That’s a great point. I was only thinking about somewhat selective schools. I bet the CSU’s can rely on GP, rigor, and knowledge of the school for instate applicants.
Also, AP scores could be more important, but I wonder how the online test fiasco will impact those scores.
@NearlyDone2024 If you look up the scores of gifted and highly gifted students you will see that “genius” isn’t tied to SES despite what some continue to push.
At HPYM, there ARE a large number of kids who are outliers by definition. And they aren’t getting the crumbs. They are entering these schools largely on the basis of being outstanding. Having to Compete in a pool where so many get knocked out means that their candicacy needs to be impeccable. They need those top test scores to give the school more confidence that podunk high’s GPA is valid.
If you run kids with the very highest scores ( the extreme outliers)you will find their rates of acceptance are not 5% they are much higher. Tons and thousands ( millions of kids with 4.0 ). These parents all think their kids are contrnders for the best schols but they aren’t. Kids with 1600 SAT, yeah not so many at all. If test scores don’t matter than why are so many kids NMF at these schools? And are you seriously going to say that a kid whose parent attended some great school isn’t likely at all to be more academic than the norm? You think all legacies are Rockefeller’s? Not true at all. And even 30 years ago wasn’t true.
So for every kid who was tutored since birth there is also a kid who reads at 2. These kids apply to HPYS also.
While there is indeed a correlation between average SAT/ACT and income, there remain many high scoring, low income applicants. You are also correct in that they do not apply to elite colleges at the same rate as high achieving wealthy students (“outliers”). That is a separate issue. For those that do apply, TO will not help them. That was my point with my hypothetical sample.
Those same students will have top end academic credentials otherwise, if they have a chance for admission. So, with or without test scores, they will have their top end academic credentials to show.
Also, this year at least, many of those from lower SES situations are less likely to be able to take the SAT/ACT at all. So even those who would have scored at the top of the range could benefit from test-optional, if they were unable to get to take the SAT/ACT at all.
“Curious about this comment. Do you think that lower income applicants are competing for different slots than full pay applicants? How do you see this working? Thanks.”
Yes - colleges, even the Harvards of the world have a budget so they cannot obviously have a class with everyone needing full financial aid. They have to prioritize who gets that aid, for sure athletes, children of staff maybe, followed by low SES, first-gen, URMs. Legacies and development of course would not get too much FA, as by definition a development kid’s family is rolling in wealth and legacies are typically wealthy as well. These are the kids where test optional will help.
ALDC are called out because they’re essentially different piles reviewed by adcoms or in the case of varsity blues, not reviewed by adcoms! By that reasoning say Harvard wants some hockey players that also have scholarship offers from BU or BC or Michigan, they have to match that with financial aid, and athletes will get that priority.
If you believe as I do that these colleges have soft but legal quotas for race, then other categories would have soft quotas - low income, pell grant, and so they’re competing with each other for those slots. It would be hard to think that Harvard would reject a full pay for a full need, in fact they would accept both so one would essentially fund the other, they wouldn’t have to dip into their budget.
“Also, AP scores could be more important, but I wonder how the online test fiasco will impact those scores”
I doubt CSUs or even UCs will use APs in admissions, the purpose of these colleges is to increase access, not limit it.
HYPSM type colleges consider many admission factors beyond GPA and test scores, and those many factors can be used to distinguish students from different HSs, beyond just looking at their GPA in isolation.
For example, Harvard’s expert in the lawsuit created a model that could explain 64% of variance in admission decisions based on a variety of factors including the ratings that admission readers gave the applicant in categories like academic, ECs, personal, LORs, etc. Test scores appeared to directly influence the academic rating, but appeared to have little direct influence on the other categories. When Harvard’s expert removed the academic rating from his model, he found that the variance explained dropped from 64% to 53%. A full summary is below:
Full Model – Explains 64% of variance
Full Model without EC Rating – Explains 55% of variance
Full Model without Academic Rating – Explains 53% of variance
Full Model without Personal Rating – Explains 52% of variance
Full Model without Teacher/Alumni LOR Ratings – Explains 32% of variance
The drop from explaining 64% of variance to 53% of variance was for the full academic rating which includes both transcript and test scores, as well as some other factors. I expect that if only test scores were removed, the drop would be far less significant than if both transcript + test scores are removed. For example, it might have dropped from explaining 64% of variance with test scores to 60% of variance without test scores. As such, I’d expect the overwhelming majority of admission decisions to in the referenced sample to remain the same, if test scores were not considered.
