Of course not. What you are saying is analogous to the “women earn 77 cents per dollar relative to men”. A provocative headline. But social scientists didn’t blindly accept this (and they certainly didn’t ignore it). Instead they created studies that adjusted for things such as occupation, hours worked, education, and found that most of the pay differential was explained by these other factors (although a small amount still remained).
This is the same thing I am advocating for here. Let’s have the DOJ lead studies that make the appropriate adjustments here and see what remains. Again, I expect any remaining effects to be small. But just as social scientists didn’t ignore the “77 cents per dollar” situation, they shouldn’t ignore this one.
@SatchelSF said earlier that Duke eventually reduces everything to a set of numeric scores. And the earlier Amherst admissions video showed that a single yellow sheet summarized the data for the small fraction of applicant decisions that went to a Admissions Committee review. And remember decisions for the vast majority of applicants are made before they even get to committee review.
We are talking about the DOJ here. The same department that we trust with investigating and prosecuting sophisticated crimes. Yet we somehow think they are incapable of doing a fairly basic social science study?
And if the remaining effects are small as I suspect, wouldn’t that increase more confidence in the system?
The Hillsdale College case has shown that the federal government can set criteria for when a college qualifies for its students to get federal student aid. There is already an existing precedent here, and it has nothing to do with the quality of the students.
@hebegebe Not sure I’m following you…when you say the point I made about the kid with the 32 vs the kid with the 35, what exactly are you saying shouldn’t be ignored?
My point was more that it’s a judgement call on the part of admissions and being the differences in the applications might be so subjective that it’s impossible to quantify and accurately portray in a “study”…and therefore, we just have to trust the decisions are being made as “fairly” as possible, in a system that we will NEVER be able to quantify exactly what that “fair” balance is. Not you, not me, and not even the DOJ.
This is what needs to be investigated, whether the admissions processes themselves are unfairly discriminating on the basis of race. The DOJ (or judge in the Harvard case) does not really need to make subjective judgments, or review actual admissions materials (although that might prove helpful later on in explaining any statistical anomalies that are discovered). The admissions offices will have already reduced their subjective evaluations - their “judgment calls” - to numbers. It is these numbers that can be compared by race for impermissible bias. We live in a country of laws; we do not have to accept or trust these decisions, just as we do not trust the decisions of private employers or landlords when it comes to race.
An illustration of the scoring system might help. The Arcidiacono study goes through the 5 measurement categories that are important to the Duke admissions office. They have told us what is important to them in their holistic decision process. These categories are: (1) academic achievement (presumably scores + HSGPA + other notable achievements); (2) curriculum (presumably rigor); (3) essays; (4) “personal qualities”; and (5) recommendations. See Table 1, p.7: http://public.econ.duke.edu/~psarcidi/grades_4.0.pdf.
This is the system Duke uses and a version of such a system will be used by any selective college facing large numbers of applicants, adjusted of course to reflect the metrics and weightings chosen by that college.
There are three basic cases with preference admits*, who everyone seems to agree will on average present with lower achievement and curriculum scores (let’s call these “observables,” since they are easily quantified and readily apparent).
Lower observables scores are "outweighed" by some combination of the other metrics such as essay, personal qualities, etc (the "unobservables"). These candidates are not really preference admits: on a holistic basis they meet the criterion for acceptance.
Certain preference groups are granted a dispensation for lower observables, but are fully competitive on the unobservables. This is the classic "bump" theory, namely that preference admits are granted a little leeway on the stats, but are otherwise "fully qualified."
Certain preference groups receive a systematic boost on both observable and unobservable metrics. In this case, these preference admits are not meeting the "holistic" criteria for admission, even with leeway or bumps for the objective stats; really *a whole different set of thresholds* is being applied to the metrics being measured.
Numbers 2 and 3 would be suspect, number 3 in particular, if race is the criterion being used to determine these sorts of preferences.
In the Duke study, it would appear that number 3 is being employed, for blacks and (to a much lesser extent) Hispanics. Blacks are lower (on average) in Duke’s own scoring system on every metric, and typically by significant amounts (ranging from 0.30 - 0.65 standard deviations, approximately). Conversely, the thresholds for Asians appear to be higher, although only marginally on certain measures.**
The above analysis hints at the real reason why the institutions have been fighting the release of information: we are going to find out that - whatever else it is - “holistic” is simply a smokescreen for creating different standards for different racial groups.
__
There will be some students who can be characterized as minorities, athletes, legacies, etc., but who meet the regular standards of admission; that is, they need no "bumps" or other special considerations in order to admitted. These regularly qualified students are not properly thought of as preference admits, but of course - especially with respect to minority students who cannot "hide" their skin color - some will suffer stigmatization because of the understanding that admissions standards have been lowered for certain groups.
** Note, importantly, that this sort of aggregate data cannot tell the whole story, because they collapse particular “spikes” in any individual candidate into the aggregate. As I always say, let’s get the individual scores by candidate, suitably anonymized of course, and then it will be very easy to test the results statistically against a categorical race variable.
