Gifted Issues Discussion homepage Forums General Discussion A bone-headed NY Times editorial

Also, we're talking about NYC, which is, as far as I can tell, a gordian knot.

Agree. To my mind, the part that might be devastating, if true, is "has not been shown to be a good predictor of student performance in high school". From a brief look at the complaint, it isn't clear whether the claim is that no evaluation has been done of the relationship between test scores and performance, or that they haven't given lower scorers a chance to show that they'd succeed at the school. The latter wouldn't be devastating in my view. The former would: I would expect the school to have collected statistics about the relationship between the scores *of those admitted* and their performance at school. If the test is an appropriate one, I would expect there to be a clear positive correlation at least towards the bottom of the scale of marks: those admitted with the lowest scores should tend to do a bit worse than those admitted with slightly higher scores, etc. (I put i this way because you might expect ceiling effects to obscure this at the top.) If this isn't the case, then indeed, it would seem reasonable to question the appropriateness of the entrance test.

But the interesting general issue is: this seems to basically be an achievement test, not an aptitude or IQ test. So, even if for the sake of argument we assume that aptitude is fixed at birth, better educated students are going to do better. "Better educated" is going to include going to better middle schools, having more supportive parents, and yes, having been tutored (whether "to the test" or "to enrich and improve the student's mind"). On average, that's going to lead to a skewed racial mix: in current circumstances, no way round that. People asking for a test that can't be tutored for are asking for (the impossible, and) admission by IQ test. But at some point, what makes for a coherent class that can all be taught at the same level isn't just aptitude, it's some combination of aptitude and achievement. So it wouldn't work just to replace this test by an IQ test (whose questions are miraculously kept secret so noone can prepare for them :-); that would weaken the school.

It's essentially the same problem that we have in the UK where the admission is to elite universities and the disadvantaged/advantaged groups are those who have been to state/private schools. There is pressure to "level the playing field" by discounting the better education that the private school students have (on average!) had up to that point, but this ignores the fact that it would have to cause either a dumbing-down of the universities' courses, or else mass failure of the students admitted who can't cope with them.

Wouldn't this force all affluent, upper middle class, and even middle class families to have their kids test prepped? If you're going to ding my kid's scores because you assume I must have test prepped her, I might as well test prep her to make-up for the lowering factor that was implemented.

-S.F.

I disagree, though I wasn't phrasing my ideas optimally when I wrote "can't be prepped for." I should have said, "Serious prep isn't a realistic option." As I mentioned, the Miller Analogies test and CTY's SCAT are relatively prep-resistant. This is because the test makers don't really release previous versions of the test. You can purchase access to three former Miller Analogies tests, but you can't save or print and they also don't tell you how the scoring system works. AFAIK, no SCAT tests are available. CTY just provides a page or two of practice questions. Likewise, the entrance exam for Thomas Jefferson is in this group (see this link). To the best of my knowledge, there was a pretty strong correlation between test scores and performance at Thomas Jefferson before they changed the admissions process. Today, 65% of the admissions decision is based on recommendations and writing samples. I don't know if they've studied correlations between test scores and other factors in their admissions process, but the remediation rate skyrocketed after they changed the process. But the number of minority students didn't change.

But the real problem is that focusing on the the admissions process ignores the root of the issue, which is that New York City 1) doesn't provide enough seats for gifted students and 2) NYC has some really crappy schools.

What large city with similar demographics does not? Bad students make bad schools. Public schools that are required to accept all students in a bad area and that are mandated to keep almost everyone in school will not be very good. I support charter schools and exam schools, but they operate under different rules, and there need to be schools to serve the rest of the population.

A thread "Are gifted education programs a waste of money?" http://giftedissues.davidsongifted..../all/Are_gifted_education_programs_.html discussed such research.

I see mithawk and Bostonian have got their polititroll hats on today.

At least Bostonian isn't being subtle about his social Darwinism this time. That bit about wealth being related to intelligence is HILARIOUS. It's almost as if he's never heard about gifted underachievement. If only there were a place on the internet where he could talk to experts on the subject, and individuals with real-world experience...

As for mithawk, I particularly enjoyed the bit about Stuyvesant being available to families who can't afford to live in an outrageously expensive city like, for example, New York City.

Enjoy your false reality.

From the Wikipedia article on "intelligence quotient" http://en.wikipedia.org/wiki/Intelligence_quotient :

"The American Psychological Association's 1995 report Intelligence: Knowns and Unknowns stated that IQ scores accounted for (explained variance) about quarter of the social status variance and one-sixth of the income variance. Statistical controls for parental SES eliminate about a quarter of this predictive power. Psychometric intelligence appears as only one of a great many factors that influence social outcomes."

One-sixth of income variance means that the correlation of IQ and income is about 0.4, since 0.4^2 = 0.16 which is about 1/6.
A 0.4 correlation is not negligible.

It doesn't equal 1, either.

Also, see correlation fallacy.

Finally, this: Psychometric intelligence appears as only one of a great many factors that influence social outcomes.

Why are we concerned about social/income outcomes here?

I can raise my income by 100% by moving to NYC or DC while simultaneously degrading my quality of life by 50%.