December 10, 2012

Education Realist responds to Unz essay on college admissions

Education Realist has a long post up entitled "An Alternative College Admissions System." Here's a sample:
….the overrepresentation of Asians is explained more by their dominance in GPA, as opposed to test scores. And that’s harder to fix. It’s easy enough to tell white kids with high test scores to go to test prep and maximize their scores, but by junior year, the GPA damage has been done. 
What that means: no more room for, say, the idiosyncratic white boy who scores 2250/34 on the SAT/ACT, scored 4s and 5s in 7 AP tests, got 780, 730, and 690 on the US History, English Lit, and Math 2c, but whose weighted GPA is a 3.8. 
So just raise the GPA, you say. White parents need to raise their expectations for their own kids. Hahahahaha. This is me laughing. Unless the white kid is ruthlessly driven and competitive on his own merit, parental pressure as a means of raising his or her grades to the degree needed to compete with Asians is a non-starter.

Okay, but this isn't to say that white parents shouldn't push farther out on the diminishing returns curve than they are doing right now on average.
Amy Chua isn’t kidding. If a white parent tried to drive her kid the way Amy Chua did hers, the kid would end up in therapy, and the therapist would make the parent stop. Asian parenting techniques are abusive in white people world. Full stop. (What disgusts me most about Chua’s story is not her own behavior, as she doesn’t know any better, but that her white husband stood by and let her abuse her daughters. But then, I’m a white parent.) 
Not only does this difference between white and Asian cultural expectations lead to lower GPAs for whites, but smart white kids with B averages are then denied access to AP classes (in most Asian schools, access to AP is strictly limited by GPA), which put even a lower ceiling on their GPA. 
And finally, understand that those Asian good grades do not necessarily translate to a well-educated student. As my primary second job, I teach enrichment at a private educational company (aka, an Asian cram school), which over seven years adds up to a lot of Asian high school students. I love them. They’re great kids. But my experience has taught me to question any straightforward comparison between white and Asian academic credentials. 
All of my enrichment kids, as sophomores, are taking honors English and pre-calc. Maybe 10% of them can reliably read a complex text and offer an interesting or informed analysis without referring to Wikipedia and repeating verbatim what they read there, and in seven years and probably 300 kids I have never once had a student who could explain the derivation of the quadratic formula (that is, the generalized case for completing the square). 
I also teach an AP US History prep course every year, at two different locations, to a dozen students per class. All but a few kids each year will have taken six months of APUSH by the time my class starts, and fewer than a quarter of them have ever known who wrote the Federalist papers, or the most important achievement of the Northwest Ordinance of 1787, when the class begins. Very few of them can even make a stab at naming the presidents in order, or even identify any of the “forgettable” presidents. These are kids attending public schools with some of the highest SAT averages in the country, more than a few of them topping out at 2400. [Northern California]
In comparison, I’ve tutored and taught (in public schools) a lot of bright white kids and their awareness and retention of their own education, including the above benchmark questions, is far superior, on average. There are, of course, white soulless swotters and creative inquisitive Korean eccentrics. But the betting goes the other way. 
So, for the grade manipulation that goes on at the bottom end of the scale, and the cultural skew that goes on at the top end of the scale, grades are just flatly useless. Unless or until we move to a system in which grades are taken out of teachers’ hands and determined by outside standardized tests, grades must be eliminated from any truly meritocratic admissions process. End rant. 
(Two points before I go on: 1) bright Hispanic and black kids are also more likely to retain their knowledge than Asian kids, but they are rarer and are going to write their own tickets regardless; 2) just as Asian test performance may overstate their abilities, black test performance may understate their abilities because the tests focus too much on abstraction and generalized situations. That’s another reason I want a much more competitive test market, to see if perhaps we can find a more meaningful way to test the bottom half of the ability spectrum. )

The real money would likely be for new tests for the high end, however. White people in Park Slope would like some objective test that shows their kids really are as amazing as they think they are.

For 45 years, people have wanted a test that closes The Gap. But, blacks really aren't going to be that big of a deal in the America of the Future. 

The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on. 

In theory, blacks and whites could wind up on the same side on testing reform. Some clever Park Slope SWPL might someday come up with a new test that cuts down on Asian advantage over whites and then push it through because blacks do relatively better on it than on old tests. Recruit a third-generation Asian-American to push it as well, and it just might work.

92 comments:

sunbeam said...

The elite schools really wouldn't be an issue if places like the state department and Wall Street would recruit and hire at places like Oklahoma State.

Instead our guy goes to work for a bank or insurance company in Oklahoma City or St Louis. Probably for a tenth the money in the long run. Plus if you really have talent at that, the leverage you can apply at a Wall Street can lead to an income that is astronomical compared to our guy.

But they don't. The problem is that you can be an Oklahoma State grad, and be perfectly capable of being an analyst on Wall Street. But they get their guys from a small cadre of schools.

You can backdoor in, but the way that has the greased rails goes through a small number of schools.

I mentioned the State Department. Also most of the really high echelon public jobs come from the same schools.

It might be more useful to do something about a system that hoards the big rewards as it does.

Anonymous said...

I've always disliked Canada's university admission scheme, being 100% based on high school grades, but the more I read about American admissions...

This is something I haven't seen: a chart comparing university admissions process by country. Canada is the only country I'm aware of that does not do SAT type testing - are there others?

Education Realist said...

Hey, all my posts are long. It's the law. And thanks for the link!

Don't think I'm certain that we can get "better" tests for the lower half of the ability curve. I would just like to see us try. I've tutored low ability kids (of all races) on both the ACT and the SAT, and they almost all do better on the former, unless they are
terrible readers.

Good essay tests would do a lot to even the playing field between whites and Asians. Check out the AP score distributions for English Lit/Lang & Comp/US History compared to math and science scores. And writing an essay about math and science? Worst of both worlds.

And on the Eng/History tests, weak writers can compensate with a strong mc performance, which won't be possible on an essay test.

Yes, I agree, the real money is going to be at the top end.

Don't make me sound like an idealist, trying to make life better for URMs! The whole point of my plan is to make it difficult for universities to hide their skew and let employers find the smart kids regardless of race and education, and smart kids not take too much of a hit when the elite schools crowd them out with diversity and URM admits.

jody said...

Educationrealist blogs anonymously, 'nuff said.

Steve, you're the only man brave enough to take a stand.

Education Realist said...

Okay, but this isn't to say that white parents shouldn't push farther out on the diminishing returns curve than they are doing right now on average.


Oops, left this off.

