4 X more "unqualified" white students admitted to Harvard than black students

[Loren Pechtel]

I don't believe they are 'massively impacted' by SES or any of those other things, inasmuch as I think most of the variability in scores is indeed attributable to the academic aptitude of the test taker.

Well, "massive" is vague enough.

You don't think a 16 percentile difference between first and 3rd attempt counts as "massive", that what you're trying to say?

And how much of this is coaching vs simply time? I scored better the second time I took the SAT--but I was also a year older. No coaching.

 

[Loren Pechtel]

From evidence presented at the Harvard trial:

View attachment 28588

Which makes it clear race is the big factor, the other things are being brought up to distract from the real issue.

 

[Loren Pechtel]

No, it only needs to maintain an average. If you get enough for 2,500 students one year and 1,500 the next but admit 2,000 donor students each year you're ok.

Pay attention - if the school only has room for 2,000 students, it will not admit 2,500 students.

Try addressing the actual issue.

The point is donor money can be used to expand the number of students they can admit. To the extent that it does donor admits are a very minor issue. He was making the mistake of thinking that it would have to vary the number each year, but all that actually matters is a rolling average, not a year-by-year value.

 

[Metaphor]

I was going by the figures in this graphic provided by Derec (year not stated) and this report (2007 to 2016 cumulative), which put the figure at 7%/10% respectively.

Even so, 14% is still higher than any plausible figure you can pull from where the sun never shines about black students admitted only due to AA - unless you want to go on record claiming that, in the absence of AA or any other non-merit-based systems, black students would constitute less than 0.3% of Harvard's students. Is that a claim you're willing to defend?

No. I posted a graphic earlier (https://talkfreethought.org/showthr...black-students&p=811630&viewfull=1#post811630) that was part of the evidence in the Harvard trial.

The table presented the statistically modelled/simulated results of various configurations of admissions policies at Harvard. The table only models four races, so the percentages should only be compared within the table itself, but it shows that if all "non merit" policies were applied, black admissions would drop from 13.8% to 4.4%, Hispanic admissions would drop from 13.8% to 8.1%, white admissions would increase slightly from 48.5% to 50.8%, and Asian admissions would increase from 23.8% to 36.6%.

Yeah, because admitting that you thought the number is lower than it actually is hurts so bad...

No, it doesn't "hurt so bad". I knew it was some fraction of 43% and the 43% was made up of three categories, and I did a rough estimate of 15% and considered that a small fraction of all white admissions.

But, here you go, if it matters to you so much. Any percentage of "non-merit" admissions is too high a percentage.

To the student who wanted to study at Harvard and now can't, what difference does that make whether he lost his place to a black or white student? None at all.

To the student, I suppose the race of the person who they lost out to probably doesn't matter much, though the reason might. I personally would feel less bad about losing a spot I might have gotten, if the spot had gone to an athletic admission, because while I think athletics have nothing to do with academic merit and shouldn't be a selection criterion, it's still a kind of accomplishment, something an individual student did.

But I'd feel more upset if I lost out to somebody who got in over me because their parents went to Harvard, or their parents already worked at Harvard, or because they were the right race and I were the wrong one.

 

[Metaphor]

Trausti might be a candidate, he is, as far as I can tell, the only person who explicitly offered defenses of legacy admissions in this thread, though I don't claim to read his mind. How is this relevant though? Where does the OP (or any other post by ronburgundy, Toni, myself, whoever else you want to classify as "our camp") attest that people on this board are examples of that?

Or is this another of your straw men?

Nobody attested it. I wanted to know who it was aimed at. If the answer is 'nobody in particular', that's fine.

Unlike you, I am under no impression that I can read other people's minds, so you'll have to ask everyone individually.

I can only attest for myself: I only joined the discussion yesterday to call out your straw men, which have reached an unhealthy level once again. As far as I can reconstruct, my first contribution was this: https://talkfreethought.org/showthr...black-students&p=811354&viewfull=1#post811354, responding to a post where you falsely insinuated that @ronburgundy 's position implies eliminating high school grades from consideration.

Within the next couple pages of discussion you insinuated that someone had claimed SAT score measures SES status, or insinuating that someone in this thread (presumably laughing dog) fails to understand that some legacy admissions are black.

