kitchen table math, the sequel: Hunter College High School

Friday, August 6, 2010

Hunter College High School

Diversity Debate

13 comments:

Crimson Wife said...

The best way to diversify Hunter High would be to offer better K-8 schools. I bet there are plenty of black & Hispanic kids in NYC who have the potential to make the cutoff *IF* they got a decent education. The problem is that too many of them are stuck in dreadful schools & their families lack the resources to make up for it after school.

Catherine Johnson said...

I'm at a loss on this one.

Now that I'm wading into the test prep situation out here, I'm not so sure better public schools can compete with high-end, one-on-one test prep.

I mentioned earlier that people are spending $25 to $45K for one year of SAT tutoring.

Parents are also spending large sums for private, one-one-one entrance exam prep.

The only answer that occurs to me is to toss the whole model out and adopt precision teaching for everyone.

Train everyone up to his or her max and then see where we are.

Crimson Wife said...

There will always be folks with more money than sense. I can think up a number of uses for $25k-45k that would be more likely to get a kid into an Ivy League school than spending it on SAT tutoring. Seed money to start a non-profit organization or something along those lines. Ivies are more impressed with actual achievements than with getting from 1500 on the SAT M+V to 1600.

palisadesk said...

I was surprised to see that they depend exclusively on a single test that is teacher-constructed (thus probably of limited validity in identifying intellectually gifted students). Without compromising quality they could make the selection process include a broader range of measures. They would probably get not only a more diverse student body that way, but also a more "gifted" one.

As a parallel, I thought of the Jefferson Scholarship at the University of Virginia. I never heard of it until one of my nephews was nominated for it, and he went through an extremely rigorous selection process. The scholarship is a very generous one, paying not only tuition but full living expenses and providing all kinds of enrichment opportunities, travel and so on. The resulting selection of students I'm willing to bet yields a much more talented pool than a single paper and pencil test would do.

For more information on the process, see:

explanation of Jefferson Scholarship

and

Jefferson Scholars foundation

Full disclosure: my nephew was chosen a Jefferson Scholar so that may affect my view of the validity of the process. He is a truly outstanding young man. I don't see why Hunter College High School could not introduce a variation on this theme.

Crimson Wife said...

I suspect it would be very difficult to do the kind of "holistic" evaluation that palisdesk mentioned given the large number of students seeking a place at Hunter High. If Wikipedia is to be believed it's around 29,000 kids who take the SHSAT each year.

Bostonian said...

It's the Bell Curve, darn it.

Average IQ by race:
Asians 105
Whites 100
Hispanics 90
Blacks 85

Different sources give numbers a few points higher or lower for each group, but the implications don't change drastically.

If the standard deviation of IQ is 15, do the math on what fraction of students of each race has an IQ above 115, or 130. The numbers look ugly, but ignoring them doesn't make them go away.

Kai said...

"Train everyone up to his or her max and then see where we are."

If we were to actually achieve this, our society as it is now would completely fall apart. We need to create a population where the vast majority is intellectually handicapped by its education so that we have enough bodies available for the menial work that needs to be done.

IQ is not immutable. I was able to raise my son's IQ by 40 points in 5 years by intensively homeschooling him. If we were to do essentially the same thing with every child in the country, I would be willing to bet that the average IQ scores would go up by *at least* one standard deviation in disadvantaged populations. And assuming that IQ can be used as a proxy for the richness of one's intellectual life, a gain like that across a population would be huge. And most unsettling to the order of things.

Crimson Wife said...

Richard Nisbett has a long discussion about race & IQ in his book Intelligence and How to Get It. He makes a convincing argument that the differences are mostly environmental (which can be changed) rather than genetic (which obviously can't). If it were genetic, then IQ in blacks would be positively correlated the degree of white ancestry- and it isn't.

Linda Seebach said...

Crimson Wife said, "Richard Nisbett has a long discussion about race & IQ in his book Intelligence and How to Get It. He makes a convincing argument that the differences are mostly environmental (which can be changed) rather than genetic (which obviously can't). If it were genetic, then IQ in blacks would be positively correlated the degree of white ancestry- and it isn't."

Well, it is, but never mind. The broader point is that Nisbett is a spirited defender of the politically correct view that racial disparities in IQ result primarily from differences in children's environments. The implication is that these disparities will mostly vanish if environments become more similar.

That's more wishful thinking than argument, and it's "convincing" only to those who already agree with Nisbett's thesis. But it's too vast a subject to settle here, so let me just offer a couple of links.

In a May 2009 working paper, J. Philippe Rushton and Arthur Jensen critique Nisbett's book point by point. If you agree with Nisbett, you won't agree with them, and vice versa, but if you're going to write about these things you ought to be familiar with arguments on both sides.

I should add the disclosure that I am slightly acquainted with both men, and admire them, their work, and their determination.

