[ricardoplouis]

Slight anecdote on test scores. In H.S. a friend and I both scored the same on a college admissions exam. A few months later after studying my score went up a couple points while his jumped dramatically to the highest percentile. I asked him how he managed to make such a jump and he told me he took prep classes specifically designed to help him score higher. I looked it up and the prep classes cost was hundreds of dollars per course.

I bring this up to say that I see many arguments here that diversity means lower test scores. But test scores are a mechanism that can be gamed with wealth. It's nice to think you scored what you did because of how smart you are, but if it's possible for someone to become dramatically "smarter" in a short period of time, then I think it speaks to the invalidity of the system.

[rayiner]

Studies show test prep makes very little difference: https://slate.com/technology/2019/04/sat-prep-courses-do-the...

Also, “hundreds of dollars” as a one time expense per kid isn’t unaffordable even for poor people. The selective admissions schools in NYC are full of poor Asian immigrants: https://nypost.com/2014/07/19/why-nycs-push-to-change-school....

[tptacek]

The point of that article, and of the first study it cites, is that the effect of test prep is radically reduced when you control for the other confounding factors. That's an important concern if you're a wealthy parent wondering whether to sink extra money into test prep, but it doesn't answer the equity concern about whether unprivileged kids are at material structural disadvantage in this testing. Those confounding factors, whose controlling in the study reduces the benefit, are likely all SES factors.

By way of example, simply taking the SAT twice has an undisputed significant beneficial effect --- potentially over 100 points on the verbal? I made both my kids take both the ACT and SAT, and take them both twice. An underprivileged kid might only be taking the one test their school asks them to take.

[rayiner]

The average score improvement from retakes seems to be modest: https://www.manhattanreview.com/sat-retaking/

As far as I can tell, most American kids don’t prepare for the SAT even with free or cheap methods. The cost of test prep seems like a red herring.

[tptacek]

I probably agree about test prep itself. But retaking the test, and taking both the ACT and the SAT (unless you know exactly where you're applying and hit your mark with one of the tests, which is something sophisticated families do but not everyone), is totally normal in high-SES households, and not normal in low-SES households. And that's just one SES-based detail; there are others (again: everything else they had to control for).

And, I mean, this kind of has to be true. Otherwise, you're ultimately going to get checkmated into arguing that low-SES working class folks are just genetically predisposed to do poorly on college entrance exams, which nobody believes.

Obviously, degree attainment tracks SES for other reasons: low SES prospective students have to be in a position to forego 4 years of income and job experience, as well as the prospect of moving away from their families to higher-SES areas in order to take advantage of the earnings potential of those degrees. But there are stark difference just in entrance exam scores themselves between high- and low-SES students (and between Black and white students, with SES fixed).

[rayiner]

You’re using SES as a proxy for cultural factors. My uncle immigrated to Toronto and never managed to get a real job. He tutors kids in the local low income immigrant community in math. Those kids are prepping for standardized tests. By contrast, my wife’s mom lives in a lower middle class city on the Oregon coast. Most of these people have been in America for 200+ years, and they’ve been poor the whole time. They don’t study for the SAT. I’ve been utterly shocked at how blasé they are about their own economic status. The folks like my wife who are exceptional leave and go become high SES somewhere else. Many other kids are smart and just dragged down by the local culture.

I don’t think we understand structural poverty. I think in part that’s because we focus on income alone, and race, but not on factors like subculture. The secret to unlocking the mysteries of structural poverty involve figuring out why white people in Oregon or West Virginia are poor while folks like Mormons managed to build solidly middle class societies despite persecution and moving to the middle of nowhere.

[tptacek]

Um, sure, yes to all of this? I was just responding to your "test prep doesn't make a difference" comment upthread.

[VictorPath]

> Studies show test prep makes very little difference

Your cited studies show a 10-37 average increase, which means some prep improves half the preppers even more.

> Also, “hundreds of dollars” as a one time expense per kid isn’t unaffordable even for poor people.

Saying this shows how out of touch you are. Speaking of New York, the recent under-10 chess championship was won by a refugee in a homeless shelter. Yes, hundreds of dollars is out of range for his family.

> The selective admissions schools in NYC are full of poor Asian immigrants

Middle class in Asia can be poor in New York City.

[rayiner]

> Saying this shows how out of touch you are. Speaking of New York, the recent under-10 chess championship was won by a refugee in a homeless shelter. Yes, hundreds of dollars is out of range for his family.

No, it shows how out of touch you are. You can’t craft general rules for society around the unique circumstances of people who are literally at the very bottom. They need special help outside the regular system.

But for ordinary poor kids, a few hundred dollars is not out of reach. What is beyond their grasp, however, is the cultural and social capital that would be required for upward mobility in the inevitability subjective regime you’d have if you got rid of the SATs. Poor asian kids, who grow up in the bottom 1/5th of the income distribution, are more than twice as likely to end up in the top 1/5th as adults than poor white kids. That’s because the gate keeping metrics in this country are objective, and not based on social connections or cultural knowledge.

> Middle class in Asia can be poor in New York City

Yes, so what? If your family is making $35,000 household income (the poverty threshold in NYC), what does it matter whether your family was middle class back in old country? And these kids are dealing with language and cultural barriers on top of the limited financial resources.

[tptacek]

But for ordinary poor kids, a few hundred dollars is not out of reach. What is beyond their grasp, however, is the cultural and social capital that would be required for upward mobility in the inevitability subjective regime you’d have if you got rid of the SATs.

Seems like a sound point.

Poor asian kids [...] are more than twice as likely to end up in the top 1/5th [...] That’s because the gate keeping metrics in this country are objective, and not based on social connections or cultural knowledge.

Seems like a reach.

[zdw]

Similarly, on the SAT I took a prep course where I enjoyed and did all the math work, kind of slacked off on the reading, and on the official test got a 800 math, 680 reading score. I finished the math before the allotted time, by a few minutes.

I don't think this has anything about being smarter - it's all about knowing the kinds of questions and their format, which is the key benefit of the prep work - you'd get to skip the time-consuming step of trying to understand the question format (analogies, really?) and just get down to solving it, giving you more time.