Hacker News Reader

Comment by vineyardmike | Hacker News Reader

> I would add a generic “culture fit” component to each candidate score which you could use as a hedge to admit legacies without calling them as such.

This is not new. This is a battle as old as time.

Want to keep out poor people? Require them to live on campus instead of locally at home. Want to keep out the wrong kind of person? Start requiring college essays to get a "culture fit". Or add "geographic diversity" to get less NYC Jews, or require "well rounded" candidates that do more than pass tests to keep out Asian Americans. Or conduct interviews so you can see their race in-person without asking for it on a form.

https://en.wikipedia.org/wiki/Antisemitism_and_higher_educat...

https://en.wikipedia.org/wiki/Students_for_Fair_Admissions_v...

View on HN

> require "well rounded" candidates that do more than pass tests to keep out Asian Americans

While I agree with you that vague assessments like "well roundedness" can and have been use for racial discrimination in the past (both intentionally and unintentionally), it doesn't mean we should throw the baby out with the bathwater and solely use standardized tests or test scores to determine admissions.

There is critical value in assessing these hard to measure qualities for creating a student body. Each student in the university is not simply consuming an educational good in isolation from one another but is also offering their experience and perspective to the community. Having everyone maxed out on test scores at the expense of such diversity would be a travesty to the thing that makes campus life vibrant.

> There is critical value in assessing these hard to measure qualities for creating a student body. Each student in the university is not simply consuming an educational good in isolation from one another but is also offering their experience and perspective to the community.

Yeah and imagine how awful it would be if they got the experience and perspective of asians.

Seriously, the supposed benefits of these things are made up and no-one ever checks whether they're assessing the things they nominally claim to be assessing. The racism isn't some accidental side effect, it's the whole point.

The “supposed benefits” of well-roundedness? As GP said, they don’t doubt it is used for discriminatory reasons as well, but are you implying there aren’t benefits to being well-rounded and it is a made up characteristic?

Yes, in my experience "well-rounded" is 100% a made up characteristic that generally means "person like the person doing the assessing". Another reply mentions "interesting conversationalist", which mostly selects for someone having the same cultural background, and is the opposite of "diversity" or whatever this week's excuse for doing this stuff is.

And your experience comprises how many cases?

You are making a lot of assumptions here. People don’t find others interesting conversationalists if they have the same background. Maybe if all you focus on is skin color, you may be right, but does somebody in Ukraine have the same background as somebody who grew up in South Florida? Does somebody who grew up in the San Francisco have the same background as somebody who grew up in Marin?

If all you look at is race, you might say yes, but these are very different life experiences. Also, there is such a thing as somebody being so different that it’s not possible for others to relate to them. Likability is not unimportant when it comes to working in a team.

Laptop professionals are remarkably similar wherever in the world you find them these days. London, New York City, San Francisco, Tokyo, Paris, etc. have all been converging on a similar set of tastes, fashions, beliefs, and consuming habits. So it's possible to have great geographic diversity, without introducing much diversity in terms of culture, class, political and religious beliefs, etc.

And conversely great diversity without geographic diversity or racial diversity. Diversity is oversimplified and measured incorrectly from the DEI perspective.

There’s definitely some racism, but intangible qualities can also boost some Asian students who otherwise look like basically everyone else applying to top schools.

Easily quantifiable check boxes don’t verify that someone is an interesting conversationalist. Arguably schools are better served by slightly lower standard and a random pick vs everyone whose parents have been min maxing the process since preschool. Overfitting arbitrary criteria is easy, but not productive.

> intangible qualities can also boost some Asian students who otherwise look like basically everyone else applying to top schools.

In theory maybe. In practice the overwhelming majority of the time it's just used to admit fewer asians.

> Easily quantifiable check boxes don’t verify that someone is an interesting conversationalist.

If we actually cared about whether people were interesting conversationalists in an objective sense (rather than just interesting to the person making the admissions decision - which mostly just comes down to having the same cultural background), we'd figure out a way to test it. These universities never tried, because they never actually cared about interesting conversationalists in the first place, it's always been nothing but a fig leaf.

> Arguably schools are better served by slightly lower standard and a random pick vs everyone whose parents have been min maxing the process since preschool.

Then make it random, if that's the goal - have an actual fair lottery between everyone who meets the standard. But again, it was never about being random.

> cultural background

Language isn’t culturally agnostic. If classes where taught in Malagasy being well read would refer to a different set of books.

> Then make it random, if that's the goal

That’s not the goal, the point is any system that can be gamed will be gamed. You can’t game random, but you can easily have someone else write a kids collage admission essay which becomes more likely the more you weight it and the higher bar you set.

> we'd figure out a way to test it.