Further links on my blog, at
http://www.lindaseebach.net/wordpress/?p=122

Crimson Wife said...

I skimmed the linked paper, and a lot of the research cited is very dated. Because of the kind of racial discrimination that used to be very prevalent, I don't think any study done prior to about 1980 is relevant to the situation today.

Having witnessed the kind of work ethic displayed by the typical Jewish student and the typical East Asian student vs. the lack thereof displayed by certain other groups, you're going to have to provide a lot more convincing an argument than merely citing a bunch of decades-old studies.

The question in my mind is how can we get underperforming minorities to adopt the positive cultural aspects of the highest-performing ones. The KIPP schools seem to be doing a pretty good job at this. I was looking at the list of schools in my area with the best performance on the Johns Hopkins CTY talent search, and there amid all the chi-chi (predominantly white & Asian) private schools were 5 predominantly black & Latino KIPP schools.

Bostonian said...

Crimson Wife wrote, "I was looking at the list of schools in my area with the best performance on the Johns Hopkins CTY talent search, and there amid all the chi-chi (predominantly white & Asian) private schools were 5 predominantly black & Latino KIPP schools."

The so-called "top schools" are ranked by the number of students who PARTICIPATE in the CTY talent search, not on the fraction of students who EARN QUALIFYING SCORES to be eligible for CTY summer courses. Furthermore, the "The Next Generation Venture Fund" sponsored by Goldman Sachs gives scholarships for "diverse" (not white or Asian) students to participate in CTY programs. Perhaps the KIPP schools are encouraging their students to go after these scholarships.

Of course, KIPP schools should encourage their students to take advantage of available opportunities. But KIPP participation rates in CTY programs are not evidence of test score gaps being closed.

Crimson Wife said...

But to qualify to participate in the CTY talent search, a student has to have a score in the 95th percentile or better on a standardized test. So even those who do not score well enough on the SAT/ACT to take the CTY courses are still doing very well compared to the overall student population.

Certainly the traditional government-run schools in my area with a predominantly black & Latino population aren't getting a significant number of their students to score in the 95th percentile or higher on standardized tests. In my local district, 6 are on CA's list of the worst 5% of schools in the state.

palisadesk said...

I suspect it would be very difficult to do the kind of "holistic" evaluation that palisdesk mentioned given the large number of students seeking a place at Hunter High

They wouldn't need to do anything of the sort for the majority of applicants, just for finalists. Now I know little about Hunter College High School apart from what I read in the NY Times story and the links from it. I checked the website, too, and saw that indeed their applicants are admitted on the basis of the teacher-created test and a written essay, no other criteria. There's nothing wrong with that, per se, but it is a selection process to identify high-performing students, not necessarily intellectually gifted students. Teachers don't have the psychometric expertise to develop tests that will identify cognitively gifted individuals.

This would not matter, except that the school bills itself as a school for academically and intellectually gifted students. If so, one would expect them to have a process in place to identify intellectually gifted students -- minority or otherwise. Many, many students who are high achievers are not gifted, and many gifted students are not the top performers in terms of GPA or even test scores. Gifted kids are very much underserved by the public school system, and are more likely to drop out of school than their less-able peers. The NY Times article suggested that the student body of HCHS is mostly middle class, which is certainly not representative of the city as a whole. Intellectually gifted students in low-SES schools may not do nearly as well on a straight achievement test because of the poor schooling they have received, but various cognitive assessment tools could identify them.

My district's gifted program has evolved over time. At one point we were using the "Renzulli Triad" to identify gifted students: in addition to high academic ability, the student needed to show evidence of creativity (not necessarily the arts) and task commitment -- a passion for something specific, whether a cause or a field of interest. I thought this combination had some merit, rather than relying exclusively on WISC scores (which is what we do now. Students must score in the 99th percentile on both scales of the WISC. Occasionally another measure, such as the WJ-III or the Stanford Binet will be used).

One thing we consistently find is that a majority of students who are referred by teachers for the Gifted program do not meet criteria. They will be bright, but often no higher than the 85th percentile. The truly gifted students are sometimes overlooked by teachers if they don't fit the stereotype. Low SES and minority gifted students are in desperate need of an enriched academic environment where they can soar, not only because their families and schools often fail to provide the needed stimuli, but also because they need the opportunity to be with others like themselves.

Demographic data about IQ of various populations shed no light on individuals. There are gifted students in all racial groups. I've referred a number of students to the gifted program in my district, and only one was white (or East Asian). One boy I referred was not considered gifted by any of his teachers. The psychologist who tested him was very experienced and had been with the district for more than 20 years. She told me afterwards that this student was the highest-scoring child she had ever tested -- reached the ceiling on every item. He was 9 years old, ESL and a South Asian refugee.. He is thriving in the gifted program.

If the school wanted to find minority gifted students, they could certainly develop protocols to do so, but would need to look beyond mere achievement testing.