You haven't defended any of those insinuations - the closest I've seen is something along the lines of "I may have read something similar elsewhere on the web" (my paraphrasis, from memory, which might be inaccurate).

Either one of those two things has to be true:

- you are arguing in good faith, but failing to actually read what people are saying, imagining they're saying things they didn't - in this case, you really, really should work on your reading comprehension skills, preferably before engaging in another online discussion, as it is, frankly, quite offensive at times and not conducive to a productive debate.

- you're deliberately using straw men to derail the discussion.

Well, thanks for presenting a false dichotomy.

Ronburgundy's position implies he ought to object to high school grades for the same reason he objects to SAT as a measurement tool. But if he doesn't object to it being used then there's nothing much to say.

The repeated emphasis that SAT scores correlate with SES status of course carries implications: that they are therefore illegitimate as measures of academic aptitude. If that isn't the argument being made, what difference does it make what it's correlated with?

Some legacy admissions are black, and laughing dog's post, and his defence of the post, sweeps aside this relevant fact. He doubled down: he said legacy admissions are

a de facto form of admitting unqualified white people.

If that is what they are, then they are also a de facto form of admitting unqualified black people, and unqualified Asian people, indeed unqualified people of any race.

 

[laughing dog]

No, it only needs to maintain an average. If you get enough for 2,500 students one year and 1,500 the next but admit 2,000 donor students each year you're ok.

Pay attention - if the school only has room for 2,000 students, it will not admit 2,500 students.

Try addressing the actual issue.

Your inability to comprehend does not mean the issue was not addressed. If there are only X spaces, then only X number of students can be admitted. It is called a binding constraint.

The point is donor money can be used to expand the number of students they can admit. To the extent that it does donor admits are a very minor issue. He was making the mistake of thinking that it would have to vary the number each year, but all that actually matters is a rolling average, not a year-by-year value.

It is not possible to admit more students than capacity, so your rolling average hypothesis requires that the capacity is sufficiently increased. And if the capacity were increased, why on earth would there be a rolling average with ups and downs at a place like Harvard? Your example is both evidence-free and illogical.

 

[J842P]

Try addressing the actual issue.

Your inability to comprehend does not mean the issue was not addressed. If there are only X spaces, then only X number of students can be admitted. It is called a binding constraint.

The point is donor money can be used to expand the number of students they can admit. To the extent that it does donor admits are a very minor issue. He was making the mistake of thinking that it would have to vary the number each year, but all that actually matters is a rolling average, not a year-by-year value.

It is not possible to admit more students than capacity, so your rolling average hypothesis requires that the capacity is sufficiently increased. And if the capacity were increased, why on earth would there be a rolling average with ups and downs at a place like Harvard? Your example is both evidence-free and illogical.

I'm highly skeptical that institutions like Harvard need legacy admissions to expand their student capacity, or that it makes any appreciable difference. I believe these institutions have plenty of resources and merely lack the will, most of all.

 

[Jokodo]

Unlike you, I am under no impression that I can read other people's minds, so you'll have to ask everyone individually.

I can only attest for myself: I only joined the discussion yesterday to call out your straw men, which have reached an unhealthy level once again. As far as I can reconstruct, my first contribution was this: https://talkfreethought.org/showthr...black-students&p=811354&viewfull=1#post811354, responding to a post where you falsely insinuated that @ronburgundy 's position implies eliminating high school grades from consideration.

Within the next couple pages of discussion you insinuated that someone had claimed SAT score measures SES status, or insinuating that someone in this thread (presumably laughing dog) fails to understand that some legacy admissions are black.

You haven't defended any of those insinuations - the closest I've seen is something along the lines of "I may have read something similar elsewhere on the web" (my paraphrasis, from memory, which might be inaccurate).

Either one of those two things has to be true:

- you are arguing in good faith, but failing to actually read what people are saying, imagining they're saying things they didn't - in this case, you really, really should work on your reading comprehension skills, preferably before engaging in another online discussion, as it is, frankly, quite offensive at times and not conducive to a productive debate.

- you're deliberately using straw men to derail the discussion.

Well, thanks for presenting a false dichotomy.

Ronburgundy's position implies he ought to object to high school grades for the same reason he objects to SAT as a measurement tool.

