Hacker News Reader

Comment by Scubabear68 | Hacker News Reader

Yep. Universities tried to track relative strength of GPA to school districts, but ultimately that is a hopeless task. Our own district has been guilty of severe grade inflation and teaching 2-3 grades behind grade level to the point where a “B” here would be a D or F in a college prep private school.

Standardized tests are imperfect, but they are not entirely flawed the way high school grades are.

This will also re level the playing field for poor families with smart kids who can’t afford all the extra-curricular wealthy families lean heavily into.

View on HN

This will also re level the playing field for poor families with smart kids who can’t afford all the extra-curricular wealthy families lean heavily into.

Is this really leveling the playing field? If a wealthy, average student can get prep work to score as well as a gifted, poor student, that doesn’t seem like it’s leveling the playing field of socioeconomic disadvantage.

Ime tuition is far more accessible than many extracurriculars, and has better feedback (assuming that standardized testing is the only metric).

Maybe not perfectly level, but more so when colleges did not look at standardized test scores.

the problem is what better idea do you have? the idea of favoring "holistic student" just selects for students who have the time, money, and parental support to juggle 2-3 extracurriculars in addition to sustained effort in classwork.

it's goodhart's law in action: any measure that becomes a target will cease to become effective. for any metric you choose, a wealthy student will be able to deploy their familial resources to fiddle the metric, because that's literally what money is - "stored power and influence". My stored influence allows a pizza to appear at my door or a lawyer to appear to defend me, and it makes a violin and a tutor appear for your kids, if that's what the metric becomes. Wealthy families will always be able to fiddle something.

The dirty truth is that a small, fixed, standardized exam is actually not a bad solution to that problem. Things that require sustained effort (grades) actually tend to bias towards wealthy families, as do extracurriculars and other "holistic metrics". Wealth buys all of those things. But getting a smart kid through a standardized test successfully actually is a much smaller, more equitable hurdle. The unpopular answer is that ultimately pretty much anyone short of the completely destitute can afford a couple hundred bucks for a month of test prep (and the book alone gets you most of the way there), and in the cases where that's not true it certainly isn't improved in those cases by requiring 6 years of violin lessons and 2 instruments etc.

I mean think of it like a job interview... is it really fair to ask the applicant to do a big take-home assignment that takes a couple nights to complete? Now imagine that you have to start preparing in middle school, and your parents have to provide financial support while you do it. We all have this intuitive sense as developers that a short, deep interview is probably better, and that a long convoluted process is neither fair to the applicant nor particularly useful due to false negatives/etc. And in fact for many positions there is likely a fairly low "good-enough" bar where anyone reasonably competent is probably gonna be fine even if they're not hyperspecialized in the exact thing you're looking for.

Honest question, if you are looking at this in the sense of "college as a hiring interview": if you want true equitability, isn't the best approach to "fizzbuzz and give an offer to anyone with no major red flags" (for whatever fizzbuzz/small task is appropriate for an interview question for your position)? Obviously college applications outnumber slots, but if you have 100k "qualified" applications for 25k slots, just give a slot to the top 10k and randomize acceptance for the other 15k? That kinda seems to be what people want for colleges, if you want a true "background-blind system" - cherrypick the best and then just give everyone else who has a reasonable chance of success an equal shot at admission.

Otherwise you do end up in these "well candidate X is 82.3% likely to graduate but candidate Y is 82.4% likely to graduate" scenarios and effectively you are making decisions inside the margin of error. And that is the point where your parents' money comes into play - even a small edge helps you in a decision that is made on these super marginal factors.

I think that's the thing people are forgetting - the whole point of college admissions is NOT to discriminate, or even find the smartest; it's to attempt to make sure that those admitted can complete the coursework and benefit from it.

Once that bar has been cleared then you can start trying to improve things above it, but a problem has been that removing the bar meant you had many people who failed out (or, worse, you watered down the degree such that graduation becomes meaningless).

At the root of all this is the "everyone has to go to college to be happy and successful" - as long as we hold that AND there are people not suited to college as it currently is, we have a major problem. We need to make sure we have a society where non-college educated people can survive and thrive, or we're going to turn college into high school.