Nevertheless, the previously linked Harvard internal study analyzed a similar set of data and found that admit rate shot up as test scores increased. It found that higher SES kids had a ~2% admit rate with a 1200, ~6% admit rate with 1400, ~11% admit rate with ~1500, and ~29% admit rate with 1600. The admit rate shot up as SAT score increased, yet the analysis suggests that removing SAT score from consideration has relatively little impact on admission decisions. This is not a contradiction. Instead I expect the relationship between SAT and admit rate exists primarily because SAT score is correlated with most of the other application sections that are considered, rather than because small differences in SAT score are regularly changing admission decisions.
For example, LOR ratings were found to be especially influential on admission decision. As listed in the table above, dropping LORs/alumni cut explained variance by half. Harvard admission readers are instructed to rate LORs according based on the following reader guidelines. Kids who ace the SAT are not doubt far more likely to get the 1 LOR rating for “best of a career” or “one of the best in many years” that are kids who bomb the SAT. A similar statement could be made for most of the other sections. You rarely have kids who bomb the SAT, yet ace all the other sections of the application and would have been admits, if test optional. The rare few who do poorly on the SAT and ace the rest of the application probably are also likely to excel at Harvard. I’d also expect ALDC hooks to be more concentrated among high SAT kids, particularly legacy and children of faculty, which further influences increase in admit rate.
MIT’s website at explains this difference between correlation and causation for SAT scores in admission decisions far more eloquently than I can in the page at https://mitadmissions.org/blogs/entry/the-difficulty-with-data/ , which is quoted below:
Except for CPSLO, the CSU frosh admission process before COVID-19 was checking that the base admission requirements (courses meeting the a-g subject areas) were satisfied, recalculating HS GPA, calculating an eligibility index of GPA * 800 + SATRW + SATM (or an ACT-based formula), applying the minimum threshold (differs for in-state and out-of-state), applying any local area preference points, and ranking applicants by this eligibility index within majors and campuses.
It is likely that with CSU going test-blind this year due to accessibility problems with the SAT and ACT, that the admission method will just use the recalculated HS GPA instead of the eligibility index. https://www2.calstate.edu/csu-system/news/Pages/CSU-to-Suspend-Standardized-Testing-Requirement-for-Upcoming-Admission-Cycles.aspx indicates that some supplemental factors may be considered (presumably translated into GPA points) at each campus’ decision.
These decision are obviously not being made on sound, data-based facts.
“ a unanimous faculty-senate vote a few weeks earlier to retain the SAT for now, after a yearlong study by a task force found the test neither “racist” nor discriminatory and not an obstacle to minorities in any way.
The 228-page report, loaded with hundreds of displays of data from the UC’s various admissions departments, found that the SAT and a commonly used alternative test, ACT (also eliminated), helped increase black, Hispanic and Native-American enrollment at the system’s 10 campuses.
“To sum up,” the task force report determined, “the SAT allows many disadvantaged students to gain guaranteed admission to UC.”
So how could the liberal governing board of a major university system reject the imprimatur of its own liberal faculty researchers and kill a diversity accelerator in the name of the very diversity desired?
The answer: The urgency of political momentum against the tests proved irresistible and swept away the research and data.”
Note the “for now” in your quote. The full recommendation was for UC to cease consideration of existing standardized test scores in the future and instead develop a new assessment system. A quote from the summary recommendations is below. The schedule was accelerated due to COVID-19. Going test optional was implemented for similar reasons.
The case that going test optional hurts middle and lower class applicants goes something like this:
This view, at least in part, is shared from someone who literally a book on admissions.