In many ways that’s why the SAT is preferred over the ACT. Having a large vocabulary, being able to express yourself, being able to think logically are all reasonable proxies. It also explains why the math section excludes calculus questions as transcripts already show if someone took calculus so they can focus on something else.

> There is critical value in assessing these hard to measure qualities for creating a student body

Value To Whom, university administrators?

Has benefit to students ever been demonstrated and proven?

Can you clearly state what is your hypothesis?

1 - it is preferable to be well rounded instead of narrowly specialised ?

2 - this preference can be taught at the age of ~18 and is not innate, and is not ingrained at early childhood ?

3 - excluding narrowly focused students and getting well-rounded ones instead benefits these students?

4 - university administrators are competent as assessing these yearning of the soul from a student’s essay?

You can join any private club you want with any composition of well-rounded people. What does that have to with higher education?

> Having everyone maxed out on test scores at the expense of such diversity would be a travesty to the thing that makes campus life vibrant.

What’s next? Cities imposing such restrictions? Should NYC or Austin require people to pass vibe check to ensure that the city life is vibrant?

> Should NYC or Austin require people to pass vibe check to ensure that the city life is vibrant?

Don't give them ideas

The ideas are there.

The willingness to overtly implement them has been relatively low since the late 1960s.

> You can join any private club you want with any composition of well-rounded people. What does that have to with higher education?

I believe that a key component of an effective education is studying the roots of philosophy. Surely you agree that the State should not prevent me from forming a private university that mandates freshmen take a philosophy class.

I also believe that a key component to an effective education is exposure to peers who come from a wide variety of different backgrounds and life experiences. Surely you agree that the State should not prevent me from forming a private university that considers the creation of a diverse student body as one minor factor in admissions.

"Surely you agree that the State should not prevent me from forming a private university that considers the creation of a diverse student body as one minor factor in admissions."

The SCOTUS decision on affirmative action in college admissions has at least restricted race from that consideration.

Looking at demographics we were beginning to approach a point were affirmative action would have helped out white communities.

The decline and fall of white power America. If only MLK had lived to see it.

That's not at all what the SCOTUS case focused on.

Realistically, formal higher education is not simply a private matter. It's a part of the complex web of accreditation, government subsidies and entrenched social institutions (not necessarily state institutions).

A university is a public accommodation. You can certainly create a book club among your friends and forbid people of the opposite gender to join or require everyone to be of a different gender, maybe you can even call it a university; but that wouldn't be the same as doing such thing on the level of a large educational facility that e.g. provides the degree of Juris Doctor that allows you to take a bar exam.

So to answer your question, "Surely you agree that the State should not...", I would say "it depends on the particulars".

Private universities aren't. They get loads of research funding, tax breaks, people paying for their education with government backed loans. All American universities are to some extent public.

>What’s next? Cities imposing such restrictions? Should NYC or Austin require people to pass vibe check to ensure that the city life is vibrant?

That's pretty much what HOAs and micro-managerial local ordinances are. The whole point of them is that they make it an expensive hassle and generally crappy to either live in an above your social class. It gets kind of plausibly deniable on a city level when you've got nice neighborhoods and poor neighborhoods and they just differentiate by the degree of enforcement.

Obviously none of this stuff is water tight. It's all a sick game of relative probability. Some low class new money professional sports/entertainment types will retire to some waspy neighborhood in the Hamptons and persist but less of those people will do so than if places like that didn't actively try and be a nuisance to live in for the "wrong type of people". Likewise some guys who have a dozen cars in their yard will persist in their locations as the neighborhood gentrifies around them, much to the annoyance of their neighbors, but most of them will cash out and move out because having your neighbors constantly calling the government to harass you using laws that didn't even exist when you moved in gets real old real quick.

This is quite similar to the strongest argument for legacy admissions, even if the sons and daughters of the wealthy and powerful don't have the best test scores they contribute significantly to the value of going to that institution for other students by virtue of offering them access to those who are going to inherit wealth and power.

Before wringing our hands too much about antisemitism or anti-Asian prejudice in universities, this is what the demographics of the Ivy League looked like in 2023:

https://archive.org/download/ivy_league/ivy_league.png

The point of this law is to reward merit and hard work and discourage universities from offering back-doors for wealthy donors and alums. The point of it is to encourage fairness.

Unless you're suggesting that Asians are overrepresented because their parents are part of an elite old-boys network that gives them an unfair advantage, I think you're missing the point here. If you want to suppress the number of Asians in school because their numbers at ivies are out of proportion with their numbers in the broader population, it sounds like you want more legacy-style admissions rules, not fewer. Maybe this is what you're suggesting and I just I'm just misunderstanding you.

I'm not suggesting anything, just adding needed context to the discussion. E.g. if you want to suggest that these institutions are rife with systemic white supremacy, be my guest. Just include in your assertions explanations for why there are, per capita, 8x as many Asians, 11x as many Jews, and 1.4x as many Blacks, as there are non-Jewish Whites, in the Ivies, despite Whites' many privileges.

