Colleges that ditched admission tests find it harder to fairly choose students • Hacker News Board

Re: Colleges that ditched admission tests find it harder to fairly choose students

#551

Post by bane » Sun, Oct 23, 2022, 3:22 PM UTC

BobbyJo wrote: post #534

bane wrote: post #506

Earlier quoted context omitted.

I grew up pretty poor, and through a longer story than I want to type here and a confluence of negative voices mostly from my school, was dissuaded from taking the SATs and even considering college. Years after high school I eventually found my way to community college, then to a 4-year, a grad program and so on. It worked out to be a blessing as I graduated without debt from a major university and entered into a rea…

Doing well in college is a great measure of how well you'll do in college, and subsequently, many good schools will accept mid degree transfers of undergraduate students who might not have done well in high school. I don't think it's viable to use that to replace the SAT and other current evaluation metrics, however.

I guess my point is, why not just shove all the students into a Community College a-like for the beginning of their undergrad? With the following features:

- Open enrollment - Cheap - Can drop out and retry later without any penalty - Entry into the final two years is granted based on first two year performance, state schools are guaranteed entry, and privates get to be selective based on major(s), student performance awards, and GPAs. - CC Associates degrees can be earned in periods much longer than 2 years. Allowing students to work, or retake remedial subjects, or try out multiple majors, without it being a huge deal.

The first two years of most undergraduate degrees are fairly similar. IIR there's been studies that show that students who arrive at the universities from CCs do as well as or better than their traditional peers. It helps with student cost and loan problems. Graduates who don't go on to their final 2 or have no plans to can simply treat the CC as a trade education. And many students who do flunk out of school due to partying, immaturity, etc. can simply come and try again when they are ready. Students actually get an education rather than studying for a mostly useless test.

Basically the CCs are a softer entry into higher education, and can be the test for performance in a way that actually is meaningful. Since student have to pay to be there, it's not just the same thing as high-school, and CC schedules are usually so flexible students can even enter the workforce while going to CC, or take more time to "figure out" what they want to really major in without it costing hundreds of thousands of dollars.

Also, as I've gotten older, and realized that tests like the SAT are just a product offered by a company, and not some Federal standard, I've started to view those kinds of tests as more of a racket than a real filter. People I know in higher education think the test is basically worthless, but also acknowledge that students have basically learned nothing over the pandemic, leaving them with very little to be selective for entry with.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#552

Post by melony » Sun, Oct 23, 2022, 3:30 PM UTC

crackercrews wrote: post #502

melony wrote: post #385

Earlier quoted context omitted.

Just ask anyone in academia? It barely counts as an open secret. Most admission forms (for the Cali and other state schools/private schools) would ask you explicitly if you have any relatives working for the school and their specific title and role.

Definitely not true for UC. Had a friend whose mom is a professor there. There wasn't anywhere to indicate this. Privates ask about alumni connection, and possibly about teaching connection. But not the public schools. Regardless what you've mentioned isn't proof that this is why so many get in. The fact is that kids of PhDs are likely to have genetic and educational advantages. That is likely much of the reason that…

Check Stanford and USC. It is a well known (though unofficial) policy.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#553

Post by rayiner » Sun, Oct 23, 2022, 3:45 PM UTC

onetimeusename wrote: post #248

rayiner wrote: post #100

What surprises me is how many Asians support these measures. Make no mistake: our ability to progress as we have in America has depended on the fact that objective metrics cannot easily be gamed. The cost of test prep is a minor barrier to entry compared to the cultural knowledge required to write an essay that moves a WASP admissions officer, or the social capital and connections required to get involved in charitab…

I don't think WASPs are given an edge here. I feel like it is actually racist to make this assumption also. White people as a broad category are the only racial group that is underrepresented relative to their population percentage in Ivy League schools. Maybe you could make the argument that WASPs are a subset of white people who are overrepresented specifically because of that faculty benefit for admissions but the…

There is no such thing as “white people.” WASPs are a distinct tribe and feel zero affinity to white ethnics, Appalachians, etc.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#554

Post by kenjackson » Sun, Oct 23, 2022, 4:32 PM UTC

brigandish wrote: post #470

kenjackson wrote: post #93

Earlier quoted context omitted.

This is a much longer discussion and one that tends to devolve. And there is a camp that believes that black peoples have it better than whites people today. I find it hard to believe, but they exist.

You made a statement that clearly begged the question, and now I see that you’ve made this comment twice to try and stop further discussion and examination of your claims, and yet I see a perfectly fine and interesting discussion has continued below and hasn’t devolved at all. Might be time to check more than one of your assumptions, but more importantly, stop trying to police what I and others here say. I’m an adult…

I never said not to have the discussion. I said it would devolve, and look at what we’re already discussing.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#555

Post by kenjackson » Sun, Oct 23, 2022, 4:38 PM UTC

kelnos wrote: post #457

kenjackson wrote: post #205

Earlier quoted context omitted.

The discussion tends to devolve right around here. Let’s just say we have different data points and agree to disagree.

But isn't that really the issue at hand? You have different data points from the person you replied to. That means that you are both correct, for some subset of people. But do your data points generalize to the larger population, or does the other person's? I don't think these conversations need to devolve. Where are the large-scale studies that show that your data point is the one that's the most common? I'm willing…

The problem is you have people who have anecdotal experience that says whites don’t act in racist ways. I have a bunch of data that I could drop here (100s of sources that say otherwise). I’ve seen these debates online before and I’ve never seen a single opinion changed.

The problem is the things being challenged for many (of not most or virtually all) people is their identity and morality. Not simply an experiment about if blueberries have more antioxidants than blackberries. Data and evidence simply don’t matter in these sorts of discussions. And that’s why they tend to devolve quickly.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#556

Post by stale2002 » Sun, Oct 23, 2022, 5:04 PM UTC

WalterBright wrote: post #527

stale2002 wrote: post #375

Earlier quoted context omitted.

