MIT to no longer consider SAT subject tests in admissions decisions

But that's just my cynical educator's view.

However, my initial thought was that MIT maybe has so many perfect scores it's useful at this point and not really a differentiating factor.

For example they can afford to have a perfect 1:1 ratio of men to women for a tech school and I don't think they lower their standards for that.

On the contrary, I went to a highly ranked engineering school of the same limitations (limited other genres of degrees) and it was 30% females overall roughly, but actually in general it was more like 6%-8% females in the actual engineering degrees.

I worked in the admissions office one summer and was hard pressed to find a B on any Q4 report cards which was one of my summer tasks to ensure no early admission students entirely flunked out. I got to see other things like scores and in general I have to say even at a lesser well ranked school, and reading the averages of every incoming class, my notion is not that there is a difference between the scores of smart hard working lower middle class vs the upper rich as much as there is an over emphasis on the exam as a whole.

I agree objective results are very important, and I don't have an answer for a way to calibrate across schools. I think AP exam scores are better indicators but not all schools have AB or IB classes, etc.

It is really at the end of the day like most things about your commitment and ability to get something done regardless of how tangential you think it is to your core goals or study. In real life it is the same way there are lots of silly things people have to do all the time to get where they want to be and the weeding out is those who don't stop trying vs those who find a comfortable spot and stay there and blame everyone else for not having the right priorities.

Regardless, I still fill like even amongst high ranking SAT scorers, students are good at acing them now. My theory though I'm not confident is true is that studying for them has been systemized so much that it is actually difficult for MIT to tell above a certain threshold of score what the bell curve is.

This is all with the exception we need to realize that the SAT is not objective or perfect (the top comment this one is in response to) it was only less than five yrs ago they pulled sailing terminology from the reading and vocabulary portions after having to admit children growing up in the Bronx indeed might not have casually sailed not had sailing related literature in their curriculums, making this a bit unreasonable even for "guessing from context".

I had a lackluster performance on the new writing section the first time around or I wouldn’t have spent any time upping the original 1520 or whatever.

The point is to brag a little yes because I really never get to tell that story now that I’m 30.

The point is that rewarding me for skipping class to smoke weed with my friends because I aced some test turned out to produce a university student who spent a lot of time partying.

I’ve changed a lot since then, and I do also have the blessing of being a rather natural and amiable person to talk too, but there was a long time where by sheer luck I was getting glided along because I ace tests like those and AP tests, but it certainly not indicative of someone who was going to spend a lot of time producing some great research paper in college or anything.

So I always thought it was kind of bullshit and I nailed the shit out of that test.

Probably just an imperfect science all in all. Much like hiring tech workers.

Edit: small grammatical errors, on a phone

I understand that it might feel like the goalposts are being moved for people who optimize to score high on such metrics, but such is the nature of this type of games. That's also why search engine companies have to keep refining their algorithms.

However, the biggest problems with these tests (and the AP tests) is that they are expensive--they cost a non-trivial amount of money to sign up for and they cost a lot of time to prepare for if your school doesn't offer direct AP classes.

The expensiveness is the barrier.

I agree that the answer is the base SAT, but with an addition of a lottery past a point threshold. Luck (of birth) got us into this mess, and luck could get us out.

> In 2010 three College Board researchers analyzed data from more than 150,000 students who took the SAT, and they found that the demographics of the two “discrepant” groups differed substantially. The students with the inflated SAT scores were more likely to be white or Asian than the students in the deflated-SAT group, and they were much more likely to be male. Their families were also much better off. Compared with the students with the deflated SAT scores, the inflated-SAT students were more than twice as likely to have parents who earned more than $100,000 a year and more than twice as likely to have parents with graduate degrees. These were the students — the only students — who were getting an advantage in admissions from the SAT. And they were exactly the kind of students that Trinity was admitting in such large numbers in the years before Pérez arrived.

> By contrast, according to the College Board’s demographic analysis, students in the deflated-SAT group, the ones whose SAT scores were significantly lower than their high school grades would have predicted, were twice as likely to be black as students in the inflated-SAT group, nearly twice as likely to be female and almost three times as likely to be Hispanic. They were three times as likely as students in the inflated-SAT group to have parents who earned less than $30,000 a year, and they were almost three times as likely to have parents who hadn’t attended college. They were the students — the only students — whose college chances suffered when admissions offices considered the SAT in addition to high school grades.

Source: https://www.nytimes.com/interactive/2019/09/10/magazine/coll...

The article goes on to explain that while grade point average is relatively consistent across income level, SAT scores are skewed towards the rich. Schools have realized this, which is why many schools no longer require the SAT or ACT.

