Accommodations are intended to level the playing field, and if your child doesn’t need them you should count your blessings.
It is not as easy to get accommodations as you think (exceptions do occur).
7 Likes
My daughter took about 5 practice tests in the few weeks leading up to her first and only SAT last year. Not every kid intensively preps these days…
1 Like
You can view Caltech’s placement tests online. The test for math 1A contains 5 questions, most of which have multiple parts. The questions look like they are what I’d expect for a Caltech class final… more advanced than AP calc or calc at the vast majority of colleges, but not so advanced that it’s unlikely for anyone to place out . Looking at the tests, I think I could pass the math 3 placement exam today with no prep even though I haven’t taken a college math class in many years, but I’d have no chance on the calc placement tests.
I don’t know the actual percentages of students who place out, but it’s enough that there are threads in which students who placed out talk about their best options and enough for several pages on Caltech’s website and documentation describing what math courses students who place out should take during freshman year.
The public colleges with the most applications almost always have some form of “holistic” application that considers more than just GPA and SAT. This includes a few colleges that have over 100k applications per year. If you have enough readers, it is possible to read any number of applications.
Rather than the colleges that have the most applications, the colleges that go by GPA and SAT alone (for all students, not just those who auto admit top x%) tend to be less selective ones that admit the majority of applicants. For example, if a college admits 80% of applicants, they can’t be as picky about each applicant having to excel in a wide variety of criteria. Such colleges tend to admit a good portion of students who are not well qualified and have a low graduation rate.
Each year the NACAC asks a large number of colleges questions about their admission system including which factors are considered and which factors are important. They weight the responses in an attempt to reflect the overall national distribution. In the most recent report I saw onlinem 86% of colleges said they considered GC LOR, 83% considered teacher LOR, 75% considered essays, etc. Few colleges just seemed to admit by a calculation based on GPA and SAT alone, and those colleges are generally less selective or non-selective. I can list numerous examples that show my point, but I imagine you are picturing a different type of college. Do you have a particular college in mind to use an example?
I haven’t looked at Caltech’s placement exams for years, so maybe they’ve gotten easier. Can you provide a link to the sample tests? In any event, I believe the placement tests have to be taken sequentially (in the same sequence as the courses are to be taken). A student can only take a particular placement test after s/he has successfully passed all the previous tests. For example, a student can’t take Math 3 placement test, unless s/he has already passed Math 1a, Math 1b, Math 1c and Math 2 tests.
For the public universities, I had in mind those that admit in-state residents based, at least partially, on some combination of course grades and standardized test scores.
However, I would not be surprised if the stats-only admission colleges (like the CSUs in California) were larger than most colleges, so it may be that the percentage of students attending colleges where LoRs, essays, etc. are considered is less than the percentage of colleges which consider those things in admission.
Also, some colleges may be stats-only for a portion of the admits, but add in the additional factors for applicants closer to the “borderline”. Texas public universities are examples of such, but there may be others that do not publicize such a methodology.
You can find Caltech’s placement exam page in a Google search. You need to register and login to Caltech’s website to view the exams, so it’s probably not a good idea to link on a non-Caltech forum. The placement exam test page states the following. They specifically say that the exams do not need to be taken in a particular order, and that you can pick and choose any combination of the 5 exams.
"You do not need to take the placement exams for Math 1a, 1b, 1c, 2 and 3 in any specific order. Therefore, you have the option to take any or all of the exams, as you’d like. "
Thanks for the correction. I mixed it up with physics placement exams. The physics placement exams have to be taken in sequence:
Ok the Cal Tech opting out seems to be inconsistent with the all students have to take the core first year, unless they expand the definition of core, which wouldn’t make much sense.
I don’t know where you’re getting the MIT 20% and 30% numbers from, including both ACT and SAT math, 70% (17/20K) scored above 34 or 750.
It is not a student opting out, but a student proving to the school that the student knows the material in the course, so the student and the school need not waste time teaching it again to the student. Caltech also has variants of Ma 1a for students of different levels of preparation, so that students who do that Ma 1a do not all take the exact same variant.
An analogous situation exists at colleges whose core curriculum includes a foreign language requirement. They typically allow students to start in a foreign language course more advanced than the beginner course if they have previous knowledge of the foreign language as demonstrated in a placement test.
Caltech clarifies on various pages of their website such as https://www.admissions.caltech.edu/apply/first-year-freshman-applicants/academic-preparation , in which they say, “Ph1 is an Institute requirement for the Core Curriculum, so you must either take Ph1abc, or place out of the course through taking the exam.”
I’m getting them from the admission statistics page of MIT’s website at Admissions statistics | MIT Admissions . A more detailed breakdown for ACT math is below:
34-36 ACT – 5,551 applicants
31-33 ACT – 1,293 applicants
28-30 ACT – 630 applicants
25-27 ACT – 304 applicants
Under 25 ACT – 114 applicants
Total = 5551 + 1293 +630 + 304 +114 = 7892
Fraction <34 = (1293 +630 + 304 +114) / 7892 = 30% of applicants who submitted ACT.
Your total comes out differently because you are assuming that students either submit ACT or SAT – never both (ACT math total + SAT math total)/ Total applicants. It’s quite common for high scoring students to submit both exams, so you are double counting a significant portion of students.
