasis on IQ is simplistic, reductionist in obvious ways, and usually ideologically driven. The insistence that IQ is both all genetic and all-important is just scientific racism.

B, tell me: is a 19 year old late starter with a weak background going to learn several foreign languages? Or even one language well, if not necessary for his major? Is a 19 year old late majoring in the humanities going to learn much math or science? Is such a science major going to learn much humanities?

The elite European systems have the advantage that everyone who comes out of them has a pretty solid general foundation in what are called "distribution requirements" (often already by age 16). The American late starters I'm thinking of breeze their way through school not studying much of anything, even though they're bright. Even American students who work hard in HS often have serious deficiencies.

Languages, no, probably not. But (for example), during my year abroad in the UK I was surprised to find that humanities majors--people doing German, for heaven's sake--didn't know who Paul Klee was. Art wasn't part of the language A levels, so they hadn't studied it. The American late starters *I'm* thinking of *didn't* breeze their way through school; they pretty much ignored it and read/learned/did things they were interested in but that weren't on the curriculum, and didn't contribute to their grades. Later, when they decide to focus on getting a formal education a few (or several) years later than their peers, they often know a *lot* more about a lot of things than the kids who followed the curriculum, got As, and didn't really explore much outside of that.

And sure, students who work hard in HS often have serious deficiencies. That's part of my point.

The IQ argument is more complex, and comes from things like studies where identical twins are separated into different family environments but end up with similar IQs. Genes are constant, environment is different, IQ stays similar, hence it must be genetic.

Incorrect in a couple of different ways; it's impossible to determine accurately how different the environments are, you're dealing with a small sample size and large uncertainty, and the idea that twins experienced different environments ignores the fact that they experienced exactly the same environment for the first 9 months of gestation.

Sorry, that second paragraph of 560 should also have been italicized.

561 is in no way meant to imply that PGD was actually advocating the point of view he went on to debunk, twin studies being a particular red flag of mine.

The American late starters *I'm* thinking of *didn't* breeze their way through school; they pretty much ignored it and read/learned/did things they were interested in but that weren't on the curriculum, and didn't contribute to their grades. Later, when they decide to focus on getting a formal education a few (or several) years later than their peers, they often know a *lot* more about a lot of things than the kids who followed the curriculum, got As, and didn't really explore much outside of that.

B's got my back.

The American late starters *I'm* thinking of *didn't* breeze their way through school; they pretty much ignored it and read/learned/did things they were interested in but that weren't on the curriculum, and didn't contribute to their grades. Later, when they decide to focus on getting a formal education a few (or several) years later than their peers, they often know a *lot* more about a lot of things than the kids who followed the curriculum, got As, and didn't really explore much outside of that.

B's got my back.

Double back!

The beast with two backs?

No way. Tweety's got a girlfriend. And have you seen her shoulders? She could kick my ass.

Seriously, though, is there anyone who's taught mature students who *doesn't* think that they generally know a lot more than traditional students? I don't mean just the fact that they have a work ethic. I mean actual knowledge, including knowledge of things outside the focus of the course, and the ability to connect course material to those other things.

re: 569

I studied alongside mature students as an undergraduate. I was borderline 'mature' myself as I'd been working for four years previous to starting the course.

I don't think it was by any means obvious that they generally knew a lot more than traditional students. A couple of them, yeah, but the rest were often pretty rusty and out of date on most things and sometimes had a fairly inflated idea of their own 'wisdom'.

They often responded to things we were reading in interesting ways, and were more interested in working out their own response to stuff than in parroting what they thought were the right answers. But I'd dispute whether they were more knowledgeable over all.

My experience teaching has been that mature students do better. It was math classes, so the ability to relate to the real world is tenuous at best, but they seemed much more able to plan ahead and defer gratification.

You have to separate the general maturity / knowledge of the world that comes with aging from the depth and quality of the educational base laid down when you were young. I think Emerson was talking about the latter.

571 gets it right. At least for undergraduates.

Master's programs, meanwhile, are kind of the opposite. They comprise people who are just out of undergraduate and are extremely enthusiastic about the material and want to learn everything...with middle-aged people who are extremely busy with their own lives, only taking the classes to further their careers, and don't have enough time to study everything, so they are always the first to get the professor to clarify whether some material will be on the test or not, or whether some "mandatory" thing is actually technically mandatory.