- ALL sections (and sub-sections) have questions that strictly increase in difficulty / projected "miss-rate" as time goes on. This is to keep test takers from coming back to answers they're unsure about but may themselves know how to solve -- so if you find yourself struggling with questions in a row, it's better to stop and go back rather than miss out on what you may already know trying to solve questions that you don't. For the reading section, the scale is scoped to each passage. For the vocab section, where there are 3 sections (vocab, grammar, and multiple-choice fill in the blank), the scale is scoped to each subsection. For the math section, it is scoped to the whole thing.

- The "Free Section" (e.g. the one that doesn't count toward your score, which instructors tell you before you start that section, so you can use it as a break if you wish) is usually section 4 or 5 of the test, to help plan your breaks. Some students, not previously-knowing or confused that the "free section" is ungraded, still take it thinking there must be a penalty of some sort.

- The word "equivocal" is tested within the SAT Vocab in around 60% of tests. Unequivocally, these questions have some of the highest wrong-rates of any question on the test.

- Within the grammar questions, Choice (e) "None of the above" is 99% of the time NEVER the answer. This is one of the most certain things on the test.

- The math questions will usually have (1) answer that is an outlier. 95% of the time, this is not the correct answer; (2) will be similar to the correct answer in different ways; and (1) will be the correct answer (e.g., say you're supposed to subtract "x" by 5 to get to the real answer. The obviously fake one might be multiplied by 5. One of the slightly-wrong answers might have 5 added rather than subtracted, another might just be off by 1). If you're ever in doubt, you can drastically increase your chances at guessing on a question by picking the question "most similar" to all of the others -- something like 65% chance, rather than 25% in the naive case.

- Again for math questions -- particularly the "word riddle" type ones -- the SAT will generally purposely pick questions that could have multiple seemingly-correct questions if you plug in 1, 2, 5, or 10 for the variables. 3 is almost always a safe bet, though I particularly liked to choose 7, because who thinks you'd ever choose to plug in 7.

- The essay is funny. Per the SAT's own published rules, they are not graded on fact at all; purely rhetoric, vocabulary choice, and clarity. All of the prompts also usually include a historical figure or event of some sort -- you don't need to know anything about them other than what the prompt tells you, but a well-known and easy way to win points with the graders is to make up a fake quote from someone adjacent to the event / historical figure as a hook: e.g. "Disconsolate upon hearing the tragedy of [EVENT X], [FIGURE Y]'s au pair journaled 'His life was short, but his memory will last forever'. Previously unknown to historians until then, Y's au pair embodied Y's belief that [SOMETHING FROM THE PROMPT]. [then THESIS STATEMENT on 3rd or 4th sentence, always]." (this is an objectively wrong and terrible sentence that I would hate to read in any other context. This is, however, similar to the SAT's example of a top-tier intro).

This is just the tip of the iceberg, too. It's a very predictable format and pattern (it has to be, as it's given multiple times during the same academic year; tests must be similar, lest one session of test takers do statistically significantly better than an equally-talented group which takes the test a month later)

So yes, while I believe the SAT does attempt to test for knowledge, it's that same pursuit of a bell curve that makes it easily gamifiable for those who know the test and not the material -- who are, once again, usually already the wealthy and connected.

There's a correlation between parental income and SAT scores.

“[S]tudents from families earning more than $200,000 a year average a combined score of 1,714, while students from families earning under $20,000 a year average a combined score of 1,326.”

https://www.cnbc.com/2019/10/03/rich-students-get-better-sat...

An "equalizing effect" would imply students of both parental income groups would have the same score.

I thought I had never heard of them, but after looking at wikipedia, it looks like they were called the SAT II when I was in high school. I remember that existing, but never knew why you would take them and never knew anybody that did.

It's not a stretch to imagine people being confused.

https://en.wikipedia.org/wiki/SAT_Subject_Tests

Yes, that's me.

> and who is comfortable making connections with others

And that will never be me.

Leadership is inherently nepotistic. You simply aren't permitted to lead if you have the wrong class background or the wrong ethnicity. (There's a reason all Senators look exactly the same.) You're literally just advocating for an old boys' club, and if that's what you want, you might as well be explicit about it.

But being the kind of articulate person college people appreciate is also a strong cultural marker that you come from the "college tribe".

What we could do is gauge students' communication ability in a way that is less subject to these factors. Which we already do, by examining students' writing.