It doesn't matter whether they do or not.

A 2400 test score doesn't guarantee admissions for a 4.2 GPA, so it won't do a thing for a 3.8 GPA.

So no, there really isn't much point in most whites going overboard to raise their test scores.

In my experience, white kids with high GPAs are competing well on test scores. Remember, universities are biased in favor of them because they don't want too many Asians. A white boy with a 4.2 GPA and a 2200 test score is golden; an Asian boy with that profile is one of thousands.

So whites with excellent GPAs and good test scores are doing well, getting into elite schools. There's just not that many of them with high GPAs, relative to their underlying abilities.

john marzan said...

"The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on.
In theory, blacks and whites could wind up on the same side on testing reform. Some clever Park Slope SWPL might someday come up with a new test that cuts down on Asian advantage over whites and then push it through because blacks do relatively better on it than on old tests. Recruit a third-generation Asian-American to push it as well, and it just might work. "

de-emphasize math in testing. this is a good start.
http://www.nytimes.com/2012/07/29/opinion/sunday/is-algebra-necessary.html

as for getting an asian american to do that, good luck. why would the asians give up their superiority to improve white people's self esteem?

Josey said...

"The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on."

It's called the LSAT


"Average scores on the LSAT, whose range is 120 to 180, vary among races. Here are averages for law-school applicants in fall 2004:

Group Average score
Caucasian/White 154.3
Asian/Pacific Islander 154.2"

Asians and whites tend to swap places at the top from year to year.

http://online.wsj.com/public/article/SB111279937006999640.html?mod=todays_free_feature%2520

http://carrefoursagesse.wordpress.com/2009/03/10/breakdown-of-sex-and-racial-group-differences-in-lsat-scores/

Education Realist said...

Educationrealist blogs anonymously, 'nuff said.

So does LaGriffe. So did the Mystery Pollster for a number of years.

Steve has done an awesome job at carving out a niche, and is making this area respectable. But I'm not a writer, I'm a teacher. I would almost certainly lose my job and not be able to find others if my name got linked to this blog. Given that my identity is pretty easy to determine, I think I'm pretty brave--well, stupid, really--to blog at all.

Besides, what does my anonymity have to do with my posts? The opinions either have value or they don't. My identity isn't really relevant to their value.

Josey--Asians are pretty close to whites on the LSAT. I still think analytical essays are going to be an important part of the equation.

And at that point, whites could also boost their scores on mult choice tests with test prep. It'd be worth their time in this system.

Anonymous said...

It might be more useful to do something about a system that hoards the big rewards as it does.

It does seem that the elite school alums are dining out for an awfully long time on a good score on a standardized test taken when they were 17.

Why might this pattern of recruiting persist? Perhaps from a moneyball standpoint the recruiters don't have enough inexpensive, fast, reliable tools to select good students from Oklahoma State. In other words, you can get fine recruits from Oklahoma State, but you might also have to winnow out more duds, and they're more worried about the duds.

Lawful Neutral said...

Perhaps from a moneyball standpoint the recruiters don't have enough inexpensive, fast, reliable tools to select good students from Oklahoma State.

Naw, they have the tools, but they're forbidden to use them. It's called disparate impact.

socks said...

The GRE was renormed not too long ago. I was surprised when I found out because I didn't remember anyone in HBD-land covering it.

The renorming went the other direction, raising what had been a low math ceiling (800 = 94%) and lowering what had been an very high verbal ceiling(720=99%).

Anonymous said...

"It does seem that the elite school alums are dining out for an awfully long time on a good score on a standardized test taken when they were 17."

Unless you take a blow to the head, chances are the IQ you have now is the same with respect to the IQ of your peers as it was at 17. But it is illegal in the US to hire on the basis of IQ (the military must get an exception), so university is a very good proxy.

Some software companies looking to get high IQ people give "logic puzzles" that are effectively those questions only answerable by people with a given IQ level.

DaveinHackensack said...

Sunbeam,

Oklahoma State alumnus can go here and sign up for the State Department's foreign service officer exam.

And Wall Street career tracks are shrinking for everyone. Hundreds of thousands of jobs have evaporated since 2008, and companies continue to announce massive layoffs (most recently, UBS and Citigroup).

DaveinHackensack said...

"Perhaps from a moneyball standpoint the recruiters don't have enough inexpensive, fast, reliable tools to select good students from Oklahoma State"

For entrepreneurship, accelerator programs such as Y Combinator and Tech Stars are pretty cost- and time-effective: about three months and maybe $10-15k per young buck. For a big company, it might make sense to start with this model and tweak it for its own purposes. For example, instead of groups of 2-3 founders, make it groups of 8 or 12 or something, so they get a better feel for leadership skills and group dynamics. And instead of having teams develop start-ups, a big company accelerator could toss them an orphan product and give them an unimportant market to run with it in.

Also, no need to put these big company accelerators in Silicon Valley or NYC or somewhere like that. Put one in St. Louis; put one in Fresno, etc. If you can churn through and evaluate 4x as many college grads in 1/4th the time, you can be less selective in who you let into the funnel.

Anonymous said...

It's called the LSAT

Asians that take the LSAT are who's leftover after PhD programs, STEM fields, medical/dental, and MBA types. So, it's not a good benchmark. In addition, although over-represented Law school, it's not nearly to the degree as in other fields mentioned.

I don't think testing is the problem. The current process is balanced enough sort out ambitious talent. It's the culture. Whites are late-bloomers, relatively, when it comes to academic/career ambition and not enough parents push their kids, except in EVERY other form of competition. I don't think a change in testing will produce results much different. White culture shouldn't have to accommodate in either direction.

What people are saying here is "Let's figure out a way to be more competitive, without being competitive." That sounds like something an affirmative action proponent would think.

The only answer is ending immigration and multiculturalism.

Anonymous said...

"I would just like to see us try. I've tutored low ability kids (of all races) on both the ACT and the SAT, and they almost all do better on the former, unless they are
terrible readers."

I hear this said a lot... that the ACT is easier on the low end. However, the ACT cannot be easier on the low end for everyone - if it is easier on the low end for some low end students, it must be harder at the low end for some low end students.

Why? ACT raw scores are converted into a score between 1 and 36 in such a way that the resulting distribution is normally distributed. So, there is a perfect correspondence between a given score on the ACT and the position/percentile you are at on a normal distribution of ACT scores. The bottom 20% of students will always be within the same score range on a test where the numerical score is nothing but a veil for the percentile of the student and the bell curve that the test is normed to match.