“research suggested… that affluent students with suboptimal scores have a better chance of gaining admission to schools that would have normally been out of reach. Colleges love high-income students and they love them even more if they don’t come with mediocre tests scores that would drag down the institution’s standardized test averages.”
Per the June 29, 2020, “Getting Into Elite Colleges May Just Have Gotten Easier: Many elite colleges and universities are not requiring the submission of SAT or ACT scores for the 2021–2022 admission season.”
www.wealthmanagement.com/college-planning/getting-elite-colleges-may-just-have-gotten-easier (Lynn O’Shaughnessy is a nationally recognized higher-ed speaker, journalist and the author of The College Solution, an Amazon bestseller)
@NearlyDone2024 i know a bazillion full pay families at our high schools whose kids with high scores don’t get into a slew of top 20 schools. You are making it sound like being full pay is a ticket and it is not.
Kids from rural areas often do not present the AP-laden, highly competitive academic transcript for a simple reason: their high school may not offer these courses. I’ve worked and lived in rural areas. I’ve seen this first-hand. The SAT or ACT is the one shot these kids have to show they’re legitimate contenders. If we care about these “outlier” kids (and I do), we should care when there will be fewer slots for them to compete for.
@homerdog
Full pay is an advantage at most colleges among similarly qualified candidates, but it’s not a guarantee of admission. At the highest levels it often does come down to small advantages. Yet, many full pay applicants, as you point out, also get rejected. All can be true at the same time.
According to your theory about test optional favoring full pay applicants over lower income applicants, one would expect full pay kids to be overrepresented among test optional admits and lower income kids to be underrepresented. However, the opposite occurs…
For example, the previously linked review of est optional colleges at https://www.nacacnet.org/globalassets/documents/publications/research/defining-access-report-2018.pdf analyzed 21 test optional colleges. At all 21 of them, the enrollees who were admitted test optional had a lower average income than the enorllees who submitted test scores. When they looked at full pay kids specifically, test submitter enrollees had a higher rate of full pay kids than test optional enrollees.
The previously linked 25 years of test optional at Bates found the same thing. Test optional admits tended to be lower income than test submitters. The test optional review at Ithaca at https://www.ithaca.edu/ir/docs/testoptionalpaper2.pdf also found the same thing – test optional enrollees had a lower average income than test submitters at Ithaca. Among retained students, 30% of test optional kids were Pell eligible compared to 15% of test submitters – a huge difference. I could list many more. This pattern of test optional kids being more likely to be lower income than test submitter kids at a particular college occurs in every study or review I have ever seen on the subject (comparing test optional enrollees to test submitter enrollees at a particular college).
Do you have an example of an existing test optional college in mind where you believe the theory of overrepresntation of full pay kids and underrpresentation of low income kids would hold up?
https://www.thecollegesolution.com/who-is-benefiting-from-test-optional-admissions/
“Rich Students and Test-Optional”
“…Tony Bankston, who heads up admissions at Illinois Wesleyan University (a liberal arts college) and has followed the test-optional debate for years, suggested in an interview with me that the real beneficiaries of this test-optional policies are likely wealthy students.”
“‘Everybody is struggling with enrollment, and colleges are looking more and more for students who have the ability to pay a substantial portion of college,” Bankston said. Test-optional policies open the door to take wealthy students who would have been borderline applicants. “I think a lot of this is going on behind the scenes.’”
“I don’t think anyone is suggesting that underrepresented students haven’t benefited from test-optional policies, but the question is whether wealthy students and the institutions themselves are benefiting more.”
Those appear to be opinions, not facts. As previously mentioned, the referenced study evaluated something different from the claims (averages for test optional colleges vs averages for sample of test required colleges). Tony Bankston who made the atypical quotes appears to currently be the assistant athletic director at IWU and no longer work in admissions – see https://www.iwusports.com/staff-directory/tony-bankston/152 . IWU doesn’t seem to have been moved by Tony’s decade long criticisms of test optional. A few years after Tony left admissions, IWU reversed their stance and went test optional,. IWU seems to have decreased full pay kids and increased low income kids upon going test optional, as one would expect.
