@HSP2019 Sounds like we move in different circles. :)>- I’d really like to believe the world of elite business is not jocky (funny that autocorrects to a word that starts with a “c”). I doubt I’ll ever get close enough to find that out, though!
It depends on the college. For example, MIT has a long history with interviews being important to their admissions decisions. MIT’s website says interviews are strongly recommended with stats on the website about low acceptance rates for applicants who decline interviews. Similarly MIT’s CDS marks interviews are “important” – the same rating they give for GPA and test scores.
I doubt Harvard gives the same weight to interviews that MIT does, but I expect they do factor in to decisions. The Harvard rebuttal portion of the lawsuit mentions the following psuedo R^2 admission decision values for different parts of the application. The Interview rating is at least as correlated with the decision as most of the other application ratings, and applicants who do not interview have a much lower acceptance rate than those who do. However, this does not indicate how much weight the interview has in decisions. It may be correlated primarily because alumni are rating on similar criteria.
Personal Rating: 0.20
LOR Ratings: 0.19
Interview Rating: 0.13
Academic Rating: 0.09
EC Rating: 0.09
Athletic Rating: 0.08
My personal experience has been a remarkable correlation between my interview ratings and admissions decisions, such that I have never been especially surprised by an admission decision among students I have interviewed. I expect this relates to rating students on similar criteria to what the college does in their decisions (and what the college says on their website), particularly the ones that tend to be more variable among applicants… such as the discussed EC criteria and what positive character factors they indicate.
^This. Not enough people put themselves in Harvard’s shoes, it seems to me. I’m not an adcom, but the evidence would suggest the following about how the admissions sausage is made:
As Harvard sees it, it’s in its interest to be as powerful and influential in all important areas of society as it can be, because that’s how Harvard sustains itself and grows. Sure, as a university, its core activities are teaching and research, and those have to be maintained and strengthened continuously, both in absolute terms and to respond to competition. That requires money and access, which depend on producing alumni who are strongly engaged, successful by various measures, distributed in a variety of influential positions and willing to get involved to help Harvard and people associated with it.
Accordingly, Harvard’s admissions goals for the college have evolved to try to admit the mix of applicants that best serves those interests. In practice, that means they admit some number of kids (I estimate maybe a quarter of the class) who are the absolute cream of the academic crop, about 200 recruited athletes, a similar number of kids who fit the legacy definition, they want a class with something like 10-15% African-Americans and around 10% Latinos/as, they need to take care of (I’m guessing) a hundred or so kids on that Dean’s/Director’s list whose parents are sufficiently important that Harvard wants them to be associated with the school in this way, they want some number of kids from square states, the appropriate number of first-gens, some with special talents of some kind (which may fit an institutional need), a few faculty/staff kids, and then they fill in the rest with a few hundred kids who are holistically great.
Why this mix in particular?
The geniuses represent the core of the academic element of Harvard and its brand; Harvard has to have a critical mass of kids who are future cutting-edge academics or researchers, or succeed by being some of the smartest people in the world.
The athletes represent another element of Harvard’s brand (the student-athlete), the alumni are energized by winning teams, and a large number of the athletes, whether or not they turn pro or go to the Olympics, are going to be successful after college, and it benefits Harvard to have all kinds of successful alumni.
The admitted legacies generally present with academic stats better than the class average, are very well-prepared for college and their social capital is such that they’re likely going to be successful after college, which is also good for Harvard. Moreover, if alumni don’t think that being a legacy counts for anything, they’re going to give less money, volunteer less time and be less inclined to provide networking benefits to Harvard grads, all of which would be bad for Harvard. The key for Harvard is to admit enough of them that the legacy preference is seen to be meaningful and enough alumni are kept happy, but not so many that they have to compromise on quality or unacceptably reduce spots available in the other buckets.
It’s deemed necessary for a variety of reasons to have an appropriate number of underrepresented minorities, and to have an insufficient number would subject Harvard to a number of negative consequences. On the positive side, many of these students will become leaders in their communities and in a variety of other ways, and it will be valuable to Harvard that they’re linked to the school.
There are any number of ways one surmises students on the Dean’/Director’s list could be useful to Harvard. I imagine Malia Obama was on the list, as well as the daughter of the Chinese premier, who attended and graduated from Harvard under an assumed name. I’d guess there are plenty of super-powered development admits, etc.
Getting a kid from all 50 states has marketing value and encourages the highest-quality kids from those states to apply; a similar logic applies to first-gens.
Admitting some number of kids with a defining special talent ensures that when they distinguish themselves in their unique way, the Harvard name will be associated with their accomplishment, which boosts Harvard’s brand. In the meantime, the Harvard orchestra may need a virtuoso bassoonist, etc.
The faculty and staff need to feel that their kids (probably only a small number of whom apply every year, but may be disproportionately talented) have an advantage - that’s just good labor relations.
The holistically great kids will be holistically great when they graduate, accomplish many things that can’t easily be predicted, become successful in a variety of ways and give back to Harvard. Harvard will benefit from having educated them.