And there is your reading comprehension problem (or your strawmanning habit, whichever it is) rearing its ugly head again. He doesn't object to scores as *a* measurement tool, he objects to using them as the sole criterion as alluded by Trausti, while pretending they are somehow floating in the ether unaffected by context, and he was quite explicit about that. He quoted like a dozen papers from the science of psychology showing that such a position isn't based in science but in cargo cult thinking, too.

You can argue against Ron's position using better science, you can accept it, or you can argue against your own brand of things he never said. One of them, and one only, makes you look stupid, but it's your choice really.

 

[ronburgundy]

I don't believe they are 'massively impacted' by SES or any of those other things, inasmuch as I think most of the variability in scores is indeed attributable to the academic aptitude of the test taker.

Well, "massive" is vague enough.

You don't think a 16 percentile difference between first and 3rd attempt counts as "massive", that what you're trying to say?

What per cent of the total variability in SAT scores do you think is actually attributable to academic aptitude? 10% 50% 90%?

The way it has been discussed by some, they seem to have no problem uttering obvious falsehoods such as 'SAT score measures SES status', which it does not.

I noticed you didn't use real quotes. Did anyone actually say that?

Also, your question is too broad to be meaningful. Particularly with any psychological or aptitude test, how much a measure captures X and X alone is completely context dependent. My prior link to the research on Raven's Matrices, often considered one of the most "pure" measures of IQ, illustrates this. If two people being compared both have zero experience with that test then a much larger % of the difference in their scores is a result of general "aptitude" differences, although interest and motivation to take that test in the given testing conditions are still factors. But if one person has taken it twice already, they may score 1/4 standard deviation higher (which is pretty sizable) which could be the exact opposite of the actual difference in the general abilities the test is supposed to measure. IOW, a person who is 1/4 standard deviation higher then another person in the true general aptitudes Raven's is intended to measure will score 1/4 SD lower if that other person has experience taking the Ravens. So, in that comparison 0% of the variance in those people's scores reflect true variance in aptitude. In fact, it's analogous to negative 100%, if such a thing were possible, which it does kinda operate as a negate when added into the computation of total variance explained in a larger population.

So, the answer to your question is that it changes with changing degrees of contextual factors that vary both across time and between sub populations, and is not constant across the full range of scores. That last point is important, b/c most practice/training designed to improve scores on some test wind up having a "rich get richer" effect, which is a standard term referring to those who already have higher scores are the one's most helped by that practice and training. This is usually b/c people whose current skill level is low (either for innate or prior development reasons) lack the neccessary general skill to gain test-specific skills from the opportunities that practice and training afford. So, paying for extra training or test attempts doesn't account for that much of the difference between those at the 30th vs. 40th percentiles. The fact that most of them are lower SES and can't afford such opportunities anyway makes that even more true. But those test takers don't wind up applying to college anyway, so they aren't central to the issue.

In contrast, for people whose true aptitude is already in the top 30th percentile, those who comprise the vast majority of college applicants, training/practice opportunities can make a big difference, boosting them from the 75th vs. 91st percentile scores, which would be the deciding difference at most schools. So, let's say that across the total variance in all SAT scores, just 15% is due to SES related factors tied to paying for test-specific opportunities like tutors, courses, more tests, etc.. And suppose that 2/3 of that 15% was concentrated in the top 30th percentiles of test takers (due to "rich get richer" effects). That means the 1/3 of the variance is due to these testing-specific SES factors, among those students for whom it matters b/c they are applying to college. It means that a sizable % of the time using SAT scores will favor the wrong student.

Now, other than that coaching and testing has large effect sizes on scores, we don't know the exact % of total variance it accounts for. But my point is to illustrate why your question isn't meaningful, b/c even if the total variance it accounts for is rather low, the % it accounts for is likely much higher in the high end of the distribution where all the stakes are.

Finally (I know, right), I kept talking about "test-specific SES factors" to distinguish from the huge influence we know SES has on the development of real academic aptitude via determining the quality of teachers, classroom materials, computers at school and home, reduced external stressors, not having to work in high school, etc.. All of those impact the very academic skills that do determine college success. It's a problem, but not a problem to be fixed at college admissions b/c it's too late by then. My arguments about SAT and ACT are not about those SES influences, but about the influence of wealth on paying for opportunities directly about boosting admission tests scores.

 

[ronburgundy]

From evidence presented at the Harvard trial:

View attachment 28588

This is a red herring. This question of the OP is not which policy affects the racial makeup of the student body more. Unless, you want to admit that conservatives are upset that affirmative action means more blacks get to go to college.

