MIT reinstates ACT/SAT test requirement

[roycroftmom]

Your guidance is invaluable!

[DroidsLookingFor]

@MITChris

I definitely wouldn’t have succeeded in the MIT undergraduate environment, with how I learn STEM subjects best.

Thanks so much for participating.

Questions about this part of your comment…how do you assess the answer to this question in an applicant? I mean I get that at some level it goes to the heart of the rationale for reinstating the testing requirement, but both more specifically and more generally - do you straight up ask applicants something about this? Are they sufficiently self-aware that they can give a useful response? What other parts of an application really demonstrate this?

[MITChris]

It’s a mix of both academic and noncognitive indicators. For example, students who are extremely motivated by getting As — being the top of the class is like a core part of their identity and what gets the best work out of them — tend to not do well in the MIT environment. We don’t have a valedictorian, we don’t have class rank, no one knows if they are better/worse than anyone else, and our grading distributions in classes look like this: MIT Admissions on Instagram: "you think you want to go to mit? here are the cutoffs for the 5.12, organic chemistry, midterm 🤠"

[DroidsLookingFor]

Love it, thanks

[LostInTheShuffle]

@GKUnion

What about students that don’t do well on standardized tests in general. My son doesn’t. He does do very well on exams he can actually study the material for. In terms of demonstrating applied knowledge, he’s exceptional.

The SAT has become a test that assesses understanding of subjects typically covered in HS courses (and for math, only up to Algebra II). If your S can demonstrate applied knowledge well, why do you think he didn’t do as well on the SAT (or ACT)?

[mtmind]

Thanks for your thoughtful reply, particularly your explanation of how MIT uses the scores and the “information” they provide. Again though, this analysis only seems pertinent to those students who choose to apply. I am curious about potentially qualified students who will see the test requirement, see the tippy-top scores of those usually admitted, and choose not to apply at all.

If students who have the potential to succeed at MIT aren’t applying to MIT because of the test requirement, then that is a loss for both MIT and these students, is it not? And even more so if these students come from under-represented communities that MIT might like to reach? Isn’t that one of the unintended consequences of the test requirement in a landscape where most of the other top schools are TO or TB?

---

On a separate note, I do appreciate you explaining (again) that MIT was never truly “Test Optional.” I think many have the misconception that MIT experimented with TO and that it was a failure, but that doesn’t seem to be the case at all.

Thanks again for taking the time to reply.

[ucbalumnus]

@MITChris

our grading distributions in classes look like this: Login • Instagram

Such grading scales are not unusual in college (not limited to MIT). They may seem very different for those used to typical high school grading (90% = A, 80% = B, 70% = C, 60% = D, or some such) which is based on having lots of easy stuff that high school C and D students can get correct. But a college instructor may include a higher percentage of A and B problems, rather than filling most of the exam with C problems.

[ucbalumnus]

@LostInTheShuffle

The SAT has become a test that assesses understanding of subjects typically covered in HS courses (and for math, only up to Algebra II).

SAT math has for a long time (back to the 1980s or before) been basically a multiple choice test on high school algebra and geometry, so good content knowledge of those subjects plus a bit of SAT test taking techniques should lead to good SAT math scores.

Of course, doing well on SAT math does not necessarily mean that one will do well on MIT math. But doing poorly on SAT math is a bad indicator for how well can do on MIT math.

[GKUnion]

I didn’t say he didn’t do well in a global sense, just not MIT well. Nor would either of my children apply to MIT, regardless of their miniscule chance of admission. As for his performance, he gets test anxiety over long, timed tests that can cover a broad spectrum of material. He’s never tested well like that. He’ll graduate Cum laude from his college of choice in May.

[bluebayou]

@ucbalumnus

SAT math has for a long time (back to the 1980s or before) been basically a multiple choice test on high school algebra and geometry, so good content knowledge of those subjects plus a bit of SAT test taking techniques should lead to good SAT math scores.

Agree partially and disagree (mostly). SAT math has historically differed from ACT math in that the SAT math built in reasoning ability. One needed raw processing power to do well (in addition to knowledge of Alg & Geom). One could spend 30 minutes on the classic, ‘if a train left Cleveland going west…’, and not solve it without good reasoning skills.

In contrast, ACT math problems were much more ‘straight forward’, but required speed. ACT also had (used to have?) 4 trig problems, two of which were always right triangles adn solved with the Pythagorean thm.

[ucbalumnus]

@bluebayou

Agree partially and disagree (mostly). SAT math has historically differed from ACT math in that the SAT math built in reasoning ability. One needed raw processing power to do well (in addition to knowledge of Alg & Geom). One could spend 30 minutes on the classic, ‘if a train left Cleveland going west…’, and not solve it without good reasoning skills.

Seems like those reasoning skills are taught in math (or they used to be). Those “word problems” that some students dreaded were presumably intended to get students to learn and practice applying math to various situations like the train leaving Cleveland or wherever.

[bluebayou]

Unfortunately, reasoning skills are in extremely short supply in HS. IMO, Geometry used to be a big teacher of reasoning, but doing full proofs have disappeared from our public high schools.

But I totally support what MIT is doing. If I was a top Eng/STEM school, I’d want SAT math, too. (preferably the old Math II subject test, but those have gone the way of the dodo bird).

btw: the ACT tried to build reasoning into its Science test…

[blossom]

I will let Chris respond with an expert’s view- but I have seen first hand (not my kid’s school) the outreach MIT did (many years ago, pre-Covid) in neighborhoods where going to college at all was not a given. The adcom’s who visited weren’t there to “sell” the college in the way that most “college fair” type set ups are done. They were there to describe the fun and whimsy and enchantment of science and engineering to audiences who had likely never known anyone who did “science” apart from a HS chem teacher.