Now, you might argue that some kids who score below the 20th percentile on the SAT score above the percentile on the ACT. But then there must some kid who scores below the 20th percentile on the ACT and above the 20th percentile on the SAT. In this case, this kid (and all others like him) constitute a new low-end, in which case you could just as easily argue that the SAT is easier on students on the low end.

In the end, it IS helpful to take both in case you do markedly better on one than the other, but not all low-end students can be winners on a test that is essentially a ranking with a number attached to it to make it seem like less of a ranking. Someone must fall below you and become a "low-end" kid if you are to move to the higher end, so I don't see how one test could be easier on the low-end kids than another.

Anonymous said...

The only answer is ending immigration and multiculturalism.

I thought we knew this.

Furthermore, even lower-income Asian parents do everything they can to help pay for college. This is a huge emotional push that not many White kids get. It's one of the many reasons why even smart White boys only maintain good enough grades that will get them into the cheaper tuition (State) schools.

Simon in London said...

Funny, I thought the main message of the Unz article was the discrimination in favour of Jewish applicants (ca 900% compared to general population) and the massive discrimination against non-Jewish whites (ca 30% vs population), not the cap on Asians (still 300% vs population).

Anonymous said...

Here is a good test:

http://www.all-souls.ox.ac.uk/content/Examination_Fellowships:_General_Information

Mwahahaha!

Anonymous said...

I think there is very interesting observation in this article:

Memory Retention

Maybe this is it? Maybe Asians don't have the long term memory retention as whites?

Creativity is like Jazz Music. You have to have a true grasp on all the chords and then see the chords as tools in your toolbox...you create from there.

Maybe that's it...lower long-term memory retention in Asians leads to lower creativity...no toolbox to draw upon.

J said...

Education Realist forgets what the tests are for. They are for predicting future performance. And reject those who education will not help. And current tests predict with high accuracy future performance. For all races. Any test manipulated so that would "close" the gap between them would automatically falsify the expected outcome. I couldnt care for a less "racist" test that measures anything but future achievements.

Peter the Shark said...

Have things changed that much? Back in the '90s elite schools routinely let in kids with stratospheric SAT scores and middling GPAs. One of my brother's friends (a white Lutheran) got into an Ivy League school on those credentials (he was a decent athlete too, to be fair, but not a superstar).

Now here's the twist - my brother's friend rode his high IQ and test scores into a very good grad school, but has continued to underachieve professionally in corporate life ever since. He didn't like hard work in high school, and he never learned to as an adult. If I were hiring a financial analyst or a marketing manager I would probably hire an Asian grind rather than a brilliant spoiled white kid.

Anonymous said...

Maybe that's it...lower long-term memory retention in Asians leads to lower creativity...no toolbox to draw upon.

I don't think that's it. If it were, then Blacks would... and Asian doctors would kill their patients...

eah said...

And that’s harder to fix.

I know exactly how to fix it.

Anonymous said...

Aren't IQ tests supposed to do this? They're designed so that you can't really study for them.

Anonymous said...

That’s another reason I want a much more competitive test market, to see if perhaps we can find a more meaningful way to test the bottom half of the ability spectrum. )

Since the market here is for standardized testing, how exactly would competition work? In order to have standardized testing, you'd have to converge towards a single, standard test. If you had many different tests, with different schools preferring different tests, you'd effectively not have standardized testing. It'd be like before the SAT when each school had its own entrance exam.

Anonymous said...

(Two points before I go on: 1) bright Hispanic and black kids are also more likely to retain their knowledge than Asian kids, but they are rarer and are going to write their own tickets regardless; 2) just as Asian test performance may overstate their abilities, black test performance may understate their abilities because the tests focus too much on abstraction and generalized situations.

If the bright Hispanic and black kids are retaining their knowledge better than the Asian kids, doesn't this mean that the bright Hispanic and blacks are doing better than the Asian kids on tests? And if the tests overstate Asian ability and understate black ability, doesn't this suggest that the bright blacks kids are more intelligent or have higher ability than the Asian kids?

Anonymous said...

"In other words, you can get fine recruits from Oklahoma State, but you might also have to winnow out more duds, and they're more worried about the duds."

I think it was Scalia who, when asked why he only recruits from two or three of the very top law schools in the country for clerks, said: "You can't make a sows ear out of a silk purse."

Anonymous said...

That’s another reason I want a much more competitive test market, to see if perhaps we can find a more meaningful way to test the bottom half of the ability spectrum.

Funny people, good writers, good artists, etc. tend to be above average in intelligence, and even if they're not, their ability tends to be quite noticeable. Idiot savants are quite noticeable as well.

Anonymous said...

"Furthermore, even lower-income Asian parents do everything they can to help pay for college. This is a huge emotional push that not many White kids get. It's one of the many reasons why even smart White boys only maintain good enough grades that will get them into the cheaper tuition (State) schools."

This seems to be much more common in both Jewish and Asian cultures - high parental investment. All else equal, this nepotistic strategy has to win out over rugged individualism, when each generation gets a leg up because its parents take a dynastic rather than individualist view.

In a world where there is not much frontier left, we need to adopt this strategy to compete.

Anonymous said...

"Asians that take the LSAT are who's leftover after PhD programs, STEM fields, medical/dental, and MBA types. So, it's not a good benchmark. In addition, although over-represented Law school, it's not nearly to the degree as in other fields mentioned."

I think this may be a big part of it. The very smartest of the Asians are so heavily over-represented in STEM, med, etc. that there just aren't that many smart ones left by the time you get to law. They are only about 5% of the population.

Anonymous said...

And if the tests overstate Asian ability and understate black ability, doesn't this suggest that the bright blacks kids are more intelligent or have higher ability than the Asian kids?

A bright any race obviously has higher abilities than an average Asian kid. Or were you suggesting bright vs bright?

Are we really going there? Fine. Bright relative to other blacks or overall? Overall, there aren't too many or is there? By what measure? Oh, back to square one.

Anonymous said...

All of these types of tests are endorsed by big business, as was the establishment of compulsory education.

They and all of the other ways our society is organized are designed to identify and cultivate profitable workers.

Intelligence is a strong part of it, but obviously they don't test for rhythmic intelligence. And EQ, creativity, and a host of other types of intelligence aren't tested for either.

Ivy league universities have a special mandate: Produce profitable leaders. That is why their admissions criteria aren't purely based on scores and GPA's.

If you want your kids to get an education, send them to a NESCAC schools. That's where most of the old Ivy crowd migrated to.

Anonymous said...