The late starters I'm referring to are the ones who don't study at all until they're 19 or so.

I was using "late starter" in a purely descriptive sense, though I should have remembered that "late starter" commonly means "people who start college late (or finish college late), to the extent that it's a technical term.

People with elite educations will inevitably have blind spots, like everyone, but not the enormous blind spots that slacker American students do. And what I'm saying is that American college education is geared toward taking 18-year-old students with weak backgrounds and turning them into PhDs in about 10 years. And I claim that that leads to specialist narrowing.

The late starters I'm referring to are the ones who don't study at all until they're 19 or so.

If you mean "study" like "study to learn the things they're attempting to teach you in school" then I really do think you are talking about me.

And I certainly have blind spots, but arguably not more than somebody who studied straight through and came out the other side at 26 with an advanced degree.

Arguably less, even, or at least I know things they probably wouldn't have come across.

That Shalizi post is a beauty. When you imagine the 'net is just a sink of hatespeech and porno condoflippers, there's a reminder that the intellectualism is right here.

You have to separate the general maturity / knowledge of the world that comes with aging from the depth and quality of the educational base laid down when you were young.

Why? The claim is that people who don't "study" until they're 19 or so know less than people who study young. Either that's true, or it isn't, and if you compare what people know, and it's not true, then there you go.

I never studied in my life until fairly late in graduate school. I *certainly* never studied shit before college.

I think the belief that people who don't do well in K-12 are dumb is a pretty broadly held one. I also think that there are a ton of people who demonstrate how fallacious that belief is. It's a stereotype, nothing more.

By "study" I mean "study", Sifu. If you were educating yourself outside of school, you were studying. If not, not.

I got my B.A. at age 32, but I wasn't really a late bloomer. I took an 8-year bypass after my junior year and then returned part-time.

Both you and B are responding to something I didn't say, though I've acknowledged that it seems like I did.

To repeat again: what I tried to say, and PGD picked up on it, is that if you try to take an uneducated American 18-year-old and turn him into a 28-year-old PhD, the odds are that he'll have a narrower background than a 28-year-old PhD who went to an elite European secondary school. "Distribution requirements" in American colleges are weak and futile attempts to compensate for the fact that colleges usually have to start from scratch in the freshman year.

And that's a bad thing. I'm frequently shocked at the blind spots of American PhDs, which are often accompanied by methodological convictions which tell them that the stuff they don't know isn't worth knowing.

The elite European systems have the advantage that everyone who comes out of them has a pretty solid general foundation in what are called "distribution requirements" (often already by age 16).

I remember a depressing moment sometime near the end of my Freshman year in college when I realized that it was completely impossible to learn everything that I would put into the category of "things a basically educated human being should know."

I don't believe that any system anywhere teaches its students a complete foundation of "general knowledge" because that's way too big.

I can easily believe that other countries teach more students to have a solid foundation in basic math.

This is what I'm talking about: Yeah, for this reason the most frighteningly impressive people I know are products of the very elite end of the French system. They tend to defy the efforts of Anglo types to classify people as quant or qual or hard science vs humanities, etc.

499

"What, actually, is the ethical argument against individual universities setting their own entrance tests to select for the kind of student they want to attract. I know there must be one because it's a practice that's been in retreat throughout my life, but I've never seen a sensible explanation."

I believe the problems are more practical than ethical. Creating a test better than the SAT is expensive and difficult. Organizations don't write their own operating systems so much these days either. And there are a bunch of logistic problems as well. How would Harvard for example adminster its test to thousands of potential applicants all over the US in an economical and secure way?

579: well, so okay, you're saying that an American 18 year old with no intellectual curiosity but the drive to do well in school will be less well educated than a European with the same characteristics? Sure, I could see that.

I don't know why you said it the way you did, but yeah, that makes sense.

I actually started out saying that the American system (weak early education followed by a second chance in college) is good from an egalitarian point of view, but has a cost. As I understand, my error was in failing to say that it was in every way good, and entirely cost free.

Emerson is Heinlein?

http://www.elise.com/quotes/a/heinlein_specialization_is_for_insects.php

If you were educating yourself outside of school, you were studying. If not, not.