> What I mean by "test prep" is learning how to game the test You'd also be wrong here. Let me give you a simple example, that might not be immediately apparent to someone who doesn't know the rules of these tests. At one point in time, on some tests, there was a "guessing penalty" for getting a wrong answer. Having a basic understanding of how that works, and different guessing strategies, can increase your score. E…

> that might not be immediately apparent to someone who doesn't know the rules of these tests. 1. Read the rules of the tests. Come on! Does one really need handholding with this? 2. Understanding the probability of success at guessing based on how many of the answers you can eliminate is trivial probability, which goes back to my advice to pay attention in math class. 3. "how long to spend on a question" I'm laughin…

> Who needs to be told

And yet despite your incredulous, fake outrage, the stats simply show that test prep helps people significantly.

If you really want more reasons, read the other response that DiogenesKynikos gave to you in the post.

My problem is with you make false statements here. Be as mad as you want about the fake that test prep helps a lot of people, significantly, but do not make false claims, and ignore the fact that the stats show that it works.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#557

Post by stale2002 » Sun, Oct 23, 2022, 5:06 PM UTC

WalterBright wrote: post #526

kelnos wrote: post #387

Earlier quoted context omitted.

Why does it seem so odd to you that learning how the test works, and learning about strategies that work for the specific test, can increase your score? It seems obvious that this would be the case, and it's pretty clear that this actually has helped many people (myself included) get a better score on subsequent attempts.

Why not just learn the material instead? It's a much more useful skill. And you can be proud that you did it without gaming the test.

> Why not just learn the material instead? It's a much more useful skill. And you can be proud that you did it without gaming the test

Because gaming the test works, as the stats prove. And these tests are very important for college admissions.

There is no reason to handicap one's self, just because someone from the peanut gallery is yelling in incredulous outrage.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#558

Post by hintymad » Sun, Oct 23, 2022, 5:25 PM UTC

SamBam wrote: post #499

hintymad wrote: post #492

Earlier quoted context omitted.

The problem of SAT is that it's too easy, or at least too easy to be gamed by tutoring. That is, family with means can really improve their kids' scores by throwing money. Make SAT as challenging as India's JEE or those national entrance exams in East Asian countries, and we will see that tutoring hardly will matter in large scale but teacher's quality will do.

Are you saying that students with money to spend on tutors don't do better at the JEE or Chinese entrance exams? I don't know otherwise, but that seems unlikely to me, so I'm interested if you have any evidence that poor kids pass these exams at the same rate.

> Are you saying that students with money to spend on tutors don't do better at the JEE or Chinese entrance exams?

Some will do, for sure. However, when tests are challenging enough, teachers and talent become dominating factors. Good teachers make sure that students work hard and have access to illuminating exercises to realize their full potential, and students with insufficient talent will not excel in those tests no matter how much tutoring they have. After all, tutoring is not education. Of course, I may have been biased per my own observation: the distribution of student performance in my schools hardly changed due to tutoring because my teachers simply gave challenging enough homework and exams, which did a good enough job to remove the advantages of tutoring, if there's any. Using an extreme example: yes you can find the best coach for you to prepare Math Olympiad, but you'd be able to cross a threshold first before you can even understand what the coach talks about. Similarly on the level of JEE or national entrance exams, I may spend 10 times as much time memorizing some patterns while my classmates whoop my ass by simply attending classes -- this is also why I think the progressives are doing a deservice to poor kids or minority groups. They could've raise standards and build better public schools to help the talented stand out, yet they choose to lower standards and really damage the perception on the otherwise truly talented group.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#559

Post by onetimeusename » Sun, Oct 23, 2022, 5:59 PM UTC

rayiner wrote: post #553

onetimeusename wrote: post #248

Earlier quoted context omitted.

I don't think WASPs are given an edge here. I feel like it is actually racist to make this assumption also. White people as a broad category are the only racial group that is underrepresented relative to their population percentage in Ivy League schools. Maybe you could make the argument that WASPs are a subset of white people who are overrepresented specifically because of that faculty benefit for admissions but the…

There is no such thing as “white people.” WASPs are a distinct tribe and feel zero affinity to white ethnics, Appalachians, etc.

Isn't that a bit like saying that there is no such thing as Asians? Both are still used for admissions it seems.

I don't think WASPs have any influence on the admissions policy at Harvard and I have seen no evidence that they are over-represented there. Others have pointed out who are. I could go on about US history, immigration, and WASPs but its better to be brief. I think they lost their power. That better fits the data than the theory that WASPs control Harvard and despite being racist and anti-semitic seem to prefer every racial group but their own. It's really just a racist trope at this point that untalented, privileged, and elitist WASPs are the villain.

Re: Colleges that ditched admission tests find it harder to fairly choose students

#560

Post by LorenPechtel » Sun, Oct 23, 2022, 6:02 PM UTC

kenjackson wrote: post #169

bombcar wrote: post #158

Earlier quoted context omitted.

I assume the stories of black questions being dropped.

Ahh. See: https://digitalcommons.law.scu.edu/cgi/viewcontent.cgi?artic... There are some good snippets, but I’m on my phone. The whole thing is worth a read though.

Discrimination is an interesting thing, look hard enough and you'll find it whether or not it exists.

An alternate explanation for what they are finding in that paper: The creators of the test favor questions whose response correlates well with how the person scores overall. A question that low-scoring people are more likely to get right is a bad question.

What the researchers should have done is looked at racial differences in answering questions given the same overall score. Of course, that won't find discrimination so they're not going to publish it.