First of all, whites should send their kids to smart white schools not ones full of Asians. let them kill each other academically. also, republicans need to push for an end to elite college discrimination against whites, under threat of loss of funding or tax breaks. Asians are massively overrepresented at ivies while they still complain about discrimination. send them home.

Orthodox said...

"And finally, understand that those Asian good grades do not necessarily translate to a well-educated student."

This is why I will homeschool.

Anonymous said...

Memory Retention

Maybe this is it? Maybe Asians don't have the long term memory retention as whites?

Creativity is like Jazz Music. You have to have a true grasp on all the chords and then see the chords as tools in your toolbox...you create from there.

Maybe that's it...lower long-term memory retention in Asians leads to lower creativity...no toolbox to draw upon.


This is an interesting idea - that intuition requires great long-term memory. It's not the first time I've heard it, either. It certainly seems like this could be the case. Obviously if you don't have the concepts stored in your memory, you can't relate them. So an excellent long-term memory is probably necessary to the sort of creative, intuitive thinking that relates known concepts to new problems. Is it sufficient?

I think to be creative or intuitive in this manner, one needs to also have the sort of mind where one can easily recall similar examples to any given idea. A sort of mental search engine, if you will. Minds without this function but with good long-term memories might be good at say, recalling the dates of particular wars, but not at say, noticing anything interesting about the wars that they have studied.

DC said...

“Very few of them can even make a stab at naming the presidents in order…”

What pct of white students could do this? What pct. of Americans can do this? What’s more is this, and the other items you mention, actually a good way to judge a creative, clever mind, or is it more of the same, the very soulless rote memorization of facts that you chastise in Asians? High school and Undergraduate school work in general is slated towards working hard at memorization of facts, less on insightful reflection and creativity. Maybe it should be if we're training office workers and technicians at these levels. Grad school is where you finally get to see if people are clever and Asians do quite well there.

The Anti-Gnostic said...

It does seem that the elite school alums are dining out for an awfully long time on a good score on a standardized test taken when they were 17.

Indeed. OTOH, we've seen that Wall Street's genius-IQ quants can crash and burn like the rest of us. Well, except the Fed bails them out.

I'm actually not as concerned about this. In my field, the big law firms chew up and spit out tons of geniuses with no socialization. Only people with the right mix of genius and 'people' skills make it. I see very few Asians left standing in the big firms after a few years though I'm sure some are. I'm guessing it's not so different in STEM companies.

I'm also not so worried because I favor white disengagement. We've lost the battle, so let the elite have their unsustainable playground. Whites will stop joining their military, stop buying their rigged financial products, and start withdrawing consent to their rigged government.

You can hire all the Asian geniuses you want. Outside their homeland anthills, they're still reliant on a cadre of smart/super-smart white guys with a clue about the real world. Maybe I'm being too sanguine about this.

Anonymous said...

@Educationrealist

Design a test which only pale skin will help. For example, small text in Asian skin color were put on students forearm for quick reading and comprehension. Lack of contrast in background in Asian skin would severe restrict Asian student performance. Be creative.

Anonymous said...

"Funny, I thought the main message of the Unz article was the discrimination in favour of Jewish applicants (ca 900% compared to general population) and the massive discrimination against non-Jewish whites (ca 30% vs population), not the cap on Asians (still 300% vs population). "

If you want a proxy for what Ivy admissions should be, look at the enrollment in top music conservatories. General interest in classical music and the ability to master it is tightly correlated with intelligence. Probably one of the most g-loaded tasks of all is music sight reading, and one of the most popular day jobs for classical musicians is computer programming.

Karen said...

DC -- my son went to Gorzycki Middle School in Austin, Texas. (Google it; it's the only school with that name in the English-speaking world.). The first day of 8th grade all students had to list the firs 16 Presidents. They will have to list the rest of them as sophomores at Bowie High School.

In fact, you all might consider examining Gorzycki and Bowie for this discussion. My son Andre made 2nd honor roll, with 3.0 GPA, eliminating him from any Ivy. Note, however, that about half his class makes the honor roll. The school is 60% white, with Asians and Hispanics at a little more than 15% each, and 5% black. (Austin is only 6% African American anyway.). I rally want to meet those laid-back white people who don't push their kids, because I really don't know any.

Education Realist said...

I was surprised when I found out because I didn't remember anyone in HBD-land covering it.


I mentioned it after the fact. I don't think it's as useful an IQ test as it was, but it's more meaningful to grad schools than it was before.

I agree, it was odd that the changes got no coverage, because they were enormous changes.

However, the ACT cannot be easier on the low end for everyone - if it is easier on the low end for some low end students, it must be harder at the low end for some low end students.


I did mention that. Slow--really REALLY slow readers, out of synch with their other abilities---will do poorly on the ACT. Moreover, many kids don't understand how it works when they take the test.

The problem with the ACT is that it has ruthless time requirements.

here's an essay I wrote on the differences between the SAT and ACT, if you're interested: http://educationrealist.wordpress.com/2012/06/22/whats-the-difference-between-the-sat-and-the-act/

Back in the '90s elite schools routinely let in kids with stratospheric SAT scores and middling GPAs.

This is *exactly* what has changed, and yes, it has changed that much. I suspect, but can't prove that universities began demanding high GPAs as a way of skating affirmative action bans, and also because of the rise of AP.

In order to have standardized testing, you'd have to converge towards a single, standard test.

No, we don't. College admissions have two such tests, the SAT and the ACT. Private school admissions have five: ISEE, SSAT, HSPT, and arghh I can't remember the other two.

Once they are normed, you can compare them by percentile standing.

What I would look for in a more concrete test is one that had less of a range because the lower end was performing better relative to a more abstract test. I'm not sure it's possible. I'd just be interested in seeing it tried.

What pct of white students could do this?

White kids who were taking APUSH and almost certainly going to get a 5 on the test? Almost all of them.

And that's not a rote fact; it's a very helpful way to hook the many aspects of American history together for analysis, when you're taking a survey course. Certainly, it's an essential big picture concept fact to know that the quadratic formula is the general version of completing the square.

And if the tests overstate Asian ability and understate black ability, doesn't this suggest that the bright blacks kids are more intelligent or have higher ability than the Asian kids?

I think someone answered this later down, but to restate: Given four kids with, say, 2100 SAT scores and passing the AP US History test with a 4 or 5, the Asian kid would appear to be far less knowledgeable than the black, Hispanic, or white kid with similar scores. There just aren't many black or Hispanic kids in that range.

