Yes, agree. But it appears there are many who are using TO as the first step. I’d bet their next step is to eliminate tests entirely and push for test blind policies. This would be mainly on the basis of, if I can’t compete on an equitable footing then you shouldn’t have an edge either.
Not for applicants, but percentages for members of the entering class as listed in freshman survey are below. I used 2018 since 2019 and 2020 have a “prefer to not say” category. It looks like about half of entering legacies say they have a >$500k income.
2018 (No “Prefer not to say” category)
Over $500k Income – 46%
$250 to 500k Income – 22%
$125 to 250k Income – 19%
80 to $125k Income – 10%
40 to $80k Income – <1%
Under $40k Income – <1%
1 Like
Whereas the tests seem “too easy” for people who read CollegeConfidential, they are “too hard” for a vast majority of high school students, so that the mean has been artificially kept at 1010-1050 but the actual raw score has already been recentered twice since the test we (adults) took. I think we can subtract about 100 points compared to the pre-1995 version, so the actual average has become in the 900s.
This could be due to more kids going to college (lowering the median) and to more kids taking Algebra2, precalculus, or even calculus and/or AP Lang (increasing the number of students who reach a high level on the math section of the test or the English section of the test.)
2 Likes
Agree that this is how it should be.
However, external standardized tests are used when high school courses and grades are not consistent or trustworthy. Seems that international examples suggest that:
- Canada: No external standardized tests for domestic applicants. Apparently, high school courses and grades are trusted to be consistent across all schools in a province. High school grades may have a partial standardized exam component.
- UK: No external standardized tests for domestic applicants, but high school grades are entirely based on standardized final exams (O/A-levels).
- China, India: University admission is based entirely on external standardized tests; high school grades are untrusted.
- US: (before COVID-19) High school courses and grades are partially trusted, but not trusted enough that most selective universities still used some external standardized tests (SAT/ACT). But the forced (by COVID-19) experiment will let more universities’ institutional research departments determine how useful the SAT/ACT actually were in predicting college performance or anything else the college looked for in admissions.
Well, they don’t need to push. TO will make fewer and fewer kids take the tests each year. The economics of testing will eventually decide the fate of these tests.
3 Likes
So there’s a huge overlap between high income and legacy. No surprise. I would have guessed 180-300k mean.
This is definitely a self-selecting group. Most parents here have high achieving (or super high achieving) kids who are looking at tippy top colleges. That’s not at all representative of the US as a whole.
5 Likes
If you want to compare old and current SAT scores…
https://files.eric.ed.gov/fulltext/ED563025.pdf contains the score remapping for the 1995 recentering (no test redesign).
https://collegereadiness.collegeboard.org/pdf/higher-ed-brief-sat-concordance.pdf contains the old and new SAT conversion for the 2016 redesign.
I recognize that you meant this seriously, and the point is taken as a serious one.
But still, it was delivered in such a droll way that it makes me wish that the CC fora had an equivalent to the old alt.humor.best-of-usenet newsgroup.
I really dislike being quoted out of context and once again the part where I say colleges are honoring work and family responsibilities was left out. I am not going to post again- this thread is pretty much done anyway 
My own kid got into Harvard from a low income community, mediocre school and financial need. So I am well aware and appreciative of these efforts. However, it is going to take time for the rhetoric to become a full reality.
2 Likes
From your lips…
From what I’ve seen of those who replied, kids who do very well with no prep have a really solid K-12 education giving them that prep. If they attended a school that dropped circles from the Geometry program and read 4th grade level books in class during 8th grade, I doubt they’d have had the same results.
Kids at our school can truly be just as capable as students from any other school, but their SAT/ACT scores don’t necessarily reflect it. If they have a good GPA, it means they’ve hit the bar for that unweighted 3.9 or 4.0. One has no idea how much they could do. It’s not difficult for a top kid to learn circles as well as “whatever” in their next class when they are needed - though they need the drive (and info) to be willing to learn the foundation as well as the new material. I suspect one might have to fall back on LORs to figure out if the drive is there?
3 Likes
Based on what I’ve seen in the “industry”, not so. They’re really interested in gathering data and, if it proves that they can select students just as well without the tests, do away with them, either to diversify recruitment (elites) or to simplify the process, especially where it’s assembly-line (large publics). Based on echoes, elites aren’t that worried about TO. Test blind would be entirely different.
Large publics are a bit unsure, because their applicant pool isn’t largely self selecting and they use few criteria: no essay or just one (or a few short ones), no recommendations… Test scores provide a quick shortcut to admit the 50 or 60 or 70% qualified under your formula. Without it, you need to spend a bit more time processing the tens of thousands of apps. Often the basic formula was some form of GPAxrigor*scores, so they’re worried they’ll have to adjust, but right now they’re gathering data.
Another area where test optional is complicated: merit scholarships, in particular automatic merit scholarships. Since there are so many A’s, there can’t be automatic merit scholarships without a limiting criterion (which would be the test score). And it’s difficult to choose who should get merit scholarships at “run off the mill” universities where you don’t have applicants who cured cancer while playing at Carnegie Hall.