Edit: Self-selection is at best an incomplete explanation. It fails to explain how, when comparing non-Jewish Whites vs Blacks, Whites' 177-point average SAT-score lead results in a 1.4x admission penalty. Meanwhile Asians' 73-point lead over Whites becomes an 8x admission advantage.

Scoring well on the SAT is an advantage for other groups, but somehow a disadvantage for non-Jewish Whites.

Legacies are not well allocated distributions either according to per capita numbers.

Arguments like this lead to universities severely discriminating against jews in the not so distant past.

https://en.wikipedia.org/wiki/Jewish_quota

I haven't made any arguments, I just presented the data.

If you hadn't said "Before wringing our hands too much about antisemitism or anti-Asian prejudice in universities," that might have been true. If you had just said "here is this data", then you would have "just presented the data".

You clearly expressed by that phrasing that you thought that the data in question would at least potentially be a reason to not wring hands about such things, and presented it for that reason.

If implying that data might maybe potentially apply to the conversation is an "argument", then you got me.

You didn't imply it might potentially maybe apply. You definitely, clearly, stated it not only applies but what the "correct" conclusion to come to was.

A policy that would exclude all Jews would potentially affect some 8 million Americans, same as a policy that lowers White participation by 3%. Which would be worse, and why?

I think the GP argument is that you can't argue going all on merit for Jews while demanding that Blacks and Whites are somehow equally represented.

self-selection.

The population that applies to Ivies is completely different from the overall population, and different from the population that gets admitted to Ivies.

If there were no requirements to be admitted to Harvard, any tom dick and harry could send his application - only then you can reasonably conclude that the admitted population should reflect the overall population.

But because there are requirements like SAT GPA etc, there is some filtering happening and population that apply is slightly different.

But the affirmative action zealots require that the admitted population must represent the overall population, despite the fact that incoming applications have completely different distribution IQ/SAT/GPA/race wise.

This leads to discrimination, where White/Asian admits, who are overrepresented among applications with high scores, are clamped at certain threshold and then other races are selected with whatever grades they have

Just to add more context to the provided data. It provides mean nationwide SAT score numbers, and the provided comparison assumes that nationwide scores are reflective of Ivy applicant scores.

Also a per capita comparison assumes that the number of qualified applications follows similar distribution, no? I'm not sure if this is reflected in the provided data.

Also, the overall analysis assumes that per capita distribution is fair but that seems subjective. Even so, two schools skew the data for Black students (75% of Ivies are <1x per capita for Black students) and of course there is no mention of Hispanic students (one of the fastest growing demographics) which is mostly underrepresented on a per capita basis.

And then it doesn't get into international students and if/how they assimilate into a "race". Nor does it reflect stickier topics such as whether Hispanic students culturally assimilate into "White", effectively lessening their numbers under a per capita comparison (it does this for Jewish students).

I appreciate what the data brings to the conversation, but don't believe others' assertions have to take any of it into account considering the number of assumptions one must make to follow a "per capita" AND population SAT = sample SAT comparison.

> It provides mean nationwide SAT score numbers, and the provided comparison assumes that nationwide scores are reflective of Ivy applicant scores.

Does it really assume that? Suppose, for the sake of argument, one group had a nationwide average SAT score of 1500, and all other groups had an average SAT of just 500. Barring any bizarre distributions of those scores, we can infer from only the averages, that the 1500-SAT group would have more individuals that satisfy a university's academic criteria, than the 500-SAT groups. It's far from perfect, but does provide a hint.

> Also, the overall analysis assumes that per capita distribution is fair but that seems subjective.

I must have missed where in those charts a definition of 'fair' is given, and then relied on for further analysis.

> of course there is no mention of Hispanic students

Hispanic students are between Asian and Black on every chart.

> And then it doesn't get into international students

That is correct, international students are entirely excluded, in the sense that all the domestic students in a school are taken to represent 100%. I don't understand how not answering all these additional questions you raise makes the data irrelevant.

> Does it really assume that? Suppose, for the sake of argument, one group had a nationwide average SAT score of 1500, and all other groups had an average SAT of just 500. Barring any bizarre distributions of those scores, we can infer from only the averages, that the 1500-SAT group would have more individuals that satisfy a university's academic criteria, than the 500-SAT groups. It's far from perfect, but does provide a hint.

There’s no need to construct a hypothetical when there is actual data to dissect. For example, in the cited links, the 25th percentile SAT score for Harvard students is ~200 points greater than the highest mean nationwide score. The middle 50% of all students (25th-75th) range is ~100. And on 7 out of 10 students admitted included SAT scores in their application. So one would have to make additional assumptions (I’m not sure what they are) to claim one’s groups scores lead to penalty and another’s leads to advantage. It could be true, but I don’t see it as a fact, hence my original position that other assertions don’t necessarily need to meet some bar.