Then why so much argument and lawsuits about Harvard and the like? After all, if “innate ability always rises to the top” as you claim, then it should largely not matter where one goes to college (even if your college choice is limited by parental actions, or not being admitted to Harvard and the like for whatever reasons you may guess but not be sure about).
You are suspecting that the differences will remain large. I have my doubts about that, particularly after including SES and chosen major as adjustment factors. But I haven’t yet read the Duke study.
Agreed, @hebegebe, that I am assuming admissions decisions will remain differentiated by race even after SES/major adjustment factors. Certainly for black and (to a much lesser extent) Hispanic applicants.
There are summary statistics in the Duke study for SES (curriculum, family income, parental education). The overall pool of enrolled students is fairly privileged (even 70% of the black students had college educated parents, and that was the lowest figure). It’s worth noting that Asian students at Duke appear less privileged than whites on all SES measures (income, parental education, private school) and yet present higher observables and even some unobservables. More granular data would answer whether the adjustment factors you mention remove the apparent disparities (for instance, STEM majors on average present higher stats and Asians disproportionately choose STEM).
As a general matter, preferences for lower SES would likely benefit Asians, because there are many poor Asians with high stats. I’m not sure about majors. Don’t most of the “top” schools say they admit without regard to choice of major?
Yes, they admit students to “The College”, and most allow students to delay choosing a major until sophomore year, and freely allow changing majors as long as the student can still graduate.
But as far as I know, every college asks students about intended major (although it allows “Undecided”), and I have to believe that is to both evaluate the applicant in context, and help fit the class to available resources.
Even if a college does not formally admit by major or have departmental capacity limits, an admission reader may find the applications of the 10,000th prospective computer science major, the 10,000th prospective economics or finance major aiming for Wall Street, and the 10,000th pre-med biology major to be less interesting than the applications of the 10th prospective philosophy major, the 10th prospective astrophysics major, and the 10th prospective art history major.
^ Wouldn’t homogeneity of applicant interests be reflected in numerical scores assigned for variables like “essay” or “personal qualities”? We could test that.
Btw, how many “social justice” essays do these guys read from certain demographics? Let’s test that and find out.
If the rest of society is always and everywhere put under the microscope with regard to race, then why not admissions offices? Given their history, they should be accorded no presumptive deference. What’s good for the goose is good for the gander, as it were.
And in addition, much of the discussion about a particular applicant is not even recorded anywhere; therefore it is not retrievable. All of this is understandable to anyone with a sufficient quota of intelligence and critical thinking ability.
@MangoLover11223 your high school might have your official race reported on your transcript/secondary school report. You might want to check with your guidance counselor to see if that demographic info is documented and if it will be shared with admissions offices.
That’s actually a question I have often wondered. Let us know if you find anything out or if anyone knows for sure, I would love to know the answer!
No one is going to ask you to prove you’re Hispanic. No college is going to ask to see a family tree or a DNA test. It’s purely a honors system. The only category where colleges do some verification is for NA, where they may ask for tribal membership.
Do you think this is abused? I mean, there are many people that legitimately check some boxes that might not be evidenced by their appearance. For this reason, a lot of people might be able to get away with it.
Our public schools where I live have the parents fill out a demographic profile for each child attending the schools which includes questions about race/ethnicity. Also has other info like contact info, etc. I’m sure there is a reason they need this info, but do you think they are allowed to share this information with colleges if they question someone’s application?
Thanks for the feedback everyone! I know my question seems odd but that’s because although I am half-Hispanic I don’t “look it”. I’m of Spanish-origin but I don’t have any of the traits commonly associated with Spanish people. I have had some people randomly start talking to me in Spanish (in stores, etc.), but in general people assume I’m white (because of my other middle eastern half).
Perhaps too many white American students have some NA ancestry so it would defeat the purpose to let anyone with some say they are NA? That’s more or less the reason NA ancestry was exempted from the “one drop” legal rule that applied to black people in the US - too many powerful colonists were already descended from native people and were not going to allow themselves to be defined as “not-white”.
Is there any other reason that particular “check box” is singled out for verification by so many colleges?
Pretty sure you can claim NA ancestry without tribal membership. They will ask you about it and may just classify you as mixed race instead of Native American. As I recall, some press releases from Pomona indicate that a large percentage of their mixed race admitees claimed NA ancestry.
There is btw an entire massive CC thread, or series of threads, devoted to race and college admissions. Just google college confidential race and college admissions.
While colleges will not ask you to prove that you are Hispanic it may not help in any way if you don’t have any cultural traits generally displayed by Hispanics. An AO may remark that you don’t display any ‘flair’ which would be code that the school isn’t considering you Hispanic.
When I applied to college back in the Stone Age I was not asked to supply Tribal Membership info but that may have changed now.