It depends upon one’s definition of serious boost.
As does a 2 in athletics.
In your opinion.
One’s reading of the Harvard lawsuit material does not make one an expert on Harvard admissions. At the end of the day, and IMO, there is no mathematical formula that will result in an acceptance.
The specific regression coefficients from the Harvard lawsuit are below (Arcidiacono analysis, full sample, full controls). Odds ratios are compared to a default of 3 rating.
Athletic = 1 (Recruited Athlete): 1800x odds ratio
EC = 1 (“Possible national-level achievement or professional experience”. Top 0.2% of apps): 39x odds ratio
EC = 2 (“Local or regional recognition; major accomplishment(s)”): 4.4x
Athletic = 2 (Possible walk on or other exceptional, few HS varsity athletes receive this): 3.9x
EC = 5 (“… family commitments or term-time work " prevent ECs): 2.4x
Athletic = 5 (”… family commitments or term-time work " prevent sports): 2.0x
Athletic = 4 (“Little or no interest”): 0.96x odds ratio (no significant impact)
EC = 4 (“Little or no participation”): 0.5x odds ratio
Isn’t the simple answer that dance, such as theater, is a major, whereas tennis or any other sport is not? So if you are applying as a dance major, your dance experience does give you a boost, or you are rewarded for it and it is considered in your admissions.
For example, my D was admitted to USC as a Theater major, but her stats were not in line with (were lower) being admitted to a non performing arts major. She got a “bump”, like an athlete, and was rewarded for her proficiency in her EC/major, much as athletes are rewarded for their proficiency in their sport.
@Data10 As you know, EC ranking not independent of Academic ranking - the two are strongly correlated, with an r square of 0.97 between Academic ranking decile and the proportion who were ranked 2 or lower on EC. Moreover, the correlation between acceptance rates based on Academic ranking decile and the EC rankings of each decile give an r square of 0.94. So it would seem that most of the variance in the correlation between EC ranking and acceptance rates can be explained by Academic ranking.
This means that you cannot use the odds ratio of the correlation coefficient of EC ranking in an multivariate regression for acceptance rates as an independent measurement of the effects of EC ranking. It is almost certainly based to a large extent on academic ranking. So the “boost” you report, of kids with EC rankings of 2 being accepted at rate that is 4.4X that of kids who rank lower on ECs is likely mostly because these high EC kids also have very high Academic rankings.
On the other hand, Athletic rankings are likely not strongly correlated with academic ranking, and as thus are much more likely to represent a boost that is not related to academics.
In any case, it still demonstrates that athletics, as a single extracurricular activity, is given the same weight as the entire set of non-athletic extracurricular activities together. If we go back to dance, a kid who spends 20 hours a week in dance, and edits the school newspaper will get a 4 in athletics, and a 2 in extra curricular activity. A soccer players who spends 20 hours playing soccer and is also a school newspaper editor will get a 2 in extracurricular AND a 2 or 3 in Athletics.
As everybody has agreed, the dancer matches the soccer player in all the aspects in which athletics is supposed to demonstrate a superior selection for a prospective student, yet these are only considered for the soccer player.
PS. The actual descriptions of the categories are slightly different than those you present.
Academic ranking was one of the control variables. In any case, regardless of whether you use odds ratio or regression coefficients directly, the conclusion is the same. Being a recruited athlete (1 athletic) is a huge boost that is far stronger than any other hook analyzed in the lawsuit, including any EC rating a dancer would receive. Receiving a 2 athletic rating is associated with a much smaller boost on a similar level to the far more common 2 EC rating. And otherwise athletic rating appears to have little influence for the fast majority of applicants. The Harvard OIR analysis had similar regression coefficients and similar conclusions.
The analysis suggests that athletics participation has little benefit beyond other types of ECs unless you get a 2 or are a recruited athlete, and few HS athletes receive 2s. 33% of applicants in the lawsuit listed their “primary EC” (first EC in list) as playing a sport, yet only 9% received a 2 in athletic. Most kids who spend 20 hours playing soccer are not going to get a high 1-2 athletic rating, and as such appear to have little admissions benefit for the time spent focusing on soccer over other types of ECs, such as dance or newspaper editor. I also wouldn’t be extremely literal in the “guidelines” mentioning a 2 for newspaper editor. Some newspaper editors might receive 3+ rather than 2-, while some dancers might receive a 2+ or even a 1, if achievement was at a high enough level. Less than 1/4 of applicants received a 1-2 EC rating, so it’s still fairly rare among an exceptional applicant pool.