Flagrant question-begging.

if you try to take an uneducated American 18-year-old and turn him into a 28-year-old PhD, the odds are that he'll have a narrower background than a 28-year-old PhD who went to an elite European secondary school

And this proves that what? If you take an uneducated French 18 year old and turn him into a 28 year old PhD, the odds are that he'll have a narrower background than a 28 year old PhD who went to an elite American secondary school. In other news, if you try to turn my cat into a PhD, she'll probably know less than PK will by the time he graduates high school.

This is obviously because she hasn't studied enough, and she's already 15. Doomed to failure!

583: What I was trying to say, Sifu, is that given that reality, American PhDs tend to be intellectually narrower than they should be. Schools have to play too much catch-up, and it's the "distribution requirements" that give. Schools water tham down, and students shop for the easiest minimum requirement.

B, I don't know if you're trolling, but you're sounding like a stupid person. And you're using that "So?" argument again.

At this point I'm not really trying to convince you of anything, I'm just trying to get you to understand what I actually said, but I'm also asking myself "Why bother?"

587: no, I understand. But you're making a point about the weakness of American secondary schools, not about older students particularly.

The reason colleges can offer second and third chances is that they aren't dependent on students having a strong background, but the cause of those schools not being dependent on students having a strong background is the general weakness of incoming student's educational preparation.

I did not use the phrase "late starter" the way it is usually meant -- I didn't mean students who didn't get their degree at age 22. I meant students who slack through high school and get serious in college. "Late starter" is a good description of the latter group, except that it's already routinely and perhaps technically used for the former group.

495

"... Have you ever read Mismeasure of Man? It seems to me that it renders Jensen's racist stuff impossible to even talk about as science."

Yes I have read it. It has been a while but as I recall he is respectful of at least some of the investigators he discusses as trying to do a good job but straying into error because of an inability to be properly skeptical of results agreeing with their expectations. This is a well known problem now (not so much 100 years ago) and comes up in non-political contexts such as refinements of measurements of fundamental constants. It is the real point of Gould's book and applies to people on both sides of arguments.

But either of those applies to me, except that while I was slacking I was learning plenty of other things.

I think the problem with your phrasing is that it could be equally true of somebody who applied themselves all the way through high school, college, and grad school but wasn't intellectually curious enough to learn anything beyond what they were taught; the level of general education in many high schools is piss-poor enough that -- even if you're focussed on getting As -- you aren't necessarily going to learn much of anything.

Sifu, my point was comparing American HS to elite French HS. Even hard-working American HS students might be pretty weak compared to elite French HS students. But American schools also allow slacking.

I really was trying to make a very specific point, which was that American PhDs seem to be pretty narrow, and that I think that this is why. Whereas, as Gonerill said, elite French PhDs don't.

593: right, I get you.

Emerson is Heinlein.

the American system . . . is good from an egalitarian point of view, but has a cost.

Agreed. But I wouldn't describe our system as "(weak early education followed by a second chance in college)." And I think doing so depends on begging the question of what a "weak education" is.

I'd say (based on "general impressions") that American K-12 education is less focused on instilling content than it is on teaching certain habits of mind (independence, creativity). And also that it's more general than other systems, rather than the opposite.

Apparently what I'm saying is terriblt foreign to people here, judging by the response. Heinlein was talking about real-world skills, which I can take or leave. I'm just saying that it would be nice if Americans weren't ignorant of foreign languages, and if humanists weren't ignorant about sciences, and if scientists weren't ignorant about the humanities. But American are ignorant in those ways, mostly because of the weakness of secondary education.

591: You oughtta read it again, James - the point you mention is very commonplace in Gould's books - indeed he tells an interesting anecdote about how a set of objective measurements he made matched his expectations rather than reality.

But that's not a terribly relevant point to the question at hand. The question is: What is the scientific basis for arguments that some races have been demonstrated to be more intrincically intelligent than others. And the answer is: There is no appreciable scientific basis.

(And once he establishes that, you're right that he goes into some discussion about how these things keep coming up in essentially the same form over and over again - which I think speaks to the point you mentioned.)

Heinlein was talking about real-world skills, which I can take or leave.

Another key difference between Emerson and Heinlein is that there's no masturbating to Heinlein any more.

OK. American education is weak in instilling content, and it's general in the sense that no specific topic is taught well.

As you see, I disagree. But I'm glad that colleges have ways to pick up the slack.