It's not leveling the playing field among social classes, but it is leaving an official door open for individuals who have a knack for tests but lack the support resources of richer folk. That's better then a closed door, and is easier to audit and evaluate than informal "holistic" admissions that may carry all sorts of implicit and hard to detect biases of their own. I'm certainly glad it was a door open to me some decades ago.

Truly leveling the playing field of society is probably not a thing that can happen.

I agree that it’s better than a closed door. But I don’t think the argument is between binary options.

>easier to audit and evaluate

This is where I think a lot of approaches go wrong. Because it’s easier doesn’t mean it’s the best or most meritocratic. At a certain point, it seems like we start optimizing to make HR personnel’s life easier, rather than the true outcome.

Not to say there's no better alternative being used somewhere, but "holistic" admissions practices often amount to rewarding nontraditional students who prove good at mimicking at WASP class signals or who fit some patronizing, objectifying trope.

If you're aiming for meritocracy, those are not an improvement over standardized tests or prior academic acheivement (and neither are they doing a good job of culturing authentic, plurlistic diversity).

Yeah, I wasn’t making a case for the current “holistic” approach

It's a big question. You could also argue that the poor student has an unfair advantage for being born with better intelligence, and the money is helping even that out.

You could but you shouldn't.

Sure. But it would also acknowledge we’re not optimizing for intellectual ability. Now which of these do you think most of society would have a problem with:

1) College should optimize admissions for intellect.

2) College ships optimize admissions for familial wealth.

You could also frame it as a dichotomy between rich kids money spent to either:

1. provide tutoring so they can go to university 2. Do something frivolous like learning to sail.

But that implies the rich kid has the choice, no? That, in itself, is part of the disparity. This is such a weird, odd take to defend.

You run into diminishing returns on test prep pretty fast. You obviously could blow thousands of dollars on a private test prep tutor but you’d be wasting your money compared to the poor kid with a $40 book grinding away at practice problems every night.

I've worked as a teacher. I realized some kids really need people to help them learn. A book wont do them any good until they have figured out how to learn from books. Having a personal tutor if struggling is really helpful.

There need to be this "click" moment.

Learning styles and all of that associated nonsense have been definitely disproven, please stop perpetuating these harmful ideas.

Does prep work have that big of an effect? I don’t think it can make an average student beat a gifted one unless the gifted one completely phones it in.

It's not perfect but IMO it's the best acceptance criteria available. The alternative is evaluating people based on subjective criteria like "well roundedness".

I think this is a false dichotomy. For example, we could also fund test prep programs for socioeconomically disadvantaged individuals.

It is graded on a curve. Raising the student body score on a synthetic measure is pointless. (Edit: You could argue "exam technique" is a good skill though, maybe?)

But ye, since rich families buy prep courses it becomes an arms race, and poor kids need them too.

The cost of test prep is a red herring. A test prep course is a couple of thousand dollars, a one-time expense per kid. From K-12, the school system will spend $150,000 or more on the same kid. If a $1,500 test pre course can teach anything a decade of schooling that costs the public $15,000-$25,000 a year can’t teach, then those test prep companies are working magic. And if $1,500 of test prep was really the difference, then it would be a no brainer to spend that money to level the playing field.

In reality, the lack of level playing field comes from cultural capital. Americans see standardized testing as beneath them and don’t prepare for it. When my uncle came to America, he moved with his wife and two adult kids into a tiny 2BR apartment in a shitty part of Queens. But I’m sure they scrounged together the $1,500 for my cousin to prep for her GRE. Meanwhile, despite extensive prodding, my wife couldn’t get her younger siblings to study for the SAT. Both were much better of economically than my cousin, and could definitely afford a prep course, they just didn’t do it.

It’s not clear to me what you can do to level the playing field in a way that’s actually fair. If you don’t have the cultural capital to know to study for a test, you sure as hell don’t have the cultural capital to know how to stand out in an admissions process based on soft factors. I spend $30,000 per year per kid to send my kids to a private school with old money Annapolis WASPs, precisely so they can get access to that cultural capital. It’s a terrible idea to make that sort of capital more important by reducing the emphasis on objective factors.

