People are always putting down the social sciences, saying they aren't real sciences. When an astronomer predicts a solar eclipse, it happens. But when a social scientist predicts something, how often does it come true?

Well, it depends how unpopular the social scientist cares to be. You don't even have to be a scientist. For example, it's easy as pie to make accurate predictions about school test scores. You just have to be willing to put up with little things like being denounced as evil and getting fired. The problem with the social sciences is that there's little demand for social scientists. What there's a huge demand for is social shamans who can lift the curse of the evil eye.

For example, last year there were complaints about admission to the "gifted" programs in New York City public schools. Admissions were done in a kind of haphazard fashion with some parents better able to game the system than others. The NYT reported at the time:

[Gifted programs] have also been controversial, as other parents say they have discriminated against black and Hispanic children. School officials say they hope the new plan, coupled with an intensive outreach effort, will increase those numbers.

So, what did school superintendent Joel Klein do? He sounds like a smart guy, right? So, he should understand simple cause and effect, shouldn't he?

Klein came up with the idea of using a standardized test and accepting only students who scored at the 95th percentile nationally or higher into gifted programs. What a brilliant concept! Why didn't anyone in New York think of that before! Surely, that couldn't cause any political problems!

Half Sigma immediately blogged in 2007:

I'm trying to figure out what this is changing. Without any affirmative action, the gifted classed will be dominated by Asian kids, and blacks and Hispanics will complain that they are being discriminated against.

Last week, the NYT reported on the results of Klein's brainstorm:

The number of children entering New York City public school gifted programs dropped by half this year from last under a new policy intended to equalize access, with 28 schools lacking enough students to open planned gifted classes, and 13 others proceeding with fewer than a dozen children.

The policy, which based admission on a citywide cutoff score on two standardized tests, also failed to diversify the historically coveted classes, according to a New York Times analysis of new Education Department data. ...

The incoming gifted class is 9 percent Hispanic, 13 percent black and 28 percent Asian. Their kindergarten and first-grade peers in the city are 41 percent Hispanic, 27 percent black and 15 percent Asian. Students admitted to gifted programs under the previous policies are 15 percent Hispanic, 31 percent black and 20 percent Asian.

See, it turns out that in a lot of NYC neighborhoods, almost nobody is at the national 95th percentile or above. Who could have imagined that? Apparently nobody in Joel Klein's office ... Of course, they could have just read Half Sigma's blog, but understanding how the world works is evil, so who would do such a thing?

It would appear that whites (and miscellaneous) make up 48% of the new gifted class accepted meritocratically while making up only 17% of the total kindergarteners. Under the more haphazard old system, whites made up only 33 percent of the gifted classes.

Today, the NYT is back with a report on the results of an expensive program that's been running for 14 years to boost test scores among Non-Asian Minorities (NAMs) so they can qualify for Stuyvesant, Bronx Science, and the six other elite high schools in NYC where admission requires taking a test. (By the way, at Stuyvesant the average SAT score of seniors is supposedly 1410 out of 1600.)

You'll never ever guess what happened!

Racial Imbalance Persists at Elite Public Schools

by Javier C. Hernandez

Recent efforts to get more black and Hispanic students into New York City’s elite public high schools have fallen short, with proportionately fewer of them taking the admissions exam and even lower percentages passing it. The performance gap persists even among students involved in the city’s intensive 16-month test prep institute, designed to diversify the so-called specialized high schools, including the storied triumvirate of Stuyvesant, Bronx Science and Brooklyn Tech.

Among the 21,490 public school students who last year took the exam, the single gateway to eight high schools, 6 percent of blacks and 7 percent of Hispanics were offered admission, compared with 35 percent of Asians and 31 percent of white students. The disparities were the worst at Stuyvesant, where 2 percent of blacks, 3 percent of Hispanics, 24 percent of whites and 72 percent of Asians were accepted. (Over all, 1 in 5 test-takers is offered a spot; racial data is not available on private school students.)

The disparities were the worst at Stuyvesant because it's perhaps the most famously competitive high school in the country.

Parents of black and Hispanic students have long complained about the lack of diversity in the elite schools’ enrollment, and the Department of Education promised two years ago to study whether the demographic lopsidedness was the result of certain groups’ doing poorly on the grueling two-and-a-half-hour test, not taking the exam in high numbers, or simply choosing not to attend the schools. The city abandoned that effort, but an analysis by The New York Times shows that not only do blacks and Hispanics lag behind whites and Asians in succeeding on the exam, they are far less likely to take it.