The question is who are taking most of the slots of more qualified white students? And that Table cannot answer that question, b/c the whites admitted under the "No legacy" are not the same white students (they are more qualified whites) as admitted under the current model. That table suggest that around 500 white students are losing slots to either blacks (about 300) or Hispanic students (200).
But it doesn't how many white students are losing their seats to ALDC students, b/c it is different white students who would be admitted under each model.

The other research shows that 32% of whites currently admitted would by rejected w/o the ALDC preferences. So, that is about 1600 whites taking slots from more qualified students. Your graphic suggest a few hundred of those slots would go to non-whites, especially Asians, leaving about 1300 that would go to more qualified non ALDC whites.

In sum, less academically qualified ALDC whites take 1300 slots from more academically qualified whites, while affirmative action only transfers about 300 slots from more academically qualified whites to blacks. And look at that. 1300/300 = 4.33 which is oddly close to my "4X" claim in the thread title.

Thus, anyone pretending that their real concern is simply more academically qualified students losing slots to less qualified people would be more bothered by the preferences mostly given to whites than by affirmative action. Thus, anyone who bitches about AA while virtually never mentioning legacies is simply a racist who is upset that too many blacks and Hispanics are getting into college.

 

[Elixir]

4 X more "unqualified" white students admitted to Harvard than black students

B-b-b-but isn't that what one would expect if it was "fair and balanced"?



Actually, it looks like about 4.7:1

 

[Bomb#20]

Naw.

No qualification ("favoring") or similar. He is pretending that it legacies are "de facto" white people, which is false.

Come on. The long history of Harvard is an institution for WASPS up until the 1960s, so legacy admissions were most certainly a form of affirmative action for children of "legacies" (who were white). Even now, since the vast majority of Harvard attendees are white, legacy admissions are still a form of affirmative action for white people.

He thinks that presently "vast majority" of Harvard students are white. That is demonstrably false.
He also isn't writing that legacy admissions are "favoring" white people; he is claiming that they are a "form of affirmative action for white people". Another falsehood.

Exactly where is he writing that last bit? Is that supposed to be a direct quote? I searched the thread and the only mention I find of that phrase is Metaphor's straw man. When I challenged him to provide a quote, he evaded saying that he's seen it elsewhere on the internet often enough (which makes it still a straw man since this discussion isn't elsewhere on the internet but here).

Maybe you can do better than him?

Instead of searching the whole thread, which is so long it's easy to miss something, try searching your own post. Yes, it was supposed to be a direct quote. That's why Derec provided a direct quote. And when you quoted Derec's post back to him, you thereby included the direct quote from ld that Derec had included. If you click on the little ">>" symbol it will take you directly to the source, post #73.

 

[Toni]

I don't believe they are 'massively impacted' by SES or any of those other things, inasmuch as I think most of the variability in scores is indeed attributable to the academic aptitude of the test taker.

Well, "massive" is vague enough.

You don't think a 16 percentile difference between first and 3rd attempt counts as "massive", that what you're trying to say?

And how much of this is coaching vs simply time? I scored better the second time I took the SAT--but I was also a year older. No coaching.

A lot of it is coaching. There are specific classes that you can take--if you have the money, transportation and time to do so. There are entire pre-K-12 school systems which are dedicated to ensuring that little Reed or Buckley will get into whichever Ivy league school their grandpapa attended.

 

[Toni]

From evidence presented at the Harvard trial:

View attachment 28588

Which makes it clear race is the big factor, the other things are being brought up to distract from the real issue.

It makes it clear that someone make some projections based on models based on their assumptions. Which include race.

 

[Bomb#20]

The point here is that the given issue (fairness) is not the actual issue. If it were, those concerned would be concerned about unfairnesses, but they more or less turn a blind eye to some and not others. 'Reflecting media coverage' seems like a poor excuse. 'Familiarity' even less so.

By that reasoning, any American or Irishman who complains about any unfairness in his own country must be lying about his motives because he apparently cares more about some triviality than he cares about the Janjaweed raping and murdering noncombatants in Darfur. That's a lot more unfair than all of our first world problems, so clearly the given issue (fairness) is not the actual issue.

