NY Times Magazine: "The New Affirmative Action" : Primrose Road

S	M	T	W	T	F	S
1	2	3	4	5	6	7
8	9	10	11	12	13	14
15	16	17	18	19	20	21
22	23	24	25	26	27	28
29	30

« Shakespeare and Derrida conference | HomePage | NY Times Magazine: "The New Affirmative Action" »

09/30/2007

NY Times Magazine: "The New Affirmative Action"

In today's New York Times Magazine (The College Issue), David Leonhardt takes into account the intersection of the categories of low-income students and minority students, and indeed points out that

"many of the beneficiaries of the preferences end up being upper-middle-class minority students, since they tend to have better test scores than poor minorities."

Even with affirmative action in place, universities tend to favor students from middle- to upper-middle-class families.

"The colleges apparently put even more stock in the polish that comes with affluence — the well-edited essay, the summer trip to Guatemala, the Arabic language lessons. In any case, the poor lose."

Then, Leonhardt alludes to what may be the most problematic aspect of college admissions: even in the era of affirmative action, they rely on outmoded forms of intelligence testing:

"Intelligence, indisputably, is in part genetic; and every intelligence test shows a gap between black Americans and others. For a long time, scientific research wasn’t very good at explaining this gap. But it has gotten better lately. For one thing, the gap between white and black adults has narrowed significantly since 1970, according to work by the noted researchers William Dickens and James Flynn. Four decades is too short a time period for the gene pool to change, but it’s not too short for environment to improve. Most intriguing, Roland Fryer and Steven D. Levitt, two economists (the latter is one of this magazine’s Freakonomics columnists), have found there to be essentially no gap between 1-year-old white and black children of the same socioeconomic status."

It seems that "socioeconomic status" is the key here: despite the many adjustments ETS has made to the SAT in the past few decades, the idea underlying the test and standardized tests like it (even if most of us no longer subscribe to this kind of thinking) is that low-income students should be kept out of the ol' Ivory Tower. Even today, students who have attended well-funded schools, can afford a $1500 test prep course, and even have a working knowledge of Latin have an automatic advantage on tests like the SAT.

Affirmative action, as it currently stands, seems to be based far more on politics and "good ideas" than it does on actual research in educational psychology. Meanwhile, the SAT and GRE are based on 1920s and 1930s testing methods designed to keep out low-income students (and there is, of course, a sizable intersection between "low-income" and "minority" in this country). In my view, an alternative to overly-politicized, not-really-student-centered affirmative action policies would be a radical rethinking of admissions testing. While SAT administrators have taken a step in the right direction by eliminating analogy questions (which often took the form of the infamous RUNNER:MARATHON::OARSMAN:REGATTA one), I believe that we would benefit from an alternative test (one that would eventually eat the SAT alive, perhaps) based on unbiased educational psychology research.

[In the interest of full disclosure, I teach test prep classes to supplement my adjunct-o-riffic income. And I've seen firsthand what a difference a few thousand dollars and a few extra hours of leisure time can make in terms of students' SAT, GRE, and LSAT success.]

11:20 Permalink | Comments (13) | Email this | Tags: education, standardized tests, college admissions

Comments

I also looked at this article. [1]

The IQ section was bizarre. The author appears to believe that if some feature does not show variation in a greatly reduced form, this means that later growth and variation is exclusively a product of the environment.

I assume that secondary sex features on newbornes vary little, if at all. Does the author then believe that the bra industry is a relic of outmoded forms of chest-size testing?

Similarly, the gap between the hight of South Koreans and Americans has narrowed significantly since 1970 (South Koreans typically no longer eat starvation rations). Does the author then believe that the difference in height between Koreans and Americans is entirely a product of culture?

[1] http://www.tdaxp.com/archive/2007/09/29/affirmative-action-bad-policy-bad-science.html

Posted by: Dan tdaxp | 09/30/2007

Dan,
While (as you probably know) I don't agree with your claims that g tells us anything at all, you do bring up a good point about the limits of "constructionism" here. If 'everything' is constructed, then isn't the idea that everything is constructed also ... constructed?

Some things (just not, in my view, intelligence) appear to be biologically-based. I am actually currently in the middle of reading a book on new media theory that discusses how constructionist theories of the "body" have actually disembodied the body, totally abstracting it from its biological basis (or, its biology).

However, I'd still argue that certain forms of testing were (socially and politically) constructed in order to justify racism (which itself was constructed centuries before to justify colonization) and classism. These arguments would, I believe, make more sense if constructionists weren't also continually arguing that gender (as opposed to sex or sexual orientation) and disease are socially constructed as well. If I said "the sky is blue and so are potatoes," some people would be inclined to disbelieve both statements; if I argue that everything is socially constructed even though everything isn't socially constructed, some people are bound to believe that all theories of social construction are to be distrusted.

Posted by: PrimroseRoad | 09/30/2007

"I don't agree with your claims that g tells us anything at all"

More precisely, it tells us a lot. It's the single most reliable predictor for a range of outcomes. There's also obviously hidden factors behind g that doubtless account for much of it.

"I'd still argue that certain forms of testing were (socially and politically) constructed in order to justify racism (which itself was constructed centuries before to justify colonization) and classism"

Well, to an extent, sure. Modern IQ testing began with the need to identify and train large numbers of civil servants and mid-level managers. To this extent, it clearly was part of the bourgeious program of overthrowing the aristocracy, furthering the market driven economy, and generally ushering in the most productive era of human history ever known (as Marx would say).

Similarly, patience was doubtless constructed in order to justify crop farming.

"These arguments would, I believe, make more sense if constructionists weren't also continually arguing that gender (as opposed to sex or sexual orientation) and disease are socially constructed as well. If I said "the sky is blue and so are potatoes," some people would be inclined to disbelieve both statements; if I argue that everything is socially constructed even though everything isn't socially constructed, some people are bound to believe that all theories of social construction are to be distrusted."

Agred.

Posted by: Dan tdaxp | 09/30/2007

I think you need to brush up on your history of standardized testing. First of all, the SAT was emphatically not "designed to keep out low-income students," quite the opposite. It was designed to open up what had been patrician world of the Ivy league by identifying capable students from the middle and lower classes and non-wasps. And it was hugely successful, as shown by the huge influx of Jews into elite institutions.

Second, the infamous "OARSMAN:REGATTA" question went away many many decades ago, and the only reason people keep trotting it out is that it was atypical and they can't find more modern example. Test designers made lots of efforts to remove these types of "cultural biases," and in fact the biggest differences between ethnic groups have generally been in the math sections that are least amenable to this type of possible cultural bias.

Posted by: ed | 10/01/2007

Hello boys ;-)
Anybody can give me url to the XRumer's homepage?
Or maybe some information...

Thank you very much!!!

P.S. Вот уроды...

Posted by: JeyramoXRu | 10/17/2008

Posted by: natalyarep11 | 05/29/2009