The typical use of the word refers to manipulative people who will smoothly pretend values and opinions they don't have in order to mislead people and take advantage of them. Alternatively it is people who talk about things they know nothing about convincingly enough for those who also know nothing about the topic.

The two are much different things. People who communicate their values and opinions have massive comparative disadvantage against them.

I've never done well on a test when I didn't understand the material, and I've always done well on test when I did.

If there's any correlation with this and academic success, it's not necessarily a positive one. There's a grain of truth to the stereotype of "nerds" not having great people skills.

The whole world is a gigantic Animal Farm, also in the West. The apparatchiks have different functions and names, but the principle is the same. The system works better though.

But it is the closest thing anybody's ever come up with.

Is it simple jealousy of rich people? Or is it a worry that their kids aren't really the most capable so society won't function so well when they're eventually in charge? I'd say their kids probably are the most capable because elite universities get their funding and fame from successful graduates so they're incentivized to choose the people with greatest chance of success. Of a well-connected rich kid and a poorly-connected poor kid with the same SAT scores, the rich kid has better chance of future career success.

That would leave a lot more space open in schools for creative work. That includes both academic research and the humanities, which are less objective but nonetheless valuable. As has been pointed out before, during the current health crisis, people turn to the arts: TV, books, podcasts, video games, etc. These are things that are hard to learn with self-study and impossible to test for -- except for the test of whether people will want to consume them.

Even programmers eventually need at least some of this. A development team requires a lot of people trained at a purely vo-tech, objective level to be experts in the various technologies. But for a product to be successful, it also requires people who know what its users want, which is much harder to judge objectively, and harder to learn from a book.

I believe too many programmers go to university to learn what they could learn on their own, or at a much less expensive school that doesn't try give a broad education. We need a lot of those, and employers make a mistake in rejecting people who don't have that university degree. Worse, even among those with that university degree, they test them purely on their objective skills, and then later complain that they produce ugly interfaces and write terribly.

ETS isn't perfect (I think the amount they charge for GRE-related stuff is sickening) but it at least does a good job of making sure the best prep resources possible are cheap or free.

Aside from the fact that this is all off-topic since MIT's announcement is about the SAT subject tests, not the SAT itself, it is not a true monopoly anyway.

MIT's web site says they accept either the SAT or the ACT. ("We require the SAT or the ACT.") The SAT is from College Board, and the ACT is from ACT, Inc., which are two different organizations. Both tests are in wide use, so if anything it is a duopoly.

People want to pay thousands of dollars to try to gain access to elite institutions. No matter what test you'll come up with, someone will offer an expensive prep course for it, because they will have willing customers.

Is there really a single company called "College Prep"?

Probably the biggest way to stop the affluent from taking advantage of the system would be to eliminate school choice.

We do have this kind of system in place, I'm closely involved in it. The Olympiads in the US basically fill the vacuum of objective assessment at a high level. The entry round for each Olympiad is pretty similar in level to, e.g. what you would get in JEE-Advanced. But they're also far less popular...

Otherwise, for non-URMs, the top schools do agree with you. SAT (and school GPA) are usually used in threshold-y way with other factors mattering more.

At the very least, given an SAT score, I can know that somebody worked alone for 4 hours with a No. 2 pencil to get it... that's more than any other indicator can guarantee.

seems like as good a way as any to predict whether you'll be able to handle the rigor of 100-level courses in your freshman year. basically answers the question: "can you read and understand what we plan to assign?"

edit: out of curiosity I perused some of your other posts. it seems like you have a pretty good grasp of english vocab and sentence construction. I'm not really sure how you could have bombed the critical reading and writing sections.

Have kids take the test, mark the score. Do absolutely nothing. Kids take test again and naturally do better.

Be a SAT prep company, do the above but insert yourself between tests and take the credit (and the money).

I have yet to see an "abolish standardized tests" type suggest a workable alternative (or even an unworkable one).

This might be great for rich kids, though in case you are ambitious, clever, gritty but come from not-so-privileged background (not necessarily low-class or third world - just temporary economic turmoil is enough) - you are going to hate every single second you spend there with all your heart.

So maybe this way it is a win-win in the end. The aristocracy get their titles and the plebs their opportunities for self-realisation. No harm done.

If you're not including self study, then maybe? But with prep and self study, everyone I knew was able to get at least a 100 point increase, I think I got a 300 point increase. Basically aced the math section through practice and study (missed 1 question).

Just going through those books and drilling incessantly on weekends and in my free time was enough to get me a 1520 on the PSAT (the max score, which helped in gaining merit scholarships, which is the reason I go to school for free right now), and a 1540 on the SAT (99th percentile). I also used Khan Academy extensively, which provides free prep sponsored by Collegeboard.