Also, while I suppose I'm saying that the tests "overstate ability", I am definitely not arguing that Asians are actually stupid but just get good test scores. The kids I work with are very bright. They are just entirely focused on getting good grades, and consider the source material completely irrelevant.

Anonymous said...

There's a significant difference between being smart and being bright. I think the education system does a decent job at identifying smarts but actually filters out brightness to some extent.

All the highest GPA kids I knew in high school (top public, class of 2005) were plenty smart, but they were dull. They had little creativity or original thinking. They never asked why or cared about much of anything besides knowing the exact requirements to please the instructor.

Most of the kids I'm speaking about went on to elite schools. Interestingly while many of them were conservative in high school, they are, to a person, full bird liberals now.

IMO The education system has been re-normed to better accommodate Jews, who "come of age" at age 13, while basically screwing over (I think intentionally but it could just be a coincidence) whites, especially white males, who do not mature until later.

The best advice I can give to parents with boys in high school is the Gap Year. DO NOT send your son to college right out of high school.

Anonymous said...

"someday come up with a new test that cuts down on Asian advantage over whites"

I thought that was the fire department test? But seriously, looking back on life the smartest people were the ones that got jobs for the government. Think about all the smart white people in Rochester that devoted their lives to Kodak(*), almost all of them would have been better of as postal workers. That might be one of the problems, white people basically know that math is unlikely to work out career wise. New Asian immigrants still believe.

* By Kodak I mean Kodak Photography, the chemicals part of the company was split off just in time not to pay into the pensions scheme. If you want to understand the difference between Asian and American career prospects in STEM compare Fujifilm with Kodak.

Anonymous said...

I teach at a competitive Univ of California campus, and about half my students are "Asian" (which usually means Asian-American.)

I would not confirm Realist's claims that young Asians are any less intellectual or interested or creative than the Whites. Maybe 15 years ago it seemed that way, when a greater proportion were foreign-born, but I don't see it now. At least half the interesting conversations that are struck up with me after a lecture (the ones that do not relate to what is on the exam) are struck up by Asians.

I am also very unimpressed with the supposedly great quirky creative intellects of the white kids, brought out by their adoring and nonabusive parents. Some white kids are great, but a lot are sort of smarmy pre-business types ("I will get ahead because I am so darned likable.") and/or self-focused Montessori school alums. "My parents told me I am the most amazing young person on the planet, so it is a privilege for you to have me in your class--I will be doing very unique projects reflecting my own very unique perspective." It all reminds me of the book Generation Me.

There is some of that with the Asians nowadays, but it is most common with the White kids.

Anonymous said...

it sounds like whites and asians are using different criteria to decide what is and isn't worth remembering.

asians cram their heads with formulae and factoids that score them scholarships, while whites retain trivia they can exchange for kudos when Jeopardy comes on. it all about who's worth dazzling.

i think the academic decline of jews is a leading indicator that the Academy is in trouble.

Anonymous said...

Bad analyses. Asians do not underperform relative to their sat scores- blacks do. Memory retention - south asians do fine on spelling bee stuff and junior jeopardy. Lsat - differential selection within the asian and white pools. Smart whites do not do stem as disproportionately as smart asians.

Dan Kurt said...

Educational realists would solve their problems with current admissions schemes by two moves:

1) use IQ testing and

2) bring back streaming in primary and secondary schooling.

Probably impossible.

Dan Kurt

Laura said...

Give me a break, making your kid practice the piano is not child abuse. And it's not like there's some Therapy Fairy that comes and snatches up the children of white mothers when they're not looking. If mom isn't interested in paying the therapist's bills, the therapist is not going to bother her. I suppose it is theoretically possible that an over zealous guidance counselor could threaten to call Child Services, but it's a one in a million shot that any particular parent is going to run into that guidance counselor with nothing better to do than harrass the mother of a well adjusted, high achieving student. I really don't think that possibility is skewing the white parenting distribution downward.

Education Realist, have you ever actually seen a white student who was getting good grades until the guidance counselor starting threatening her mother, and then suddenly the student started getting mediocre grades?

TontoBubbaGoldstein said...

If I were hiring a financial analyst or a marketing manager I would probably hire an Asian grind rather than a brilliant spoiled white kid.

You might be better off hiring both--as a team. The grinder will give you consistency and reliability that the high IQ guy won't. The high IQ guy will bring innovation etc that the grinder won't.
Synergy, baybee!

Anonymous said...

in seven years and probably 300 kids I have never once had a student who could explain the derivation of the quadratic formula (that is, the generalized case for completing the square)... These are kids attending public schools with some of the highest SAT averages in the country, more than a few of them topping out at 2400.

That's absolutely astounding for me.

Wow.

So much for the SAT as a substitute for an IQ test.

We're talking Pavlovian Dawg territory here.

Wow - again - wow!

Florida resident said...

Here is Charles Murray's reaction to the article by Unz:
http://www.aei-ideas.org/2012/12/at-the-ivies-asians-are-the-new-jews/

Anonymous said...

"Black test performance may understate their abilities because the tests focus too much on abstraction and generalized situations."

It depends on how you define "abilities". If by abilities you mean "g" then forget it - intelligence is all about abstraction and the ability to generalize. The only way to increase black relative performance is to test for something other than g. People have been trying for almost 100 years now and it always comes out the same, so to keep trying meets the definition of insanity. However, if we are willing to play Alice in Wonderland and define "intelligence" as "sprinting ability", then sure, no problem.



K

Anonymous said...

If you want your kids to get an education, send them to a NESCAC schools. That's where most of the old Ivy crowd migrated to.

Actually it isn't where the "old Ivy crowd migrated to." The NESCAC schools aren't much different than other schools. They're just smaller and have fewer or no grad depts. and no science research.

Josh Yellowfever said...

The article from "Education Realist" reads like a sour grape. In short, it fits very well with the stereotype of a typical white guy. Please check "the stuff white people like". There is a way to make white and black test scores better: Combining SAT score with maximum number of push-up achieved in single try, and femural bone density test. If you have each category of these three tests equally weighted, you will dramatically inprove white, black and hispanic scores, and level the playing field with Asian Tiger Mom.

Josh Yellowfever said...

The examples produced from Education realist seems quite irrelevant. What is the point to retain a long term memory of US presidents in order? The history course is so selective to favor whites, there is not much incentive for Asian to retain it. The problem with white kids in math is their lag in problem solving capability, not memorization. It, in fact, starts with teacher such as Education realist himself. Most of the white math teachers I met are quite bookish and formula based, and not very problem solving oriented, let alone being creative in problem solving, in comparison with Asian math teachers (not even close). The more I read his posts, the more I feel Education realist a hidden anti-Asian racist, despite his claims of otherwise. I have to point out those sat scores listed by Education realist are likely the scores on tests without any review.