You don’t need a prep course to do well in the SAT. I did quite well with no work at all beyond school work, and I know kids today where it’s the same.

As others have indicated, you can just buy a prep book if you want and be done with it. The prep courses are just separating parents from their money.

You get way better result by practicing on old standardized tests before the real one.

Preparation for those is the most value you can get from your time studying.

I.e. study for the exam, not for what the SATs try to measure.

Yes. Just take lots of exams under strict time constraints, preferably under more adverse conditions (a public place with distracting chatter, a room that’s too cold, etc.).

And don’t bother with fake exams written by book publishers. They just aren’t authentic.

The beginning of the last paragraph made sense to me:

> It’s not clear to me what you can do to level the playing field in a way that’s actually fair.

But the ending confused me:

> It’s a terrible idea to make that sort of capital more important by reducing the emphasis on objective factors.

In a world with disparities of "cultural capital", not to mention $30K prep schools, which factors do you consider to be objective?

> In a world with disparities of "cultural capital", not to mention $30K prep schools, which factors do you consider to be objective?

$30k/year private school helps very little for standardized tests or Math Olympiad or things like that. In part because elite WASPs consider those things gauche. A kid from public school can be on a level playing field with my kids on the SAT by investing a couple of hundred bucks in buying old administered tests to practice on.

$30k/year private school helps massively on “soft” metrics colleges look at: essays, teacher or counselor recommendations, weird coursework, work in public service, “leadership,” entrepreneurship, etc. All the shit WASPs think make for a “well rounded” student. For example, a recent thing is trying to get your kid “published” while still in high school. How many professors does a bright kid in rural Iowa know? I know dozens.

> A kid from public school can be on a level playing field with my kids on the SAT by investing a couple of hundred bucks in buying old administered tests to practice on.

I thought you were dismissing test prep? "If a $1,500 test pre course can teach anything a decade of schooling that costs the public $15,000-$25,000 a year can’t teach, then those test prep companies are working magic."

I’m not dismissing test prep altogether. My point is that test prep isn’t the huge unfair advantage it’s made out to be. It can help you familiarize yourself with the format of the test, and learn some techniques for certain kinds of problems. But at the end of the day, what’s being tested is your vocabulary, basic high school math, logical reasoning, and reading comprehension. The prep classes don’t help you with that.

And insofar as the prep classes are helpful, you can get nearly all the benefits with a small investment. When I was prepping for the LSAT, I got a 167 (out of 180) on my first practice test. Nearly all my missed answers were on the “logic games” section. So I went on some Internet forums and found recommendations for PowerScore’s Logic Games Bible, which is $34 used on Amazon. I’ve got a relatively poor visual-spatial memory, so I learned some tricks in the book for using diagramming as a crutch for that. I ended up with a 176, entirely due to improvement in logic games.

So I can’t say prep never helps—the difference between a 167 and a 176 is the difference between a top-25 law school and a top-10 one. But the vast majority of people can do what I did to prep. You say “what about homeless kids or kids who don’t have Internet access” but at the end of the day you’re talking about a process that’s extremely accessible to the overwhelming majority of even poor kids. Which is a lot more than you can say for any of the alternative metrics for filtering applicants.

Ok, public and private school kids could get the same test prep for a relatively low cost, but does that mean public and private school kids are on a level playing field with regard to the test? In other words, do years of private prep school have no advantage with regard to performance on standardized tests?

> In other words, do years of private prep school have no advantage with regard to performance on standardized tests?

I would say that standardized tests is where years of private prep school makes the least difference. My wife and I are both public school kids, and our kids go to an expensive private school, and as far as I can tell there’s nothing magic about what’s taught or how it’s taught. Our motivation for private schooling is that they can counsel-out disruptive boys (so they can’t interfere with the learning of our orderly girl) and so our kids get used to socializing and dating within the right social class. I don’t think it’ll affect our kid’s SAT scores one bit.