I will hazard a guess that if a kid was motivated enough from one of these presentations to learn more about MIT, the hurdle over testing is one that the adcom can address. I will also hazard a guess that if a kid from the South Bronx or South side of Chicago shows a high degree of promise, MIT is going to do some homework on that kid (and the HS) before an auto-reject because there were no accompanying SAT scores.

I interviewed a kid for a job at my former company who was an MIT grad. He had been educated on an Indian Reservation, in a tribal school which does not send many kids to college- and had never sent a kid to MIT. No, he’s not typical. Yes, he was admitted and graduated. They all figured it out apparently.

[ucbalumnus]

@bluebayou

If I was a top Eng/STEM school, I’d want SAT math, too. (preferably the old Math II subject test, but those have gone the way of the dodo bird).

The Achievement / SAT Subject tests were the non-incumbent tests. So even though some colleges did find them to be better predictors of college performance than the SAT (or ACT), requiring them tended to be access limiters among students for whom Achievement / SAT Subject tests were not on the radar. So they became only “recommended” by many colleges… but as colleges dropped requiring them, their use fell to a point where the College Board dropped them.

[MITChris]

@mtmind

Thanks for your thoughtful reply, particularly your explanation of how MIT uses the scores and the “information” they provide. Again though, this analysis only seems pertinent to those students who choose to apply. I am curious about potentially qualified students who will see the test requirement, see the tippy-top scores of those usually admitted, and choose not to apply at all.

I agree with you. This is where our recruitment efforts — and particularly our partnerships with CBOs, Questbridge, etc — come in. As for the tippy top scores, I’ve been trying to explain that for 10+ years now (one must imagine Sisyphus happy) The Difficulty With Data | MIT Admissions

Isn’t that one of the unintended consequences of the test requirement in a landscape where most of the other top schools are TO or TB?

Possibly, yes. As we say in the post, when we stopped considering Subject Tests, it was largely because they had become so esoteric that they were too big a barrier to justify our usage of them given what else we could consider. Over the pandemic, we looked to see if we could get rid of the SATs, and decided we couldn’t, at least for now. But we’re data-driven about this, and as facts change, we’ll change.

[quote]
think many have the misconception that MIT experimented with TO and that it was a failure, but that doesn’t seem to be the case at all.[/quote]

Yes, it is a very unfortunate misreading of the scenario. I don’t know how we could have been clearer in our post(s) that this was about the pandemic, and that it was nothing about the admits from the last two years. So it goes.

[LostInTheShuffle]

I see. Thanks for clarifying, @GKUnion.

[Htas]

Actually it does. A kid who works until 9P, gets home at 10ish then does homework until 1-2 might be tired and have lower grades than a kid who gets home at 3PM. Ask me how I know? Not every kid has the same circumstances. Colleges do account for this. Always have.

[Htas]

Sorry, but I’d doubt a kid at that level could keep up at all. Would be interesting to see. But it seems highly unlikely that a kid who comes in at the bottom an keep up with the level of kid that is at MIT. There’s a certain amount of tenacity that helps a lot in learning but there’s also a skill set. Since the SAT teaches only the most rudimentary math as it relates to MIT, that score would mean there is a huge gap between that level and the level needed to survive in the first year. I don’t think there are any resources that can fill in that much gap. Might be wrong, but that’s a least 1 full year of math plus

[Aguadecoco]

Removed

[Data10]

@Aguadecoco

I envision a student that does well in his HS math class, is taking Calc senior yr and scored a 650-720 on the SAT math applying TO to a school like MIT. They like STEM and do well at school but they aren’t at the top of the applicant profile and would probably require some resources to keep up at a school like MIT.

@Htas

Since the SAT teaches only the most rudimentary math as it relates to MIT, that score would mean there is a huge gap between that level and the level needed to survive in the first year. I don’t think there are any resources that can fill in that much gap. Might be wrong, but that’s a least 1 full year of math plus

The SAT/ACT is a single test that emphasizes rapidly answering short and simple multiple choice algebra/geometry questions correctly. I’d expect MIT math to instead emphasize longer and more complex calculus+ questions that are not multiple choice. While there is no doubt a correlation between the 2, there are also no doubt exceptions of certain students doing relatively better on one than the other. It’s not a simple map such that a “650-720” math means you can’t handle MIT. There was a particularly easy math SAT test not long ago for which 3 careless errors would give a student a 720 math. That’s hardly definitive proof of there not being “any resources that can fill in that much gap” and requiring “1 full year of math plus.”

I think the more important question is if the kid who was not adequately prepared for MIT math would have been admitted, regardless of their SAT score. Would the 650-720 math kid who is too far behind to possibly catch up that you describe be admitted under a test optional system? Or might a kid who had not mastered basic algebra/geometry be likely to have something else on the application that was non-ideal besides just the SAT math score, such that they wouldn’t have been admitted as part of the incoming class?

This isn’t purely a theoretical discussion. Many colleges have been test optional prior to COVID, and they generally do not show decreases in freshman retention or graduation rate upon going test optional, including tech colleges. For example, I mentioned WPI earlier. WPI had a significant increase in both freshman retent8ion and graduation rate upon going test optional, without a decrease in portion choosing tech majors. I don’t know if they increased remedial classes, but I certainly would not make that assumption. Whatever they did do seems to be working for them.