JE, your are being more earnest than usual. The Heinlein thing was a joke. The trend to specialization even in elite liberal arts colleges is more a function of modern life than poor education policies. With facts at your fingertips via google, what does smart mean anyway?

I do think that Europeans are a lot more likely to speak more than one language, and to travel. But I don't think that's because of their formal education system as much as it is simple geography.

(Don't most college prep US high schools require some foreign language? And don't most European schools not?)

(Don't most college prep US high schools require some foreign language?

My (very good! public!) high school required "some" foreign language instruction. Two years. That basically works out to none, though the people who actually did do four years of the same language got good.

I satisfied it with Latin, of course.

I'm not really talking about a trend. I think it's been true all along, though it may be getting worse.

In general I think that overspecialized people do worse work even in their own fields. Some of the stuff that psychologists and economists say is horrendous.

603: ditto! I learned nothing.

My elementary school also required two years of foreign language instruction -- either french or spanish -- again, I learned nothing.

My (initial) college required one semester of foreign language instruction. I took Chinese, and learned a surprising amount.

603: Well right, mine too, and I did Spanish from kindergarten all the way through my freshman year of college. But my point was that I don't think foreign languages are required at all in the UK, at least. Dunno if they are in European schools, though my guess is probably. (Hence, I was once again using "most European" to mean "UK," because I am a racist.)

Even smart people can be stupid, sometimes. I don't think anyone has the capacity to be smart about everything, all the time. Present company excluded, of course.

I've heard that colleges routinely (in the absence of better information) covert four years of HS language to one year of college language. I went to college with two HS years, and I had to start from scratch after basically failing French 200. The whole episode was very damaging to my college career.

Don't they do placement tests for things like foreign language? Even with multiple years of a language why not just start with 101, and ace it? You had a bad counselor, John.

I managed to convert one year of HS German in to one year of college German. (It was second-year German, though.)

Big win!

Now we understand why John thinks that US K-12 education is worthless.

The quality of language study varies quite widely. Even within colleges "1 year" can have hugely different meanings. I did one year of beginning russian one summer at a highly respected school and found that I fit better in 3rd year than in 2nd year at another highly respected school. And then later found that at some schools the 1 introductory year I did - if you measured by how much of the textbook is covered - takes 2 years.

You're talking like a stupid person again, B. I wish D^2 were here to kick your conventional educationist ass. I don't have the beans.

Oh, and at Berkeley my four years of HS Spanish would have been treated as one year, had I decided to take more Spanish (and at that point, I could read stories and essays in Spanish and had learned all the grammar there was to teach at the high school level). I'm pretty sure that kind of thing is fairly common.

498

"Shearer, just because results of a test are consistent over time doesn't mean it's a good test for what you want to test for. It means it's a test for something, but doesn't tell you anything about what. You could give me the SAT over and over until I plotzed, and I'd keep on getting top-end scores. This would not make me any likelier to be a creative mathematician, nor would it make Jane Smith, who makes calculation errors under pressure but has more mathematical aptitude than I, any less."

The thing the SAT math and similar tests measure is not the only thing that contributes to being a successful mathematician but it is a contributor and the more you have of it the better. And it has been a long time since I took the SAT but I don't recall the math exam as being particularly calculation heavy. Anyway can't you use a calculator these days?

552

"ummm, one could say that they are. Or at least that g is so essential to the measurement of IQ that the two stand or fall together. Are you sure you're familiar with psychometrics?"

You could say that but you would be wrong. There are people in the field (g-men) who make a fetish of g and thus make easy targets but g is not really neccessary.

I have a lay knowledge of field but don't follow it closely.

613: John, don't be silly.

Seriously, B., You're very status quo on education, except that you want more money spent on it, and you want it to be egalitarian -- which it can't really be because grading is what education does.

598

"But that's not a terribly relevant point to the question at hand. The question is: What is the scientific basis for arguments that some races have been demonstrated to be more intrincically intelligent than others. And the answer is: There is no appreciable scientific basis."

This is a strawman based on the strength of the word demonstrated. You could equally well ask: What is the scientific basis for arguments that the races have been demonstrated to be intrinsically equal in intelligence and receive the same answer. A more meaningful question would be what is the evidence for each position. Gould does not really deal with this, much of the book is shooting fish in a barrel, criticizing 100 year old work which no one relies on today.