A lot of the nonsense comes from the bonkers categories. Hispanic origin is sort of like being Jewish from a statistical perspective - it’s a layer, not a state.

We’re also decades after civil rights. People don’t fit in these boxes.

My nephews dad is Irish, mom is black Puerto Rican. Racists would consider him black, his name is Irish, mom is of African and Spanish origin. They don’t speak Spanish at home or really have a deep connection to Hispanic culture. Wtf is he for the college demographic survey?

Likewise for my Filipino friends… are they Asian? Pacific Islander? This particular family speaks English, Tagalog and Spanish at home. Culturally they are very much into traditional Filipino traditions, but their Catholicism practice is close to Spanish style.

> The point of this law is to reward merit and hard work and discourage universities from offering back-doors for wealthy donors and alums. The point of it is to encourage fairness.

The point of donor child admissions, as I understand it, is to bring in additional money. Which could potentially enable more low-income students to attend.

The reason why some universities have ended legacy admissions recently seems to be that they have concluded that it doesn't bring in enough money. But they usually still have donor admissions.

It's not clear what your point is exactly. It's still possible that Asians, despite being overrepresented, receive some sort of penalty in admissions.

For any penalty Asians receive there are other ethnicities that appear to receive a much harsher one.

Which ethnicities would those be, and how does this follow from the data above?

Your skill at crimestop is commendable.

What's the implication here, that asians should just suck it up because others have it worse?

We're stuck with a math problem. If you admit purely based on test scores then black and latino applicants would be significantly underrepresented and non-Jewish white applicants would be slightly underrepresented. If you address these problems by insisting on a full balancing then you're going to have to reduce Asian and Jewish admittance. But increasing black and latino admittance at the expense of non-Jewish white admittance when the latter group was already underrepresented is not only the same kind of quota system but not even satisfying the goal of proportionate representation -- and that's the one that seems to be happening. So what do you want to do?

Addressing the underlying barriers preventing blacks/latinos from getting into elite universities, rather than trying to fix the symptoms by instituting illiberal policies like reverse discrimination.

Then you need people to stop getting mad at universities when admission based on test scores causes those groups to be underrepresented. They're not the ones who can fix the test scores and nothing is going to fix them before the next round of applications.

Arguably, the problem of underrepresentation became already unsustainable during Affirmative Action by the classification of African immigrants as black. Because the latter displace African-Americans who do worse in test scores than Africans.

https://www.theguardian.com/education/2007/may/29/internatio...

break us colleges into tiers based on academic rigor. assuming you met XXXX standardized test score you are automatically sorted into one of the appropriate schools at random on your 18th birthday.

no take backs, safety schools permitted.

Looking at mean SAT grades is close to irrelevant when your (supposed) strategy is to pick Top-N candidates. If for example every single person got exactly the mean score allotted to them by their race, you'd actually expect that the top 5.9% of universities would all be 100% Asian, the next 57.8% of universities would be purely "White (incl. Jewish)", the next 18.7% of universities purely Hispanic and bottom 12.1% purely Black.

By that math, every single slot in an ivy league school "belongs" to an Asian candidate and even a single white person is already over-representation.

(Didn't separate White from Jewish since they don't have mean scores for just Jewish, possibly they should be first according to the estimate in the text below, but that doesn't really matter for the point made)

> Looking at mean SAT grades is close to irrelevant when your (supposed) strategy is to pick Top-N candidates.

Not if the mean SAT score predicts how many students surpass some academic cutoff used by the universities. It's like predicting who will win a best-out-of-3 100m dash, when all you have are the runners' mean 100m times.

I find it baffling how people become incapable of the simplest inferences, and capable of the smallest nitpicks, when they don't like where the data leads.

As a note of caution: these demographic statistics are somewhat misleading because they use the demographics of the US as a whole, but the correct demographic set is to use for 18 year olds (70 year olds generally aren't applying to college)

if you look at the ethnic breakdown of children in the US in 2023 (https://datacenter.aecf.org/data/tables/103-child-population...):

48% non hispanic white

26% hispanic

14% non hispanic black

6% asian

5% mixed

1% native

When you look at these statistics through this lens you see that white relative underrepresentation is slightly reduced, hispanic significantly increased. broadly speaking, when you look at younger ages the country is a little bit less white and a lot more hispanic.

Perhaps we should investigate the systemic discrimination that is causing lower birth-rates in whites. There are whole academic fields for similar disparate outcomes in other groups, so why not treat this case the same way?

We are witnessing late-stage higher education before AI guts out the education parts.

Reply on news.ycombinator.com