I believe IWU went test optional in the 2018-19 year since the percent of students submitting ACT dropped from >90% in previous years to 66% in 2018. I haven’t seen a breakdown of test optional vs test submitter at IWU. However, from the CDS I can review a comparison of FA before and after going test optional , which is summarized below. During the 3 years before going test optional, an average of 31% of students were full pay. During the 2 years after going test optional, an average of 23.5% of students were full pay. There was a significant decrease in full pay kids after going test optional, suggesting a similar relationship to other examples.
2015 (before test optional) – 33% of students determined to not require any FA (full pay)
2016 (before test optional) – 33% of students determined to not require any FA (full pay)
2017 (before test optional) – 27% of students determined to not require any FA (full pay)
2018 (after test optional) – 22% of students determined to not require any FA (full pay)
2019 (after test optional) – 25% of students determined to not require any FA (full pay)
@theloniusmonk: “Those same students will have top end academic credentials otherwise, if they have a chance for admission. So, with or without test scores, they will have their top end academic credentials to show.”
I didn’t actually post this, it was ucbalumnus.
“Kids from rural areas often do not present the AP-laden, highly competitive academic transcript for a simple reason: their high school may not offer these courses. I’ve worked and lived in rural areas. .”
It applies to urban areas as well, and I agree that standardized tests have been a way to help highlight those kids.
“Going test optional decreases the odds for a financially needy, high testing applicant who is now competing for fewer spots in any given cohort”
I think you’re making a connection between test optional and financial aid that’s not there, until we see how this year shakes out. If my budget dictates that half the class gets FA and other half is full pay, test optional won’t change that, or if it does, only at the margins. Most selective colleges have diversity goals, first-gen targets, athletes,even geography goals that requires FA. Pell grants are now in the US News rankings, so if anything you may see low income get a small boost, as data10 mentioned with the Harvard case.
@Data10
I am citing a 49 page study on how high achieving, low income high school students are disadvantaged in the admissions process at academically competitive colleges.
https://www.jkcf.org/wp-content/uploads/2018/06/JKCF_True_Merit_FULLReport.pdf
“…high-achieving, low-income students are less likely to have access to college level coursework than other students. This is one of the key factors in constructing most Academic Indexes. Low-income students are one third as likely to take Advanced Placement or International Baccalaureate (AP/IB) courses as other students. This is partially a function of attending rural and small schools, which are less likely to offer AP/IB programs, and partially a function of being steered away from higher level courses even when AP/IB programs are offered. The failure to take AP/IB courses means that a significant percentage of low-income students who could perform well in highly selective colleges will never even be considered because they have a low Academic Index score.”
“THE DECK IS STACKED AGAINST HIGH-ACHIEVING, LOW-INCOME STUDENTS
By the time all of the admissions preferences and processes (such as early decision and demonstrated interest) are added up, there is little room left for high-achieving, low-income students, nor do these students receive any preferential treatment of their own for having overcome the often steep barriers that result from growing up in poverty. In fact, analysis of 13 selective institutions’ admissions data suggests that while being an athlete, underrepresented minority, or legacy provides considerable advantage in college and university admissions, emerging from low-income families and being a top student provides no boost whatsoever (Figure 17).”
“We conclude that the preferences and some admissions processes in selective colleges and universities, taken together, have resulted in a surprising, and probably inadvertent, result:
Being admitted to a selective institution is actually harder for the high-achieving, low-income student than for others.”
“Many selective colleges and universities would dispute this conclusion. They would claim that they already provide a boost for economically disadvantaged students because their admissions process recognizes that merit should consider not just a student’s endpoint but the distance traveled from low-income high school to elite college or university. But what institutions say is very different than what they do, and when the data are examined, multiple researchers have found that being in the bottom inccome quartile (relative to the middle quartiles) has no positive effect on admissions.”
"In fact, being low-income may actually constitute a disadvantage in admissions. While the nature of the data does not yet provide a firm statistical basis to draw such a conclusion, anecdotal evidence is rife. When all these admissions preferences are summed up, the number of unhooked application slots is profoundly limited; estimated by one university chancellor as comprising only 40 percent of the available seats at Ivy League institutions. In the words of another college official, to be accepted at an elite institution an unhooked candidate “‘has to walk on water.’”