I'm status quo on education? Because I'm the one arguing that people without formal educations are often as smart and well-informed as people without? Remember I'm the one that chose to send my kid to a school with shitty test scores rather than the "best school in the district."

I am very much *in favor of* formal education. But that's not the same thing as being in favor of the status quo.

447

"... Especially since he seems to hold me in much the same contempt that I hold him."

Well B you are a bully which I don't care for. And I am biased against people who are rude to me. Not that you are the only offender but you do stand out.

I have never seen you receptive to a specific criticism or proposal for change.

Education is often a rightwing political football and when that routine starts I loyally defend the schools, but I doo think there's something seriously wrong with American education, especially ages 14-22. (And something different wrong 22-infinity).

I have never seen you receptive to a specific criticism or proposal for change.

I've made specific criticisms and proposals for change. What I'm not receptive to are easy cliches.

Anyway, the thread's getting too long and I'm wondering if I can get a nap in.

A nap in here? It seems sort of noisy.

You're not receptive to others' easy cliches, B. 596 was adequately cliched for most purposes.

You could say that but you would be wrong. There are people in the field (g-men) who make a fetish of g and thus make easy targets but g is not really neccessary.

If you really want to make an argument that "intelligence" can be usefully reduced to a single metric then I believe you need g or something similar. If you simply want to give a test and say the score on it will be correlated with something useful then you don't. The controversy around IQ is driven much more by the former claim than the latter.

625: Whereas "American education is worthless, and kids arrive in college requiring remedial teaching" is a shockingly original statement.

So maybe we could have argued about the content of what was said, instead of talking about easy cliches?

To say that 18 year old American students from good schools are usually two or more years behind students from elite European schools isn't meant to be original. It was meant to be a truism.

I do think that Europeans are a lot more likely to speak more than one language, and to travel. But I don't think that's because of their formal education system as much as it is simple geography.

Well, the traveling makes the language skills more useful, but where do you think they're learning these languages, if not in school?

I can only speak to Germany, but even in the non-college track schools, English is a requirement.

where do you think they're learning these languages, if not in school?

From tourists.

Dunno if they are in European schools, though my guess is probably

Yes, in most countries compulsory English. Whether this is a good idea (as opposed to allowing an option of e.g. Mandarin or Arabic) I'm not sure.

630: when I visited Angkor there were these little street (dirt? Forest?) kids who could recite a few facts about your country of origin, in your own language, in flawless [ your language ]. They came up to us and -- upon discovering we were from the US -- recited the population of Washington DC and a couple other surprisingly detailed facts. I can only assume they had the same spiel ready in many different languages, given the multinational character of the tourist population there. It made it all the more depressing to realize they would probably never see the inside of a school or make more than a dollar a day.

Still, I coulda taken 'em.

I coulda taken 'em.

"I coulda tooken 'em" is correct when speaking of five-yearolds.

626

"If you really want to make an argument that "intelligence" can be usefully reduced to a single metric then I believe you need g or something similar. If you simply want to give a test and say the score on it will be correlated with something useful then you don't. The controversy around IQ is driven much more by the former claim than the latter."

Well, I don't really see the distinction between the first claim and the second claim unless you are taking the first claim to be saying the single metric captures everything you could want to know about a person's intelligence which I think is a strawman which no one is actually arguing.

As I understand the argument about g, a problem with IQ is different IQ tests measure slightly different things depending on the mix of questions. From a practical point of view this doesn't matter too much as the different IQ tests tend be highly correlated and it doesn't make much difference for most purposes which one you choose to use. However from a theoretical point of view it would be nice if there were some objective procedure to come up with an unique definition of IQ. The g-men claim to have such a procedure but there are certainly doubters. The details are highly technical and tend to make my eyes glaze over but it would not surprise me if the skeptics are correct at least to some extent in claiming that the g-men are overstating their case.

I don't really see the distinction between the first claim and the second claim unless you are taking the first claim to be saying the single metric captures everything you could want to know about a person's intelligence

No -- that's not the only distinction at all. Say I wrote you up a test that quizzed you on the names of the President, VP, Majority leader of the Senate, Speaker of the House, members of the Supreme Court, and also asked you to name several elements off the periodic table. I think it's possible the score on such a test would be pretty correlated with earnings, and with a number of other life outcomes (e.g. years of education). But no one would take it is a measure of "innate intelligence" or a similar type concept. IQ doesn't have power because it's a test score that's correlated with stuff. It's powerful because we presume it measures something innate, inherent, unchanging. The "g" argument is absolutely crucial to making that case.

