Comment by Manuel_D | Hacker News Reader

Comment by Manuel_D | Hacker News Reader

Parent Full thread

I think you're still not not getting the point Abbot is making.

> They already tried to build a system that uses merit and qualifications alone, and the complaint is how doing that has led to extremely biased results.

How did it produce "extremely biased results"? You're being vague here, but I suspect by "biased results" you mean they produced a racial (and perhaps gender) makeup you don't like. Abbot rejects the idea that a non-discriminatory admissions process is biased because it produces student body with the "wrong" racial makeup. His view is that picking a "right" racial makeup and utilizing discrimination to achieve it, is a form of bias.

Also how is it "intentionally inflammatory and insulting"? Are you aware that 62% of African Americans do not support the use of race in admissions?[1] Perhaps what's insulting is and admission process that looks like this: https://www.economist.com/img/b/1280/1482/90/sites/default/f...

1. https://www.pewresearch.org/fact-tank/2019/02/25/most-americ...

>You're being vague here, but I suspect by "biased results" you mean they produced a racial (and perhaps gender) makeup you don't like.

No, you are completely and utterly wrong on that. This is more of the same intellectually lazy assumptions, please stop this.

And I get his point perfectly, it is still insulting and dismissive. When someone starts talking about school vouchers and "merit" in the context of this then it's mostly an admission that they don't know what they're talking about. It has absolutely nothing to do with the problems faced by a school like MIT that already has an extremely low acceptance rate anyway. They know how to test students on academic rigor, that isn't the issue. The fact is, they already cannot accepts all students who pass based on academic merit alone. Your chart even demonstrates that is the case with Harvard, the issue is explicitly not that racial makeup is being considered over academic achievement.

> No, you are completely and utterly wrong on that. This is more of the same intellectually lazy assumptions, please stop this.

Then can you explain what was biased about an admissions process that does not discriminate on the basis of race, gender, and other identity characteristics, if it wasn't the demographics of the student body? It's still vague how a race-agnostic admissions process would be biased.

> The fact is, they already cannot accepts all students who pass based on academic merit alone. Your chart even demonstrates that is the case with Harvard, the issue is explicitly not that racial makeup is being considered over academic achievement.

I'm not sure how one arrives at this conclusion. An African American student in the 4th decile (as in, is scoring above 40% of students and below 60% of students) has the same chances at admission at an Asian student in the top decile. Even at that top of the academic performance an Asian applicant has a ~15% chance at admission while an African American student at the same academic performance has nearly a 60% chance at admission. There is nowhere near equal chances at admission between racial groups at the same academic performance.

The fact that they have many more qualified applicants doesn't make their admission practices any less discriminatory. If I have 500 qualified white applicants and 5,000 qualified Asian applicants and I admit 100 whites and 100 Asians have I not discriminated against Asians? Just because I'm picking both from the same qualified pool doesn't change the fact that I'm making a 10x disparity favoring whites.

>Then can you explain what was biased about an admissions process that does not discriminate on the basis of race, gender, and other identity characteristics, if it wasn't the demographics of the student body? It's still vague how a race-agnostic admissions process would be biased.

There can still be implicit bias. Please read this publication from MIT on this exact subject: https://tll.mit.edu/teaching-resources/inclusive-classroom/i...

>An African American student in the 4th decile (as in, is scoring above 40% of students and below 60% of students) has the same chances at admission at an Asian student in the top decile.

Ok, but that still doesn't change the facts. The issue is not that there aren't enough qualified students or that unqualified students are being placed above qualified ones or that merit is being ignored. That just isn't what's happening at all. That graph is missing a lot of important context, like the actual court case it was talking about that is currently headed for the Supreme Court:

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

The whole thing about "we need more merit" is a misdirection. It's a bad, irrelevant take. This professor should be ashamed to have put his name on such dreck.

>If I have 500 qualified white applicants and 5,000 qualified Asian applicants and I admit 100 whites and 100 Asians have I not discriminated against Asians?

It very well might be so, but now you're coming back around to the same thing about implicit bias again, the very thing that the original article was dismissing as not an issue. The plaintiff even argued in court that Harvard was doing this...

> There can still be implicit bias. Please read this publication from MIT on this exact subject: https://tll.mit.edu/teaching-resources/inclusive-classroom/i...

If anything, this lends even more support to Abbot's proposed changes to admissions criteria. Removal of race from applicants' applications would eliminate the implicit bias, as admission officers would not know the race of applicants. This is like putting the screen between orchestra performers and reviewers in auditions.

> The issue is not that there aren't enough qualified students or that unqualified students are being placed above qualified ones or that merit is being ignored.

Then why are Asians at the top of the class still only half as likely to get admitted than Black applicants in the middle? The fact that you proclaim that all applicants are qualified does not alter the racial disparity in admissions rates.

Here's another blunt example: I have some job that requires a fitness test. I have 1,000 applicants run a a mile. If they can do it in less than 8 minutes they're qualified. 400 applicants make the 8 minute cutoff. 100 of them Black, 100 white, 100 Asian, 100 Latin. I hire 100 white applicants and no one else. When people complain that I only hired people of one race I point out that all of them passed the 8 minute qualification test and no unqualified white applicant was hired over a qualified applicant of another race. Is that a reasonable way to argye that this hiring process was not racially discriminatory.

Reply on news.ycombinator.com