Anthony said...

Peter the Shark says: "Back in the '90s elite schools routinely let in kids with stratospheric SAT scores and middling GPAs. One of my brother's friends (a white Lutheran) got into an Ivy League school on those credentials (he was a decent athlete too, to be fair, but not a superstar).

Now here's the twist - my brother's friend rode his high IQ and test scores into a very good grad school, but has continued to underachieve professionally in corporate life ever since. He didn't like hard work in high school, and he never learned to as an adult."

Brother's friend is like me. I didn't have to work in high school to get pretty good grades (A in anything scored mainly on tests, B in things where I actually had to do homework at home.) I got into U.C. Berkeley with a 3.5ish GPA, and got clobbered because I actually needed to work to get decent grades.

The problem is that there's no good way to test for conscientiousness. GPA approximates it, but you don't know how much of a grade is based on tests vs work completed, and you can't really compare grades across schools (or even across majors within schools).

This may be one reason for people preferring to hire graduates from their alma mater - I know what someone with a 3.5 from Berkeley had to do, but I don't know how easy it would be to get that 3.5 from Stanford.

I suppose the best measure of conscientiousness would be to compare an individual's college grades to their predicted grades based on their SAT score. But the data to do that on a college-by-college basis isn't publicly available.

Anonymous said...

Grad school isma better indicator than HS and undergrad, but Nobel prizes and startup companies are where you _really_ see cleverness and creativity. How do things look in those departments?

Anonymous said...

@Josh (10:24am): I think Ivy League quarterbacks would do pretty well by your combined index. They also do pretty well in the real world, so maybe you are onto something...

Hacienda said...

A lot of Asian students don't do test prep, go to after schools, or have Tiger moms.

Probably most, I would think. I guess that tosses a monkey wrench in here. So solly.



Anonymous said...

the overrepresentation of Asians is explained more by their dominance in GPA, as opposed to test scores.

I don't think he understands the implications of his own claims or what representation means.

His first clause - "the overrepresentation of Asians" - refers to Asian overrepresentation relative to their proportion of their population.

"Dominance in GPA" doesn't mean overrepresentation relative to GPA. Overrepresentation relative to GPA would mean that for a given level of GPA, Asians would have a higher likelihood of admittance relative to non-Asians with the same level of GPA. It doesn't mean that there are more Asians at a given level of GPA. For all we know, Asians may be underrepresented relative to GPA.

Asians are underrepresented relative to test scores. Which means that for a given score, there is a lower likelihood of an Asian with that score being admitted relative to other populations.

Anonymous said...

They're just smaller and have fewer or no grad depts. and no science research.

Sounds like an environment where teaching and learning actually happens...

jody said...

"The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on."

we have this already, it's called real life.

Truth said...

" (Austin is only 6% African American anyway.)."

We have a word for that around these parts, "black."

Education Realist said...

Asians are underrepresented relative to test scores. Which means that for a given score, there is a lower likelihood of an Asian with that score being admitted relative to other populations.

They are overrepresented in universities, particularly public universities, because of their dominance in GPA, not because of their test scores. I agree that they are underrepresented in elite private universities relatives to their test scores; that is well documented.

A lot of Asian students don't do test prep, go to after schools, or have Tiger moms.

A lot of Asians don't have really good test scores, either. A lot of Asians aren't at HYP. Lots of things that lots of Asians aren't.

Most of the white math teachers I met are quite bookish and formula based, and not very problem solving oriented, let alone being creative in problem solving, in comparison with Asian math teachers (not even close).

The example I give is exactly the opposite of a formulaic math problem. As for presidents, while I agree that presidents are not something that the general population should know off the top of their head, it is something that any decent AP US History student realizes is important, interesting, and relevant information during a survey course.

intelligence is all about abstraction and the ability to generalize.

I said abilities, not intelligence. Do we need to have an IQ of 100 in order to perform certain tasks, or is there a way to capture ability, once learned, that doesn't have to do with generalized situations?

It's a bit interesting that everyone's focusing on the Asian part, although it makes sense given the population here.

But it's worth noting that my proposed system would do relatively little to end Asian over-representation in universities, and indeed, I didn't not propose it with that purpose.

What I want is

a) end using grades as a metric. This will hurt Asians, but not terribly. It will help whites, as a rule

b) end the use of public universities as dumping grounds for near illiterates, as well as forbid them from using foreign students as piggy banks.

c) force schools to release data so their skews can be identified.

d) change the testing requirements to add essays.

I don't think most people understand the degree to which many Asians work and prepare for a test. Our current system isn't really designed for it. Rather than change the system, I'd rather add in a type of test that's more difficult to prepare for, and give other students an incentive to do more test prep.



Anonymous said...

Or swpls could simply start voting to restrict immigration, the where you were born test does tend to discriminate against non-whites.

Anonymous said...

FAO Anthony:

I would be confident in the opinion that getting that 3.5 from Berkeley or any good state school in the top 50 would be tougher than Stanford or H or Y.

Anonymous said...

"Good essay tests would do a lot to even the playing field between whites and Asians. Check out the AP score distributions for English Lit/Lang & Comp/US History compared to math and science scores. And writing an essay about math and science? Worst of both worlds."

Yes, but why is the number of Asian National Merit semifinalists so disproportionate? This test should be biased in favor of whites since the selection index double weights the verbal score.

Whiskey said...

Cui Bono?

There will be NO ZERO ZILCH NADA NONE reform of test scores, admissions process, and the pattern of Ivy League admissions (or other universities down the line as commenter Sunbeam notes) because it benefits the Elite as they now stand.

What is a major problem for most to conceptualize is the elite is not exactly like the English Landed aristocracy. The Elite is quite willing to take on the son of a Nigerian transport minister at an elite level, who then creates massive trading losses (at UBS) or the Red Princes and Princesses of China or say, Barack Hussein Obama. But it won't take on either Jim Bob Rural from Nebraska or Menachem Orthodox from Blue Collar Staten Island. It certainly will take a Buffet or say, a Speilberg.

Fair and open selection of talented people by Ivy League and other universities, and employers, is a DAGGER TO THE HEART of the GLOBAL ARISTOCRACY. And those who hope to join it, principally White women who cannot get enough of kings, princes, and global celebrities (and fantasize Lena Dunham style about sex with them).