635

"... IQ doesn't have power because it's a test score that's correlated with stuff. It's powerful because we presume it measures something innate, inherent, unchanging. The "g" argument is absolutely crucial to making that case."

I don't see this at all. "innate, inherent, unchanging" are empirical qualities which are largely independent of whether intelligence is best modeled in a single factor or multi-factor way. A single factor could in theory depend on how much of some nutrient x is in your diet and multi-factors could in theory all be strictly genetically determined.

As I understand the argument about g, a problem with IQ is different IQ tests measure slightly different things depending on the mix of questions.

I don't think this is the problem with g. I thought the problem with g is that no one can identify anything to measure that can't be taught.

There are lots of problems with g, both what Heebie said and in Cosma's discussion of factor analysis.

But in response to 635, the larger point is that what lifts IQ above being just another test score is its literary/metaphorical power. It is supposed to represent "inherent intelligence" or some such. But there is little or nothing objective about IQ scores that justifies such a claim. It's heritable, but so is lots of other stuff. There is no direct evidence at all that it's "genetic". It is modestly correlated with some life outcomes, but so is lots of other stuff.

If you really wanted to come up with a good predictor of life success, it wouldn't look much like IQ. It would be multi-factor, include emotional measures, and include measures that were based on observing actual behavior over time instead of just in a testing environment.

But if you wanted to come up with a relatively quick paper-and-pencil test for use in sorting many applicants in a large bureaucratic environment, like say the military, it might look a lot like the IQ test. It might be useful in such a setting too.

637

"I don't think this is the problem with g. I thought the problem with g is that no one can identify anything to measure that can't be taught."

This isn't true at all. There have been lots of programs which try the raise the IQs of low IQ groups with little long term success.

638

"If you really wanted to come up with a good predictor of life success,"

IQ is not supposed to be the best possible predictor of life success. It is trying to isolate and quantify the trait commonly called intelligence. No one disputes that there are other traits such as conscientiousness which are also very important for life success.

Pick your battles, James.

There have been lots of programs which try the raise the IQs of low IQ groups with little long term success.

This is not true at all. The IQ test has to be recalibrated every few years, because everyone keeps marching upwards. Just because the low end is still the low end doesn't mean they're not learning the test.

I mean, pick them better.

194: "you're"

642

"This is not true at all. The IQ test has to be recalibrated every few years, because everyone keeps marching upwards. Just because the low end is still the low end doesn't mean they're not learning the test."

The causes of the long term rise in scores are uncertain. Better nutrition and health care has been suggested as a cause. As has greater intellectual stimulation from things like TV. Causes like this do not constitute learning the test. People have had little luck taking a group A and trying to teach them to do better on IQ tests than a matched control group B with no special instruction. When gains are seen they tend to be temporary.

I do not believe this to be the case. Picture me writing an impassive set of facts disputing your statement. Thank you.

People have had little luck taking a group A and trying to teach them to do better on IQ tests than a matched control group B with no special instruction. When gains are seen they tend to be temporary.

I haven't got the time to look up references here, but if you're actually interested in this stuff, Shearer, I'd look into your basis for this. As I recall the research, interventions that raise IQ are a dime a dozen; the problem isn't that such interventions haven't been devised, it's that they're prohibitively expensive on any useful scale.

647

"I haven't got the time to look up references here, but if you're actually interested in this stuff, Shearer, I'd look into your basis for this. As I recall the research, interventions that raise IQ are a dime a dozen; the problem isn't that such interventions haven't been devised, it's that they're prohibitively expensive on any useful scale."

Here is a generally favorable review of Head Start which concedes the IQ gains do not last. I believe this pattern (initial gains that fade away) is typical for interventions targeting children.

And of course if some expensive intervention actually produces permanent gains in intelligence which are reflected in IQ scores this is not a defect in the tests. I took heebie-geebie's 637 to be a claim that IQ tests are defective because they are easily gamed without actual increases in intelligence.

Now that you have seen what happens to the threads you kill, James, the ghosts of threads present and future, and you will learn how your life of long-winded trolling will eventually kiiiiiilllllllll yooooooouuuuuuuu........

.....uuuuuu.....

uuu.....