It does seem like an odd way of rating applicants to have one of the 4 main categories dedicated to athletics like this, rather than separating recruited athletes as special applicants, and treating athletics like other types ECs for everyone else. There may be historical reasons, or a special dependence on admitting enough possible walk-ons due to Ivy League athletic conference uniqueness ( huge number of less popular sports, without scholarships or comparable 4-year commitment factors).
Most of my descriptions were direct quotes from the document you listed. The two that were not quotes were based on influence from other documents. For example, Harvard’s Statement of Material Facts more explicitly implies 1 is a recruited athlete saying, “a “1” athletic rating is used to refer to an athlete recruited by a Harvard varsity team.”
@Data10 When two variables are as collinear as EC rating and Academic ranking, it is almost impossible to tease them apart. Arcidiacono’s claims did not require that the effect of each be separated, and. In fact, he used this strong collinearity to support his claim that high ECs do not come at the expense of academics for Asians (or any group, for that matter).
I used “20 hours” because most youth soccer organizations post 20 hours a week as the maximum hours that a high school soccer player should spend training per week. Since training is generally done on school grounds, no extra time is spent back and forth. If 20 hours in the max, most will spend less, but time is also spent going to away games. So it’s not unreasonable to assume that a student who would get a 2-3 score will be putting an average of 20 hours a week into soccer.
"33% of applicants in the lawsuit listed their “primary EC” (first EC in list) as playing a sport, yet only 9% received a 2 in athletic. " A 3 in Athletic is still better than a 4 in Athletic. Moreover, they also seem to be “double-dipping” in that they are using their sport to boost both their ECs and to get a higher score in athletic.
The discussions in The Crimson and elsewhere implied that kids with an Athletic rating of 2 received a serious boost in application chances, but they were evidently wrong.
The rating method is not odd if you accept the claim that it’s used as a way to boost the number of upper middle class/wealthy white kids who are accepted, since the majority of the kids who are being accepted as recruited athletes are upper middle class/wealthy White kids. That’s the claim that Saahil Desai presents in his Article in The October 23rd edition of the Atlantic (Education section), titled “College Sports Are Affirmative Action for Rich White Students”. It seems reasonable, but I have not verified and tested all of Desai’s claims, so I cannot say for sure.
I posted the exact quotations because I think that they highlight the special status afforded to athletics, compared to other EC, and it seemed to me that your quotes were from a different section or document (I know that they were variously presented with slight rewording). You are correct though, that yours were quoted directly from the same source as my quote.
However, I did not with to imply that you misquoted the documents, either by accident or purposefully, and I apologize if it seemed to be the case. While I disagree with your interpretation here, I know that you do your best to present the most accurate data available across the different thread in which you participate, and do a good job of fact checking your claims.
The answer is simple. There’s no scoreboard for dance or singing. If the glee club is a little weak, they still sound great.
In contrast, if the hoops team loses every game by 50 points everyone easily can tell. So you can’t have a completely awful team.
So to be acceptably competitive, Harvard let’s in some ballers with lower sat scores. Now Yale hoops starts to lose. So they admit some ballers. Rinse and repeat.
Eventually the Ivy League comes up with the AI system to keep it within limits. None of that is needed for dancers.
Schools don’t have to have sports teams. But if they do, then this stuff is inevitable.
Also regarding title ix.
To make sports compliance easier, some schools tried to make dance/cheer a competitive college sport for girls.
That effort did not withstand legal challenges. But if dance/cheer had been able to make it to varsity sport and ncaa status, then it would have resulted in admission tips and scholarships like other sports.
Because, well, scoreboard.
You looked at the correlation between AI decile and percent EC 2+ & Academic 2+. This isn’t identical to looking at the specific academic and EC ratings of individual applicants, among other relevant controls, as was done in the analyses. There were 3 referenced analyses in the lawsuit – SFFA-Arcidiacono, Harvard-Card, and Harvard OIR. All 3 used different controls. The Harvard OIR was more notably different from the others and used a different set of years. In spite of these differences, all 3 models suggest a similar relative contribution between academic 2 and EC 2. Of course they don’t find identical regression coefficients, perhaps partially due to the limitation you suggest, among other model differences.
The referenced analysis found that there was not a statistically significant difference between 3 athletic and 4 athletic, in terms of influence in chance of admission . Yet a 4 in any ratings category besides athletic seems to be quite meaningful and in some cases seems to almost doom chance of admissions. For example, non-recruited athlete applicants with a 4 academic had only a 0.07% admit rate (recruited athlete admit rate with 4 academic had ~1000x higher admit rate). In the final 5 years of the sample, nobody was admitted with a 4 personal rating – hooked or not.