Context matters. Media coverage matters. Familiarity matters. Some might regard those as poor excuses for the world mostly ignoring the Janjaweed; the point here is that the inference leap from "I feel that's a poor excuse" to "Your real reason is whatever the hell I say it is, and I say it's because you're a racist!" is not a logical inference leap.

This entire thread is a painful exercise in Whataboutism.

 

[Loren Pechtel]

Try addressing the actual issue.

Your inability to comprehend does not mean the issue was not addressed. If there are only X spaces, then only X number of students can be admitted. It is called a binding constraint.

The point is donor money can be used to expand the number of students they can admit. To the extent that it does donor admits are a very minor issue. He was making the mistake of thinking that it would have to vary the number each year, but all that actually matters is a rolling average, not a year-by-year value.

It is not possible to admit more students than capacity, so your rolling average hypothesis requires that the capacity is sufficiently increased. And if the capacity were increased, why on earth would there be a rolling average with ups and downs at a place like Harvard? Your example is both evidence-free and illogical.

May I suggest a very simple thing you're overlooking? Spend money, increase capacity.

 

[ronburgundy]

He thinks that presently "vast majority" of Harvard students are white. That is demonstrably false.
He also isn't writing that legacy admissions are "favoring" white people; he is claiming that they are a "form of affirmative action for white people". Another falsehood.

There are 2-4 times as many whites at Harvard as any other racial group. It is invalid to clump all other races together when evaluating whether one groups gets more benefits than any other. In addition, legacy admits are based upon who among today's 18 year old applicants had Harvard graduate parent. Harvard grads rarely have kids during school, so most of today's applicants would have had parents who got admitted to Harvard around 1980. So, then what was Harvard's ethnic breakdown in 1980? Whites only dropped below 50% at Harvard a few years ago. It's hard to find stats on 1980, but this shows that Asians were only 13% in 1992. So, whites were likely around 75% of students of back in 1980 (and probably higher), making current white applicants about 7 times as likely as any other group to get a legacy preference, probably 10 times as likely as blacks, even 3 times as likely as all other groups combined.

 

[Bomb#20]

First and most obviously, a donor-based admission brings in enough money to create a new seat, so the kid isn't taking anyone's seat away.

There are plenty of flaws in that argument. First, I don't know about Harvard, but in most physically defined universities, there is an actual limit to the number of students. Second, there is no guarantee that the "added" seat (if there is one) will go to a more academically qualified student. Third, there is no guarantee that donor money is sufficient to fund another "seat" or that the donated funds will be used to add seats.

Way to miss the forest for the trees. The number of seats at Harvard has grown many times over the last 384 years, and adding resources is what made it possible. Whether one more seat is added in one particular year because of one particular donation isn't significant; what matters is the long run effect on overall resources of making some of the seats available to rich parents to buy their undeserving kids' way in. Now, if you can show that Harvard is in the habit of selling those seats too cheaply to finance the fraction of its expenditures those seats account for, you'd have a case; but that would take research and statistics, not just a "there is no guarantee" comment.

 

[laughing dog]

First and most obviously, a donor-based admission brings in enough money to create a new seat, so the kid isn't taking anyone's seat away.

There are plenty of flaws in that argument. First, I don't know about Harvard, but in most physically defined universities, there is an actual limit to the number of students. Second, there is no guarantee that the "added" seat (if there is one) will go to a more academically qualified student. Third, there is no guarantee that donor money is sufficient to fund another "seat" or that the donated funds will be used to add seats.

Way to miss the forest for the trees. The number of seats at Harvard has grown many times over the last 384 years, and adding resources is what made it possible. Whether one more seat is added in one particular year because of one particular donation isn't significant; what matters is the long run effect on overall resources of making some of the seats available to rich parents to buy their undeserving kids' way in. Now, if you can show that Harvard is in the habit of selling those seats too cheaply to finance the fraction of its expenditures those seats account for, you'd have a case; but that would take research and statistics, not just a "there is no guarantee" comment.

Way to shift those goalposts. The discussion was not about the long run.

 

[Bomb#20]

Way to miss the forest for the trees. ...

Way to shift those goalposts. The discussion was not about the long run.

Excuse me? Shift what goalposts? I didn't set any goalposts; I explained why rb's argument didn't imply what he claimed it did. If you think your or his choice to ignore the long run somehow makes rb's inference correct, show your work.