> there’s nothing magic about what’s taught or how it’s taught

Why does it have to be magic? It's simply "culture", as you've said. Indeed, your reply alludes to this:

> they can counsel-out disruptive boys (so they can’t interfere with the learning of our orderly girl)

Private schools usually have more disclipline, higher expectations of students, a better learning environment, and selectivity. Public schools have to take everyone, no matter how disadvantaged, disruptive, or dumb. And whatever the cause, the stats do show that private school students score significantly higher on average on the SAT.

Of course disparities exist among schools, whether private or public. Education in the United States is extremely decentralized. I've never attended a private school, but I recall that when I switched public schools as a result of my family moving to a wealthier city, I was suddenly far behind in math and lost. Though I eventually caught up, I suspect that if I had remained in my old school district for my entire childhood, I would have been at a disadvantage, through no fault of my own.

By the way, a few months ago I read the book "Quiet Street" by Nick McDonell, which recounts his experience at the Buckley prep school in New York. You might find it interesting, and in any case the book is quite short, so you can breeze through it.

> Private schools usually have more disclipline, higher expectations of students, a better learning environment, and selectivity. Public schools have to take everyone, no matter how disadvantaged, disruptive, or dumb.

These things make for a more pleasant learning experience, and more well-connected kids, but I don’t think they confer any particular advantage on the SAT, as compared to any other academic or intellectual metric you might choose to use to select some applicants while rejecting others. And while the average, unmotivated and undirected kid might do better in a private school environment statistically, I don’t think there’s anything a motivated kid needs to do well on the SAT that isn’t available in the vast majority of public schools in the US. My wife went to high school in rural Iowa. Two neighboring towns had to share a high school, and the nearest “big city” was Sioux Falls, SD, 70 miles away. She wasn’t missing any resources needed to do well on the ACT.

In fact, the SAT is probably fairer than most of those other metrics, because it tests a pretty lowest-common denominator set of math, vocabulary, and logical reasoning skills. It doesn’t rely on books some kids might not have read or advanced math that might not be available at particular high schools.

> And whatever the cause, the stats do show that private school students score significantly higher on average on the SAT.

Because parents who can afford private school are smarter and more sophisticated than parents who can’t.

> And while the average, unmotivated and undirected kid might do better in a private school environment statistically, I don’t think there’s anything a motivated kid needs to do well on the SAT that isn’t available in the vast majority of public schools in the US.

How many kids are self-motivated? I certainly wasn't. I did the required work, the minimum expected of me, getting by on raw talent. Nobody ever pushed me, and I kind of regret that now. I didn't acquire self-discipline and motivation until my mid-twenties. My observation has been that the majority of "driven" kids are driven by somebody else.

> In fact, the SAT is probably fairer than most of those other metrics, because it tests a pretty lowest-common denominator set of math, vocabulary, and logical reasoning skills. It doesn’t rely on books some kids might not have read or advanced math that might not be available at particular high schools.

Ok, but what if colleges eliminated the "non-objective" factors and made the standardized test the sole criterion for admission? You can't just change the system without consequences. There would probably be more perfect SAT scores, because everyone would be focused on the SAT as the sole source of college admission, ignoring grades, extracurricular activities, recommendations, etc. Prestigious colleges would start to require perfect SAT scores for admission, given the pool of candidates with perfect SAT scores and the lack of other criteria; after all, with only one criterion, they couldn't justify letting in someone with a lower score than other applicants. In that scenario, it's questionable whether the SAT could even remain as "easy" as it is currently. They'd probably have to make it more difficult in order to produce a wider range of scores.

> Because parents who can afford private school are smarter and more sophisticated than parents who can’t.

How exactly would that matter? Many parents who can afford private school are also very busy, sometimes absent. We're not talking about home schoolers. Regardless of wealth, most parents seem to rely on the schools to babysit their kids.

I have been a standardized test coach and I can assure you that standardized tests are not as coachable as people think they are. Students plateau pretty quickly and you simply can't move their scores higher.

Reply on news.ycombinator.com