That's why Korean Rapper Psy faces no real backlash for rapping about killing Americans and US military -- ordinary people ARE the enemy.

What is needed is to "burn it all down." Destroy totally the University system, by first and foremost flooding Harvard with as many Black people as possible. Make Harvard admit say, 85% of its student body as Blacks, and people with merit and skill will flock to online certification via places like Stanford, MIT, Caltech, and possibly Oxford, Cambridge, Humboldt, Milan, and other old European universities.

Husker Don't said...

This isn't my original line, but I can't remember where I heard it so I can't give credit: Harvard is a hedge fund with a side business in education. Jews may be over-represented at Harvard, but they're also over-represented in the United States' business elite, which reflects well on Harvard and provides them with massive amounts of money they can then donate to Harvard. Asians are also over-represented at Harvard, but that hasn't translated to nearly as much success after school. Harvard has definitely noticed this, and is redeploying capital to better uses.

Harvard, and similar schools, are not in the business of admitting the smartest people they can find. They're in the business of admitting the people who are most likely to benefit Harvard in the future. And the way they figure that out might, just might, differ group by group.

Anonymous said...

They are overrepresented in universities, particularly public universities, because of their dominance in GPA, not because of their test scores. I agree that they are underrepresented in elite private universities relatives to their test scores; that is well documented.

"Overrepresented in universities" means that the Asian proportion of university student bodies is higher than the Asian proportion of the general population.

Asians are underrepresented relative to their test scores. This means that the current Asian proportion of university student bodies is lower than what it would be if Asians weren't underrepresented relative to their test scores.

I don't know if there have been studies on representation relative to GPA. It may for all we know be the case that Asians are underrepresented relative to GPA as well.

Anonymous said...

a) end using grades as a metric. This will hurt Asians, but not terribly. It will help whites, as a rule

Why would ending grades and going exclusively by tests hurt Asians rather than help them and hurt blacks and Hispanics?

Cornelius said...

The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on.

In theory, blacks and whites could wind up on the same side on testing reform. Some clever Park Slope SWPL might someday come up with a new test that cuts down on Asian advantage over whites and then push it through because blacks do relatively better on it than on old tests.


An exam that somehow tests the same skills required for public debate would satisfy this criteria.

Anonymous said...

i"n seven years and probably 300 kids I have never once had a student who could explain the derivation of the quadratic formula (that is, the generalized case for completing the square).."

Sounds very implausible. If those kids had an average iq of 105 then you're essentially saying that iq 146 kids (1/300) did not nail the question. This is assuming the population was symmetric in distribution, whereas there surely must have been a lower iq threshold of say 95. Even otherwise, a 1/1000 rarity among whites and kids couldn't do that problem? Hmmmm ...

Anonymous said...

The math taught in high school tends to actually be applied math, not pure math. It's not surprising that most students can't derive the quadratic equation.

Anonymous said...

"Harvard, and similar schools, are not in the business of admitting the smartest people they can find. They're in the business of admitting the people who are most likely to benefit Harvard in the future. And the way they figure that out might, just might, differ group by group."

And this is why Gordon Gee persuaded Vanderbilt to discriminate in favor of Jewish students, tripling their numbers in the space of an Olympiad.

Jews now qualify as legacy admits, as special cases (sports, music, etc.), and as the children of major donors. Asian-Americans have only their high grades and test scores.

Dan Kurt said...

re" Anonymous said...
i"n seven years and probably 300 kids I have never once had a student who could explain the derivation of the quadratic formula (that is, the generalized case for completing the square).."

Sounds very implausible. If those kids had an average iq of 105 then you're essentially saying that iq 146 kids (1/300) did not nail the question. This is assuming the population was symmetric in distribution, whereas there surely must have been a lower iq threshold of say 95. Even otherwise, a 1/1000 rarity among whites and kids couldn't do that problem? Hmmmm ..."12/11/12 4:08 PM
and
“Anonymous said...
The math taught in high school tends to actually be applied math, not pure math. It's not surprising that most students can't derive the quadratic equation."12/11/12 4:47 PM

If these kids actually were incapable of understanding the quadratic derivation they had to have been 1) robbed of a primary and secondary education, or 2) dumb as posts. Take your pick.

Graduated from HS in 1959. Anyone who got through Algebra II HAD TO UNDERSTAND how to complete the square to get through the first semester. It is not applied math but elementary algebra. It is not that hard. Go to the Kahn Academy and watch a tutorial on the topic or read this article:
http://www.basic-mathematics.com/proof-of-the-quadratic-formula.html

Applied math ( which existed only on the graduate school level ) centered on matrix algebra used to solve simultaneous differential equations when I went through the system in the 1960s. Applied math was taught to engineers and scientists so that one could solve problems using math. Applied math gives one intellectual Tools to do the work an engineer and scientist needs to be a success in his career.

Dan Kurt

sunbeam said...

I'm struck by something that is kind of weird if you think about it.

Intelligence should be valuable right?

But let's take a super smart kid who just graduated high school. He reads and studies on his own.

What good is it? I'm sure he will do well, but unless something statistically unlikely happens he isn't going to invent the better mousetrap, or write the screenplay for a mega-blockbuster.

If you don't go the college route, your only real route to wealth is to have outstanding people skills. Being the best machinist in the state of New York might be good for a little over 100k, maybe.

You could have made a lot more money running a team of Mexicans roofing houses when the boom was going.

My belief is that if you don't get your ticket punched at the right school, brains are not as useful as you might think. They might be useful for someone else, but it's not your ticket to ride.

On the other hand reading and manipulating people takes real intelligence, and it is the only way most people can get ahead. But I'm not sure that getting A's in AP Calculus in High School really helps you.

Truth said...

"Make Harvard admit say, 85% of its student body as Blacks..."

Hey Sviggey, you've made this exact "point" numerous times. You can no longer deny that you and Whiskey are kindred spirits!

Anonymous said...

"Ivy league universities have a special mandate: Produce profitable leaders. That is why their admissions criteria aren't purely based on scores and GPA's."

Yes, they've done a good job of leading this country down the drain.

Eric said...

It does seem that the elite school alums are dining out for an awfully long time on a good score on a standardized test taken when they were 17.

Why might this pattern of recruiting persist?


Griggs v Duke Power. That's the bad meat that poisons the well, and until the courts decide companies can give IQ tests without getting sued employers will continue to use college admissions as a proxy for intelligence.

Cathar said...

If you want a test that both whites and blacks will do better than Asians on, why not test football skills, pct. body fat, size of one's John P. Wadsworth, and stick in a grandfather clause?