Obviously, in all cases (this is what @lookingforward talks about), Harvard looks at the whole package to see if it hangs together and to assess institutional fit before moving applicants to committee. They make small top-down adjustments toward the end of the process to get to a preferred mix, considering average scores, aggregate financial aid, geographic distribution and a bunch of other things, including racial balance (which the current lawsuit is all about).
I think that’s pretty much the way it works.
Of course there is! There are even second place trophies.
There are millions of high school players who never play in college. Most of the college players won’t go on to play professionally, but that doesn’t mean there wasn’t a benefit to them playing in college. Scholarships, exercise, travel. How about just for FUN? Should all musicians stop playing in college because they aren’t going pro? Should artists pack away their paint when it is clear they aren’t going to be paid? How about those pre-meds that never become doctors?
You clearly don’t get it.
There are a significant number of D1 non-ivies that give medical redshirts/hardships and honor scholarships. Notre Dame definitely does.
“But there are still a lot of poor students who wouldn’t be in college at all without athletics.”
That’s a stain on our society. How many poor students DON’T go to college because they’re incentivized to put too many eggs in the athletic-scholarship basket?
People in every other rich nation are stunned to learn that we distribute academic opportunity on this basis. Shame on us.
And in those other rich nations, do all poor students get to go to college?
Some D1 conferences (Big 10, Pac 12) now offer multi year scholarships. There are rules about keeping them, but things like injuries are protected. On my daughter’s teams, only one player lost her scholarship and was dismissed by the coach for off the field reasons during the 4 years. All other players had their scholarships renewed or increased during their time on the team.
And some coaches are just great human beings. DU had a hockey player who had a serious heart condition discovered in his freshman year. He hadn’t even played a season -or maybe even a game. The coach granted him a full scholarship for 4 years and he started working as a coach. Still there.
"People in every other rich nation are stunned to learn that we distribute academic opportunity on this basis. Shame on us. "
My husband is English and he still doesn’t really believe me when I explain this aspect of American college admissions, especially for elite schools. He gets that slight smile and patient look which I know is a signal that I’m saying something he thinks is complete BS but he doesn’t think is worth arguing about so he’s humoring some sort of crazy idea. It is totally inconceivable to him that athletics is at all part of the admissions process.
Can’t speak to the other rich nations but yes, in England poor kids do go to college. And since Oxford and Cambridge cost essentially the same as the least selective local college and all are at fairly low cost to UK residents, the poor kids don’t even have to chase merit by going to lesser institutions. And admissions to top unis are purely merit based, so no admissions even for development type cases; that’s why the royal kids went to St Andrews (they had the stats to get in there on their own) rather than any of the more prestigious unis. It’s not a perfect system, though. There are other aspects to their admissions that many Americans would hate, like the fact that substantially all results are based on a single, year end exam. High stakes, little chance for error and not very forgiving for smart kids who don’t test well for whatever reason.
“And in those other rich nations, do all poor students get to go to college?”
Of course not. But they don’t pick the ones who do on the basis of their being 7 feet tall.
Isn’t it better to have 10 basketball players get to go to Harvard who otherwise wouldn’t get to go to college at all?
If you watch the docu-series Last Chance U, you see that for some kids, sports really is the only way out of a life of poverty. Some of the players at the junior colleges are there because they were expelled from higher ranked schools. Some come from some money. Jim Kelly’s nephew Chad went there after her was expelled/suspended from a better school, and now is playing for the Broncos. Other guys are there because their grades were not high enough for a better school. They need constant tutoring about school and life. They know they are only in school to play football, but at least someone is paying attention to them and their education.
“Agreed. I’m still curious, though, about what need some of the athletes fill.”
Best explanation on this I’ve ever heard.
Colleges (even small time athletics schools like the Ivies and NESCACs with no money and no pro careers) give breaks to athletes because you keep score in sports. Even though no one really cares about Amherst basketball, Amherst has to be at least competitive. It can’t lose to Williams 100-20 every year.
If one school cuts its admit standards to compete, then the others have to as well. That’s the whole point of the Ivy League and its AI system – the schools all agree on the extent to which they are allowed to vary their admit standards.
They don’t do that for the student orchestra or glee club. If they sound a little worse than last year, there’s no scoreboard that tells the world that they stink.
Skimming through some additional lawsuit documents, I see that Harvard standard rating scale is 1-4 with +/-, and ratings of 5 to 6 are reserved for special situations. One of those special situations is students who have too little time for typical ECs due to home obligations like the quote above:
In Arcidiacono’s model with full controls, this EC rating of 5 was associated with a notable boost in admissions, more so than all athletes except for the 0.8% of applicants who who received a 1 in athletic (recruited athlete) and the 8% of applicants who received a 2 (among the strongest HS athlete applicants).