I agree that recruited athlete preference at Harvard and similar primarily benefits wealthy white kids. However, that would also be true if Harvard applied a recruited athlete hook, without dedicating one of the four core ratings categories to athletics, like Stanford and Duke do with their applicant rating categories. Why does Harvard need to differ from Stanford, Duke, and others by having a special ratings categories for athletics that doesn’t seem to distinguish between the ~90% of applicants who receive 3 or 4? And instead only seems to serve as a way to flag the ~9% applicants who receive a 2 in athletic that would not have been flagged as recruited athletes?
Harvard’s >40 teams are mostly in preppy less popular sports, where the recruited athletes are expected to be mostly wealthy and White kids. However, the kids who aren’t recruited athletes aren’t as limited by what sports are offered by Harvard and instead often play more popular sports, which tend to be less wealthy and less White (soccer, football, basketball, …). If the portion receiving 2’s primarily focused on excellence in athletics, regardless of sport, then it wouldn’t show the same degree of bias towards wealthy/White as recruited athletes. However, if the kids receiving 2’s are primarily limited to potential walk-ons, then it would again emphasize mostly wealthy/White sports For example, I walked on to the crew team at Stanford without any previous rowing experience… Walking on crew is also common at Harvard. This wouldn’t have been possible in more popular sports like football, basketball, or soccer. I’d expect a similar principle to apply at Harvard with walk ons not being practical in the most popular sports at Harvard.
I think it is largely because the NCAA does not have dance as a sport.
@Data10 Although I’m really enjoying the discussion (I mean REALLY), I’m worried that we’re delving into realms of statistics that others may find boring.
I’m not sure that I agree with you about similar contributions of ECs and Academics. I am now leaning towards thinking that these measure different aspects the same thing, at least in the context of high schools kids who are applying to elite colleges.
So ECs like math and robotics clubs and competitions, as well as various other academic “Olympiads”, essentially showcase the same academic skills as do the academics. Those would also be true of ECs that require writing and organizational skills. Moreover, in many of these ECs, leadership positions are most commonly bestowed on students who are doing well, and the better a student you are, the more likely you are to have a leadership position. Rank in the EC category is enhanced quite a bit by a leadership positions. So, once again, there is a strong connection between academics and ECs.
When we look at the other common types of ECs, related to music and fine arts, the correlation between music and academic achievement has been demonstrated a number of times, as has the correlation between an education in fine arts and academic achievement.
That’s not to say that these two measurements are replicating each other and only one of them should be used, but rather, ECs are good “supporting evidence” that the academic standing of a kid is not only a measurement of how good they are at testing (both standardized and in classrooms).
However, Athletics seems to be unrelated to either
Your analysis and discussion of the reasons and possible reasons for having “Athletics” as a separate major category sound pretty sound, and I have no better explanations or counter arguments.
“ECs are good “supporting evidence” that the academic standing of a kid is not only a measurement of how good they are at testing”
ECs of whatever kind are good supporting evidence that a student can spend 20-30 hours per week on something (often unrelated) outside their core academic classes and still keep up good grades and strong test scores. But that just demonstrates that almost all high school classwork and testing is incapable of distinguishing between truly exceptional students (top 0.1% and beyond) and hard working, very competent (top 1%) students.
I think much of the high apparent strong correlation between academics and ECs when looking to % 2’s ratings in different AI deciles relates to % 2 or better being an average over a huge sample, rather than individual applicants. If you you look at averages over a large sample rather than individuals, correlations with scores (2/3 of AI is based on test scores) often seem remarkably strong. For example, it’s not unusual to have a R^2 of 0.8 to 0.9 between average graduation rate of a college and average SAT score of a college, among a large list of colleges. One might think this means test scores can explain the vast majority of variance in a particular student’s chance of graduating. However, when one tries to instead predict chance of graduation for individual students with their scores rather than the average graduation rate for the full college, the results are completely different. R^2 often drops to <0.1, with some other factors typically being stronger predictors than scores, such as HS GPA.
The specific numbers mentioned in the Arcidiacono Harvard applicant analysis are quoted below for unhooked White applicants in the baseline sample. All ratings appear to have notable correlations with AI decile, yet AI alone only explained a small portion of variance in admissions decisions.