Test-blind at a large public system will be tested in California. If they pull it off, “designed by psychometrician” tests will slowly disappear, used only in instances such as a “very smart boy who was a chronic underachiever during freshman and sophomore years but turned things around and look, now he’s got a 1400!”
Blockquote
“Would you focus on who can answer the most simple multiple choice algebra/geometry/trig questions quickly without making careless errors?”
If the SAT Math or even Verbal were this simple, everybody would pretty much be getting over 700
Blockquote
It really is NOT that simple for the vast majority of HS students. They finish Algebra2 junior year, not always honors, take the SAT in May, miss a lot of questions that they don’t recognize (including anything involving stats, trig, and more complicated algebra/functions); they read the passages too slowly and don’t distinguish between all the answers because “they all look true”. They get tired because they’ve never taken a test that lasts 2 hours, let alone 3 or more, they lose focus, forget to turn the last page in the booklet, bubble in wrong.
Now those students would not apply to “top schools” to return to the threads’s topic, but they’re more representative of what HS students find easy or hard… the students most CC parents know (their kids, or kids in their area…) are not representative of what high school students find easy or hard- most find it easy, or what they find hard is breaking 1450 or higher. This skews “our” view. I tried looking at @ucbalumnus’ grids with the 2016 redesign/recenter and the 1995 recenter: if I’m not mistaken, a 1050 today is what a 850 used to be in 1990*, and the test had to be recentered so that 50% reached the 1010 threshold because too many students didn’t. In short, for a majority of US students, it’s not a simple test.
@michaeluwill explained upthread how the SAT was the hardest thing he’d ever done, he studied hard for it, yet on test day he didn’t think it was “this simple”.
*1050 today used to be 970 before 2016. 970 “recentered”, with a 430 math (student struggled in Algebra2) and 540 verbal (did okay with Verbal but failed analogies and synonyms…) to old non recentered pre 1995 scores of 390 and 460.
@Happytimes2001 you described friends and yourself that rose from low to lower middle income to upper middle to upper income Was grad school and merit based education the key to progression?
I’ll try to summarize - there seem to be three basic opinions here:
A. Academic ability can be quantified in a single number (or two), the number/s can be measured using standardized testing, and the standardized testing we have is fair and equitable
B. College-readiness can be quantified in a single number (or two), the number/s can be measured using standardized testing, and the standardized testing we have is fair but not equitable, because K-12 education is not equitable.
C. Standardized tests are crude tests of college-readiness at best, because standardized testing generally only tests the ability to do well on standardized test of the same type, and the ability to do well is affected by SES and cultural background, and therefore the tests are neither fair nor equitable nor effective.
Did I get it right?
Not at all. But I think everyone has already expressed lots of opinions so I’m out as the topic has been beaten to death, IMO.
3 Likes
I’ll summarize the opinions differently:
-
A. Standardized tests have some value. They help determine students’ readiness for college and therefore should be used as one of the measures for college admissions at most colleges.
-
B. Standardized tests have some value. However, they are unfairly influenced by accessibility issues and excessive test preps. Therefore, they’re unfair and shouldn’t be required for college admissions.
-
C. Standardized tests have little value. They only capture students’ ability in testing, and not much else. Therefore, they shouldn’t be used, or at least required, for college admissions.
-
D. Standardized tests have no value. Making students take tests is an antiquated way of measuring their readiness for college, or even how much they’ve learned.
8 Likes
That works too - a more nuanced view than mine.
Probably more like a combination of value/usefulness and fairness, both of which can vary based on the specific test and the specific college. Note that usefulness to the college may be based on things beyond student success in college.
Fair Partly Unfair
Fair
Useful 2+2 2+1 2+0
Partly Useful 1+2 1+1 1+0
Not Useful 0+2 0+1 0+0Someone who believes that standardized tests (or a particular standardized test) are not useful (the bottom row) or (not correctably) unfair (the right column) is likely to oppose standardized tests (or a particular standardized test). Someone who believes that standardized tests (or a particular standardized test) is useful and fair (the top left corner) is likely to favor standardized tests (or a particular standardized test). Someone in one of the other parts of the matrix may have an intermediate opinion.
4 Likes
Yes, making a sweeping statement about all standardized tests is unlikely to be accurate. Instead you need to define which standardized test you are using for what purpose and review that specific case. For example, using the AP Physics C score to estimate whether the applicant will be successful in freshman physics, a math placement score to select most appropriate first math course, and using the combined SAT score to predict chance of graduation at a highly selective college are not all the same thing. The 3 different situations will have very different degrees of predictive ability and different sets of limitations, external influences, and “fairness.”
It also is important to look at how much the test adds to that desired prediction beyond the other available information used by the admission system. For example, the “top colleges” that are mentioned in the subject of the the thread and first post use many different factors to evaluate the applicant as part of their holistic admission system – transcript including course rigor, which classes received high/low grades and how relevant they are to planned field of study, HS profile including grade distribution for that HS, essays, LORs, ECs/awards, interview, etc. The relevant comparison for test optional is what SAT score adds beyond the combination of the other factors used in evaluating the applicants.
2 Likes