Perhaps most surprising is a close look at the students enrolled in the city’s Specialized High Schools Institute, created 14 years ago to prepare students for high school and recently expanded by Schools Chancellor Joel I. Klein. Black and Hispanic students who attend the institute are more likely to succeed on the test. While 90 percent of Asians and 85 percent of white students at the institute take the test, 65 percent of blacks and 70 percent of Hispanics do; last year, of the institute graduates taking the test, 58 percent of the Asians, 49 percent of whites, 21 percent of Hispanics and 19 percent of blacks were offered admission.

Let's do the math. Among Asians enrolled at Specialized High Schools Institute, 52% pass the test (90% times 58%), as do 42% of whites, 15% of Hispanics, and 12% of blacks. It looks like another example of La Griffe Du Lion's Fundamental Constant of Sociology—the gap between blacks and whites in average scores on tests of reasoning is about a standard deviation.

Deputy Mayor Dennis M. Walcott said the data showed there was work to be done both to get black and Hispanic students to take the test and to help them pass it.

”I’m not ever happy when I see a low percentage of those students participating in schools that are high rigor,” he said. ”It’s important for the halls of Stuyvesant, the halls of the Bronx High School of Science, to be reflective of the city itself.”

Instead, the schools that make up the upper crust of the public education universe belie the system they are part of and the city where they reside, and the disparity between the races has grown even more pronounced over the past decade.

In this city of 1.1 million public school students, about 40 percent are Hispanic, 32 percent are black, 14 percent are Asian and 14 percent white. More than two-thirds of Stuyvesant High School’s 3,247 students are Asian (up from 48 percent in 1999). At Brooklyn Technical High School, 365 of the 4,669 students, or 8 percent, are Hispanic; at the Bronx High School of Science, there are 114 blacks, 4 percent of the 2,809-student body.

The other schools in the elite group, considered a second tier, are more diverse: Brooklyn Latin School, for example, which became a specialized high school in 2007, is 23 percent Hispanic and 32 percent black (though it has 183 students, a fraction of the top three).

The portrait of test-takers from public schools is closer to the overall enrollment, but hardly a mirror: 28 percent of last year’s were black, 23 percent Hispanic, 30 percent Asian and 19 percent white.

Marcia V. Lyles, deputy chancellor for teaching and learning, acknowledged that racial diversity at the schools ”is not where we would want it to be.”

Elizabeth Sciabarra, who oversees student enrollment planning, said the city had increased its efforts to inform families about the test, with the hope that interested students of all backgrounds might start preparing earlier. But, she noted: ”It is a choice. There are kids who might be wonderful candidates for this who will just not sit for the test. That transcends ethnicity; that’s across the board.”

The test-prep institute, which includes a full-time five-week summer session and twice-a-week workshops during the school year, was a core part of the city’s strategy to diversify the ranks of the elite schools. But the intensive program has been hampered by a Supreme Court decision last year that ordered districts to remain race-neutral in efforts to diversify schools. Now the program gives preference to students based only on family income, not race.

And enrollment in the institute has fallen to 2,800 students at 10 sites this year, from 3,800 students at 17 sites in 2006. Education officials said that they reduced the number of sites to standardize the curriculum and that despite the drop in enrollment, more students were currently receiving the full test-prep regimen.

The test itself, consisting of 45 verbal questions and 50 math questions, measuring students’ ability, for instance, to put sentences in order and discern geometrical angles, has also become a subject of criticism.

Joshua N. Feinman, an economist who graduated from Stuyvesant and is the parent of a Bronx Science junior, recently released a study challenging the validity of the test, saying it had not undergone normal predictive bias studies to see if it was skewed toward any gender or racial groups. The study revives complaints from the 1960s, when civil rights groups charged that the tests were unfair to black and Puerto Rican children and should not be the only criterion determining access to the schools.

Department of Education officials said they were confident that the test, which is manufactured by Pearson and has been used since the 1970s, was reliable.

These results are all straight out of La Griffe du Lion 101, but it constantly comes as a big surprise to people like Joel Klein.

But, of course, now that Obama is in, we have Hope and Change, so everything will be different Real Soon Now.