Portion of White Unhooked Applicants Receiving a Good 2 or Better Rating
Top AI Declie – 97% academic, 33% EC, 48% LOR, 30% personal, 63% interview
2nd AI Declie – 93% academic, 30% EC, 43% LOR, 28% personal, 57% interview
3rd AI Declie – 82% academic, 27% EC, 38% LOR, 26% personal, 49% interview
4th AI Declie – 68% academic, 27% EC, 33% LOR, 23% personal, 43% interview
5th AI Declie – 50% academic, 25% EC, 30% LOR, 23% personal, 38% interview
6th AI Declie – 26% academic, 24% EC, 25% LOR, 20% personal, 33% interview
7th AI Declie – 9% academic, 22% EC, 22% LOR, 19% personal, 26% interview
8th AI Declie – 2% academic, 20% EC, 17% LOR, 16% personal, 19% interview
9th AI Declie – 0.4% academic, 16% EC, 12% LOR, 13% personal, 13% interview
Last AI Declie – 0.1% academic, 11% EC, 7% LOR, 8% personal, 7% interview
I agree with many of the reasons you listed for the correlation. The academic type ECs have a more obvious reasons for the correlation. However, as summarized below, the most common “primary ECs” (EC listed first or 2nd on application) were athletics and community service. It seems less obvious that a high AI decile would have a strong correlation with athletics and community service success. There may be some secondary influences, such as the kids with higher AI deciles are more likely to be encouraged to pursue extreme community service activities and have opportunity to do so. Kids in a high AI decile are also more likely to have financial support from their parents, which is relevant for both athletic and community service success. There are also personality traits that are relevant to both athletic and academic success, such as dedication and tenacity.
Most Common “Primary” ECs Among White Applicants
MODERATOR’S NOTE:
Or, more to the point, off topic. Let’s move along from the analysis paralysis, please.
@MWolf s19 didn’t have time for any academic ECs because he’s a three season athlete, artist, and community leader. When I mentioned early on that he may want to do one of the academic teams like math club, he said “why do I need to show a school more academics? My academics are shown in my rigor and grades.” And I think he was right. So far, he’s five for five in admissions with seven more to go. I think the sports and art helped.
D21 will hopefully stand out with ballet vexing her main EC. But if she doesn’t then oh well. It’s what she loves to do!
agree with @BookLvr . After Title IX was passed in the 70’s mens and womens sports programs started to merge and like any two cultures merging, the dominate one got to make the rules. So sports that were important to men, became the most important to all. Womens Basketball and Softball teams grew significantly while gymnastics teams were dropped. There wasn’t a compatible football counterpart, so Womens Rowing, because of the numbers (you can have 60 people in a crew program) started to explode, as did double counting distance runners in cross country, indoor track and outdoor track, to balance out the numbers. There are some schools that pre-title IX had strong mens rowing and no womens rowing, and now have athletic scholarship womens rowing and club mens rowing. The other thing that happened at most large schools is the number of funded sports started to shrink as schools tried to have the minimum number to support in order to have the maximum fund for Mens Football and Basketball (Stanford still holding on to breadth and depth). Dancers will never be given scholarships or slots by NCAA Athletic Departments. Academic Departments might support dancers or actors as performer and reward them with admissions, but they won’t be provided with athletic department preference and they probably have to major in that department.
I would echo Data10 analysis that with the exception of RA or serious walk-ons the athletics in elite college admissions like plays very little role. My DD was also a three season athlete in HS and varsity player and her athletics ratings in Harvard admission were a 3 and 4, presumably having no impact on her acceptance.
The dancers I have seen had most success in elite college admissions are cheerleaders. Though they are not recruited every elite college needs about 5 freshman with that profile to field a 20 person cheerleader squad.
Dance is typically a major or a minor usually within the Fine Arts college. Some schools offer fine arts scholarships. It’s not considered a sport but rather an art form.
Cheerleaders are not dancers.
Most colleges that are emphasized on this site do not offer dance as a major or minor. Among the minority that offer dance, few students usually choose than major. At a good portion of selective colleges that offer dance, most years have 0 dance majors. According to IPEDS, only the following selective colleges (using 25th% ACT of >= 30 as proxy for “selective”) had >2 graduating dance majors in the most recent year. HYPSM… all had 0.
Barnard – 5 Dance Majors
Berkeley – 7 Dance Majors
Case Western – 3 Dance Majors
Michigan – 7 Dance Majors
Middlebury – 3 Dance Majors
Northwestern – 3 Dance Majors