In my opinion, prep courses are just a means for unmotivated students to put in the same amount of work as a highly diligent, self-studying student. There’s plenty of free and relatively inexpensive resort resources out there nowadays, it just requires time and attention (which sadly a lot of my generation doesn’t have).

But if you define the variable like this, you end up opening a can of worms.

Do rich kids have more scholastic potential because they have better access to resources? Or should scholastic potential be the measurement of how a child performs assuming they had the same access to resources? Should children who belong to a racial/ethnic group that have social stereotypes of having more scholastic potential be defined as having more scholastic potential because of the effect of stereotypes and labels, or should the measure of scholastic potential be of the underlying potential if the effects of stereotypes and labels were equal?

Normally measurements of potential aim to be to measure the factors least impacted by other factors, because if we account for those factors we are measuring actual instead of potential.

Maybe the problem is most students just don't give a damn about the classes their parents make them take? Can someone explain the findings in this paper?

The elephant in the room in these kinds of discussions is the validity of the tests to begin with. MIT makes a statement that the general tests are predictive in combination with other criteria, but how predictive?

Generally studies of these sorts of things find that they're moderately predictive of first-year GPA (like .4 correlation), and then trail off to zero as the interval between testing and outcomes increases. The effects are even smaller in studies in relatively unselected samples (due to court orders or legal decisions, for example).

So one argument always goes that you should use whatever empirically-supported stuff you can to make fair decisions, but we treat them as if they're more than they are. Sure, you could have a lot of things that are significantly predictive but with small effect size, or where there's lots of noise, but why as a society to we pretend these are huge effects?

The other thing about that paper is the hint that the practice effects are larger in higher-performing examinees, which also makes sense and is consistent with other studies. Why is this a problem? Because those are exactly the types of students for whom these issues are more applicable. A 12 point average gain with practice doesn't matter if you're including people who never had a chance at MIT anyway, but a 20 point gain might in a highly competitive group where small differences are being magnified tenfold.

The issue in all of this isn't the people in the 99th percentile versus the 50th percentile, which is the bulk of what's going into these predictive models and effect sizes, it's the 99th versus the 95th. There's a ton of real-world noise, but society acts as if the noise is nonexistent. It's like we're idolizing the outcome of some kind of survivorship bias process.

Is there any indication that they actually broke down any further, or is this just limited to the impact of average coaching and thus not a good representation of those who receive the top tier coaching?

And of course you wouldn't want to report to the college if a person has a disability, are you trying to scarlet letter everyone who needs a little more time to read from dyslexia?

Yes, this another big problem caused by (too much emphasis on) standardised testing. I learnt the most in classes that didn't do this. But people do better on tests when they do do this. Standardised tests are destroying our education system.

I get that college admissions is a contest and performative but hot damn this thread is full of people mad at other people's happiness.

It's a weird world where people are mad that those with means are using them to educate themselves and their children.

As an adult I choose to pay for lots of different forms of education that gives me "advantage" -- bleh -- over my friends, peers, coworkers, etc.. Why should I feel guilty that I have the means to buy culinary textbooks so I can bake better?

At the GRE level this is even more extreme. For competitive programs a perfect score doesn't distinguish you so taking the test can only harm you. When applying to grad school I was explicitly advised not to take the GRE subject tests.

It's not all benefit: you can be taught to be lazy, not to be self-reliant, you can lack tenacity and resilience, you can have perverse expectations and a sheltered view of the World. Such things can turn out to be a hindrance to self-fulfillment.

> predict performance about equally well

having no tires or no engine predicts performance of a car--or rather the lack thereof--equally well. Yet grip and horsepower are independent variables. I think that means SAT scores don't predict success too well at all beyond a certain threshold.

Having the right motivation (haha, a pun) for a certain disciplin might make a huge difference. So you can test e.g. vocabulary learning in general, or top9cal knowledge, which requires precise choices of vocabulary nonetheless, but one not found in a general dictionary. It's more like knowing which dictionaries exist, and what texts are referenced therein.

This is point where you take a step back and conclude that if this is your measured result, you may have been measuring nothing at all.

If I was an investigative journalist and or I had significant amount of time, then I would look into it.

Much truth is accessible just by looking.

It's fairly easy to spot the "heavy-hitters" just from recommendations, undergraduate publication record, thesis, and GPA. In short, MIT and its elite cohort (CMU, Caltech, etc.) only accept undergrads who found their way into their professor's research groups -- this feat in and of itself is a superior indicator.