Or just admit that if its unacceptable favoritism for blacks to get AA over whites because of poorer performance , then on principle it should also be to cook up tests to benefit whites over Asians because Asians do better. The other option is to say f*ck principles, I do it because I want whites to do better.

ben tillman said...

"Black test performance may understate their abilities because the tests focus too much on abstraction and generalized situations."

It depends on how you define "abilities". If by abilities you mean "g" then forget it....


Right. La Griffe covered this years ago, at least with respect to Blacks whose SAT scores are on the right tail of the bell curve. It's sampling error where, typically, n=1.

ben tillman said...

Why would ending grades and going exclusively by tests hurt Asians rather than help them and hurt blacks and Hispanics?

Because grades are largely based on homework, and Asian parents do their children's homework for them.

Anonymous said...

Because grades are largely based on homework, and Asian parents do their children's homework for them.

Yes, because their Engrish is that much better.

Let's give credit, where credit is due.

Cail Corishev said...

In my first- and second-hand experience, smart white kids have a lot of intellectual interests outside school in which they invest a lot of time and energy. I spent many hours learning to program computers, reading all sorts of books, designing adventures for role-playing games, building things, learning to grow things, and so on. Those activities all made me more knowledgeable and well-rounded (the elusive quality schools supposedly want), but not in any way you could easily measure. Since smart kids generally aren't "joiners," they may not even do those things through a group like 4-H or Scouts that can be put on an application.

I didn't know any smart Asian kids growing up, but my impression from reading about them is that they don't tend to have as wide a range of interests. They're more likely to focus on the areas of knowledge that will directly help them get a degree or a job, and disdain knowledge that isn't obviously useful.

So if you have 100 white kids and 100 Asian kids, all with IQs of 120, the Asian kids will probably have a better average GPA and more AP classes under their belts. The white kids will probably be more interesting to hang around with, and will have a better chance of finding alternate employment and having a good life if the target goals their parents set when they were toddlers don't work out.

Anonymous said...

The larger problems seems to be the numbers of people gaming the admissions system. The metrics used to determine admission were supposed to measure some underlying qualities, but they've become ends in themselves. Worse, the gamers tend to have all sorts of negative qualities; they're a bunch of group-think conformists who are all asking "will this be on the test?"

One of the great things about the US is that you can be a complete screw-up at 17, obsessed with choom and chicks and Chevys, and still get your shit together at 23. This ethos seems to be lacking at the elite level, though. The pipeline for the elite runs through elite schools, and that's gated to select the gamers described above.

Anonymous said...

"In my first- and second-hand experience, smart white kids have a lot of intellectual interests outside school in which they invest a lot of time and energy."

Then, how come so many white college kids all think, party, and talk alike? It's PC and pussy.

I don't see much of intellectual interest in anything by anyone. Among conservatives, it's always been the case, and I think the problem got worse because many of the most interesting thinkers on the right said stuff about race, sex, and etc. But since those became taboo subjects, rightist thinking got dumb with MLK worship and free markets as solution for all problems.
I mean does anyone really think 'school choice' will fix problems? And do white people really want black kids to CHOOSE to come to their schools?

Cail Corishev said...

"The future of testing will arrive when somebody comes up with a test that Asians can't crush whites on."

Wouldn't straight IQ tests come the closest on this? Assuming it's true that Asians average a few points higher than whites, that's hardly "crushing" them, and at least you can't inflate your score by cramming. Also, since we're talking about the right end of the bell curve, the shape matters. The Asian average may be a little higher, but if the white curve is wider, whites could regain the advantage at the higher levels. Has anyone actually looked at that?

This is why I will homeschool.

That's an interesting point. Asian parents have basically determined that they'll do whatever it takes to get their kids the magical scores that colleges want: cram school, tutors, hiring out the homework, whatever. What happens if white parents decide the same thing, but instead of manipulating things within the school system, they just homeschool and hand their kids straight A's?

If the colleges simply don't want whites, it's easy: don't accept homeschool grades. But they already know from the last couple of decades of performance that homeschool kids tend to be pretty sharp and good students. If they do want the good ones, they're going to have to depend on the tests to sort out the real gems from the fake.

Cail Corishev said...

"In my first- and second-hand experience, smart white kids ..."

"Then, how come so many white college kids..."


There's a very important word in my sentence that's missing from your sentence. A hint: "college" doesn't mean the same thing at all.

Anonymous said...

"Asian parenting techniques are abusive in white people world. "

today's white people world?


"Education Realist forgets what the tests are for. They are for predicting future performance.'
"And current tests predict with high accuracy future performance. For all races. "

do they? I recall hearing complaints on how they didn't, and for sexes as well. My explanation was perhaps even with recentering, SATs couldn't keep up with the grade inflation?

"and Asian doctors would kill their patients... "

for a moment thought it said 'parents', amy chuas beware!


"This seems to be much more common in both Jewish and Asian cultures - high parental investment."

many white boys would have at least have the ambition to better their old man, if there was an old man around.

The Wobbly Guy said...

A test that Asians can't crush whites on? I think it's quite easy. When devising a subject test (eg. Chemistry), contain the following types of questions:

1. Some knowledge questions. No more than 40% of the score. Memorise, ho hum types. Easy as heck.

2. Some application/data response questions. Again, no more than 40% of the score. Requires some brainpower, but set at a level that the slightly above average student can work through.

3. Have 20% of the score based on questions that teach a concept right on the spot, and ask the candidate to solve them using the concept outlined. It's great fun for the intellectually curious, but not so fun for the grinders, who might do well for the other question types but fall apart here. It's the content version of the IQ test, assessing the student's ability to identify and apply the patterns taught. In essence, the raw core of intelligence.

To avoid disadvantaging the non-grinders (eg. whites), you can make questions belonging to type 1 and 2 really damn easy, or even make it pseudo-open book (you can bring in an A5 sized card, for example) that enables good students to be able to go well even if they didn't really prep hard enough for it.

Then toss in type 3 questions to sieve out the really brilliant who deserve to get into the elite ranks. Works every time, and almost impossible to prep for, because the type 3 question content is often university level and there's just too much content at that level to grind through.

BTW, I believe virtually every A level student (ages 16-18) in Singapore can explain the derivation of the quadratic formula - that's 30% of the cohort. Either the teaching is just too lousy, or the foundation just isn't there.

And they have to be able to explain it, because maths at their level is quite horrible.
http://www.seab.gov.sg/aLevel/2013Syllabus/9740_2013.pdf