1 Like
@Data10 your math may be slightly there, MIT received total applications of approx 20K, they received total 22K scores, so the overlap is not a lot, only 2K scores or 1K applicants if you assume two scores per applicant. About 80% of SATM submitted were 750 or over and 70% for ACTM 34 or above. You have to assume that there are some kids that also submitted a 770 and a 35, so looking at the proportion that way, there may be max, 250 who submitted below 750/34, so about 25% of the applicants.
Your math may be slightly as well.
17/20 = 85%, not 70%
Suppose 1k applicants submitted 2 scores as you claim, and the other 19k submitted one score. The total number of scores would be (19k)*1 + (1k)*2 = 21k. The total would not be the actual 22k scores.
Yes, that was what I wrote in the original post – 20% <= 740 and 30% <= 33. If you want to estimate for either SAT or ACT, then one might roughly estimate roughly 25%. The larger fraction of SAT to ACT pulls the average down. And kids who score well are far more likely to submit both SAT and ACT than are kids who score poorly, which pulls the average up.
2 Likes
How valid is this assertion? A kid who achieved a perfect, or near perfect, score on one of the standardized test (whichever s/he happened to take first) would be unlikely to take, let alone submit, the other test. Why should s/he, unless s/he wanted to leave the impression that s/he overemphasized and/or overprepared for these tests (and perhaps also over-tested)?
2 Likes
Avoiding the bottom 30% of MIT students on one of the 4 ACT sections is not the same as getting a near perfect ACT composite score on first try.
If a kid with 2 middle/high scores (by MIT standards) submits both, it shows that he/she has consistently good scores. If a kid with 2 low scores (by MIT standards) submits both, it shows that he/she has consistently poor scores. I’d expect that the applicant is more likely to showcase consistently good scores than consistently poor scores.
2 Likes
Seems true that someone who scored very high would unlikely take both the SAT and ACT. But, there must be kids who sign up for and take both about the same time (maybe signed up hoping to have a couple of options). I don’t think it looks that bad if they have two scores.
Then again, seems like once and done is the best way to go. But in 2020, anything goes.
I can’t imagine kids intentionally trying to send in two scores to “doubly impress” seems a bit odd.
While some kids would think it’s doubly impressive, I think the bulk just want to showcase what they’ve got. (Or at least, in the past.)
Where it helps to send both is when one test number is off but on the other test is higher. Eg, a 720 M but a 35 ACT M. Maybe they don’t omit the SAT because the CR score was sweet.
And with many kids taking one test or the other earlier in hs, you just can’t count on one and done.
1 Like
@Happytimes2001 I agree that some students may have signed up for both tests and took them in two adjacent months. If they couldn’t decide which one to submit, they may submit both. This type of situations should be obvious to AOs as applicants also had to supply the test dates for those two tests.
@lookingforward The case of kids receiving inconsistent sectional scores between the two tests is interesting. How should a school like MIT look at it? Should AOs there look at them favorably (these applicants managed to at least score higher on one type of test), or unfavorably (these applicants had to rely on a different type of test to remedy a potential deficiency instead of relying on superscoring from the same type of test)?
I’ve heard that too. Send both if you have a section that’s low on one but high on the other. But what does that say about these tests? What kind of math student does decidedly better on one test than other and how is that even interpreted? That’s another reason why I think these tests leave something to be desired.
S19 got a 1540 (800math/740 EW) on his first real SAT after a summer of studying and taking practice tests. Could never break 31 on a practice ACT (although admittedly only took two practice tests before throwing in the towel). He thought the ACT was a cinch and could get all questions correct when he gave himself more time but wasn’t interested in trying to come up with a strategy to go faster. How is his test taking experience interpreted exactly?
D21 also could hit a 35 given all of the time in the world. We had her do that just to see how she felt about the actual content. Speeding up was a problem and she couldn’t get close to that. What does that say about her? She knows the content but can’t read fast enough or doesn’t use tricks to go faster? Her SAT practice tests were better within the time limit but, after studying on and off for months and four cancellations, just didn’t score as high as she did on practice tests when masked up and stressed about how late in the process her two sittings were. She’s going TO everywhere because her scores do not lift up her app and don’t match her GPA and rigor.
1 Like
I agree that the tests leave a lot to be desired, particularly when used as a part of MIT admissions. As I’ve mentioned in this thread, just 1 careless error on the SAT could be enough to put you below MIT’s 25th percentile in some years (raw score conversion varies by year). 3 careless errors could drop you as low as 720. However, on ACT math, you can usually make 1 careless error and still get a perfect score. The most relevant difference is probably time. On the ACT you have to race through and average 1 question per minute to get a MIT-level score. In contrast, the SAT gives you 33% more time for fewer questions.
Of course neither test in any way resembles the format, level, or content of questions a MIT student would see in math course. MIT level math isn’t about answering simple multiple choice questions as fast as possible, without making any careless errors. If an applicant has MIT level math scores on one test and not the other, I don’t think you can reliably draw much of any conclusions.
1 Like
The SAT M or SAT Math 2 subject test does very little, if anything to differentiate kids that get into MIT or Cal Tech, HMU, they use other tests like AMC/AIME and national or international awards, like ISEF, Olympiads to figure out the academic part. That’s why these colleges have a section for AMC scores, Stanford may as well. The verbal maybe used more, but again, most of the applicants are submitting 750/34 or higher.
1 Like