Having special meanings of 5-6 indicates that the ‘4’ rating is the worst possible standardl rating. The SOMF states the same about the 1-4 standard rating scale. It describes the academic rating criteria as quoted below. Among applicants who received the worst possible academic rating of ‘4’ in this criteria, 70% of recruited athletes were admitted, and 0.08% of the rest of applicants were admitted.
Harvard’s representative in the lawsuit created a model that was able to explain 64% of the variance in admission decisions based on ratings of applicants received in the categories above and others (not including overall rating), as well as various additional controls. The percentage of predictive ability lost by excluding ratings from one category were as follows. LORs and alumni were grouped together in this model. The earlier pseudo R^2 numbers I listed suggest LORs have ~1.5x the predictive ability of alumni interviews, which would suggest that alumni interviews had similar predictive importance in the model to personal rating and slightly more than academic rating.
Excluding LORs and Interview Rating: Lose 50% of predictive ability.
Excluding Personal Rating: Lose 19% of predictive ability
Excluding Academic Rating: Lose 17% of predictive ability
Excluding Extracurricular Rating: Lose 13% of predictive ability
Excluding Race: Lose 10% of predictive ability
The relatively strong importance of things like interview rating and personal rating compared to things like the academic rating in the model seems counter-intuitive . I suspect this result occurs in the model because there is more variation Only 0.5% of applicants receive the academic max 1 rating that often involves faculty review, as quoted above. The admitted unhooked group is primarily composed of the 51% of unhooked applicants that receive a 2. This huge group with the 2 academic rating likely has relatively small differences in academic stats, but has much larger differences in the categories that were more influential in the model. For example, only 21% of unhooked applicants received a strong 1-2 personal rating, and having a high personal rating was less correlated with academic stats (academic index) than any of the other ratings; so the personal rating becomes a powerful force for distinguishing among the 51% of unhooked applicants that received a 2 in academics.
northwesty #53, test scores are correlated with family income, but the released data stop short of Harvard score level. It is not clear whether the correlation continues all the way up to include the extremely well off, or whether it tapers off, or even reaches a maximum at some fairly high, yet not stratospheric income.
@Data10, this doesn’t surprise me at all. Harvard long ago said, they take about 200 of the brightest minds of their generations, but everyone else it’s a mix of factors. I also believe that most applicants have great grades and scores, but they aren’t doing the kind of academic work that clearly sets they way ahead of their peers.
“northwesty #53, test scores are correlated with family income, but the released data stop short of Harvard score level. It is not clear whether the correlation continues all the way up to include the extremely well off, or whether it tapers off, or even reaches a maximum at some fairly high, yet not stratospheric income.”
The whole highly selective admissions system advantages the advantaged. Even with big hooks and big financial aid for low income, first gen and URMs. The legacy pref is merely one part of that. Simply requiring very high academics is a huge selector for higher SES. Earmarking 15-ish percent of your seats to recruited athletes is another high SES selector.
In 2015 dollars, median family income at:
H $170k
P $186k
Y $193k
Brown $204k
If you eliminated legacy preferences, I bet those numbers would not move very much. Since you’d probably end up swapping out high SES legacy kids for quite similar high SES non-legacy kids.
@Data10 - related to this, I see that Card’s expert report (https://projects.iq.harvard.edu/files/diverse-education/files/expert_report_as_filed_d._mass._14-cv-14176_dckt_000419_033_filed_2018-06-15.pdf) notes the following on pages 28-30:
- Over 42% of applicants in the six years of data analyzed got an academic rating of "1" or "2".
- Applicants with at least one rating of "1" (academic, personal, athletic, extracurricular) are 48% to 88% likely to be admitted.
- Applicants with three or four ratings of "2" or better are 43% to 68% likely to be admitted.
I also understand the Arcidiacono report states that 33% of legacies are admitted (I’ll assume the Crimson has it right: https://www.thecrimson.com/article/2018/6/20/admissions-docs-legacy/).
Connecting the dots here, now we can make an educated guess about how legacy preference works in practice. If you’ve got “1”-level strength in any of the four areas, or “2”-level-or-better strength in any three of them, and you’re a legacy, you’re probably more likely than not to be admitted. If you’re a legacy, the odds are you’ve got bright and ambitious parents, who have provided you access to the kind of education and other opportunities that give you a much better shot than the average person to be a “2” or better on multiple dimensions - and then the legacy tip seals the deal.
Yes, I know we don’t have enough data to see how the extent to which these factors are interrelated, and correlation doesn’t equal causation, but…
I think it’s @northwesty who likes to say that the legacy effect is powerful because it’s a tiebreaker in a game with a lot of ties. If 42% of the applicant pool to Harvard is an academic “1” or “2”, that’s a lot of ties. Add “2”-level strength on one or two other dimensions, throw in the legacy tip and it’s a good bet that the tie is broken in your favor.
Well, they do have to beat Yale after all. You can’t do that if you don’t have athletes.
“they do have to beat Yale after all”
In football and crew. Nobody but swimming alums cares who wins that swim meet.