I agree @theloniusmonk that Pell grant recipients may now be getting a second look. At whose expense? Those just above the Pell grant level.
https://www.insidehighered.com/admissions/article/2019/01/28/study-pressure-enroll-more-pell-eligible-students-has-skewed-colleges
“…the colleges were admitting almost twice as many students with incomes just below the Pell threshold as their relevant pools would have predicted, while they were admitting far fewer (still needy) students with incomes above the Pell threshold.”
Once again, the non-hooked lower and middle class applicant find themselves competing for less of the pie.
I do appreciate the feedback from others like @data10. I’m going to look very carefully at the nacacnet study cited. My initial thought is that colleges admissions officers have a vested interest in portraying test optional in the most favorable light. But I’m open to reexamining this thought in light of compelling and objective evidence. I hope everyone feels the same way, defense mechanisms notwithstanding.
Another interesting point:
“While barriers to admissions often deter low-income students from applying to institutions of higher learning, the ones who do might still be at a disadvantage due to admissions processes within universities. For one, the ‘holistic review’ approach many schools employ takes weight away from test scores and GPAs, which may hurt low-income students who lack diverse extracurriculars due to the aforementioned ‘activity gap’. To be sure, there is indeed an income-correlation with standardized testing as well, but it’s far cheaper in terms of study resources to prepare for the SAT than it is to fund an expensive and involved extracurricular career.”
https://publicpolicy.wharton.upenn.edu/live/news/2302-impacts-of-lower-socioeconomic-status-on-college/for-students/blog/news.php
In making an argument against test-optional, the College Board provides some data that suggests that, among discrepent students (i.e. those where HSGPA and SAT score are significantly different from the expected correlation), those with higher SAT are more likely to be from advantaged demographics (>$100k parent income, Male, Graduate degree parent), while those with higher HSGPA are more likely to be from disadvantaged demographics (Black, Hispanic, <$70k parent income, AA/AS or less educated parent).
So there are not many disadvantaged students for whom SAT scores are their primary way of getting attention by colleges.
Nobody said low income kids are not disadvantaged. Low income kids often face many major obstacles beyond the ones typical high SES kids experience. My dispute was with the claims about test optional favoring full pay kids over low income kids, specifically with the implication that test optional admits were more likely to be high income and less likely to be low income than test submitters.
As ucbalumus touches on above, SAT scores are correlated with income. As income goes up, average SAT scores go up. As income goes down, average SAT scores go down. It’s far from a perfect correlation. There are some low income kids with great SAT and some high income kids with poor SAT, but the average SAT score is significantly higher for the high income kids. The reasons for this correlation are complex and multifaceted, with many different contributing factors.
In contrast GPA shows less correlation with income than SAT scores. For example, table 1 of the study at https://files.eric.ed.gov/fulltext/ED502858.pdf lists the following correlation of to family income among tens of thousands of UC students in Table 1. It’s not even close.
Correlation with Family Income: UC Students
SAT Verbal – 0.32
SAT Math – 0.24
HS GPA – 0.04
SAT is much more correlated with income than HS GPA. So SAT is much more likely to an abnormally weak point on application compared to GPA for low income kids. And SAT is more likely to be an abnormally strong point on the application for high income kids. A similar statement could be made about several other key admission considerations.
Given that SAT is more likely to be the weak point on application for low income kids, removing that weak points by going test optional tends to favor low income kids. There are individual outliers, including individual low income kids who have higher scores than predicted by GPA. However, these outliers are not good representations of the typical student. When you compare large groups at a test optional college, you consistently see that low income kids are overrepresented among the test optional enrollees, and high income kids are overrepresented among the test submitter enrollees. This relationship was shown in many previously linked studies, which I will not repeat again here.
A similar principle applies to URMs, first gen students, rural students, and others who are more likely to be lower income, and/or more likely to have SAT be an especially weak point on the application. These groups tend to be overrperesented among test optional enrollees at a particular college, although the relationship is often not as direct and consistent as for income in isolation.