Why is it impossible to admit STEM majors purely by merits? [implied: elite or highly selective / rejective]

Parents Forum
201

@MWolf,

I am sure your post took a lot of energy, but it’s unfortunately rather inaccurate.

It’s certainly true that MIT has long been focused upon fields that require a high level of math, and certainly true that Harvard has long emphasized a Liberal Arts education.

And yet, prior to 2012, Harvard and MIT matriculated roughly the same number of the very top high school math contest winners. If you look back at the history of the Putnam awards, there are more considerably more winners from Harvard than from MIT.

This all changed around 2012, when MIT decided that it simply wanted these USAMO/IMO winners, made an active effort to recruit them, and pretty much got all of them since then. And after that, almost all of the Putnam winners have been from MIT. What you are seeing is the result of a conscious recruiting decision.

It’s important to understand that what is merely “a trick” for Terence Tao is likely something that mere mortals cannot understand even if they spent a lifetime trying to prepare for it. There is a reason why so many very capable students get a 0 on these higher level exams, and it’s not due to lack of effort.

It’s pretty clear that you don’t understand the process here, as it’s only the very first set of exams (AMC 10/12) that are multiple choice. The next level (AIME) is fill in the blank. Everything after that (and there are roughly four levels between AIME and IMO) is proofs based. Roughly 250 students each qualify for USAMO (plus roughly another 250 for USAJMO) to take a 9 hour exam over two days to solve six problems, and their solutions are graded manually because, being proofs, there are multiple ways to get at the correct answer. Each problem is scored from 0 to 7, and by far the most common score for each problem is 0.

Bringing this back to a running analogy the best I can, doing well on the AMC 10/12 is the equivalent of being a great sprinter, but in this case needing fast-twitch math brain cells to solve 25 problems in 75 minutes (3 minutes per problem). AIME is 15 problems in 3 hours (12 minutes per problem), so perhaps it is the 800 meters. USAMO as I mentioned above is 6 problems in 9 hours (90 minutes per problem), so perhaps it’s the 3200m.

Math research is the equivalent of a marathon runner, and it favors the slow-twitch math brain cells that can look at a problem for days or weeks and come up with creative solutions.

My son was pretty good at math competitions but is better at math research. But he also happens to know a few people that have been on the US IMO team over the years, and some of them happen to be outstanding at math research as well. There are certain people who can be the Bo Jacksons when it comes to math, excelling at both competitions and research, but the key is that they have an athleticism for math.

4 Likes
202

Before a kid can excel in math competitions, they have to even know they exist. Most don’t. There’s roughly 15 million HS students in the US. Around 300,000 participate in AME competitions.

I’m not saying they aren’t cool, fun, or that kids without facility in math can excel. They can’t.

I’m saying it’s a skill that may or may not signal potential in higher level math research and that it isn’t a necessary badge to be very successful at a higher level.

203

Yes, we disagree. Very much so. In fact, while I can wax passionate about what I think universities in the world should and shouldn’t be doing as far es exams are concerned (I’m an equal opportunity mouther-offer), in this case I am passionate about misrepresentation.

You are presenting the Oxford and Cambridge admission process entirely based on the paradigm of holistic admission. It is not holistic. High stakes end of year exam
scores aren’t just used to make faculty involvement manageable. Faculty wants to see that the students can sit long, difficult exams and do well in them. They ascribe value to them and what they test for. They are not another data point. They are required. No exam scores, no interview.

Similarly, the entrance exams are required. While they are not like a French concours or the Chinese gaokao, ie they don’t just take the best from the top and afaik there is no fixed cutoff and the lowest score that makes it through may vary from year to year but they want to see you do well in them. No entrance examination, no interview.

Faculty involvement isn’t just “critical” in Oxford admissions. Faculty IS admissions. After having chosen the interview candidates on the basis of their scores in both school administered and university administered exams, the faculty holds the interview (about the subject, they are not interested in unrelated ECs or athletics and admits. Or doesn’t. There is no one else.

Once in a blue moon, some math genius will be admitted to Cambridge on the basis of their insanely difficult math entrance examination alone. But that’s just about the least holistic admission in the world.

1 Like
204

I never said Oxbridge admission process is “holistic” in the sense of US elite college admissions. “Holistic” admissions, perhaps by design, mean different things to different people. At one end of the spectrum, a college can look at an applicant “holistically” based entirely on academics: it may just want to take into account an applicant’s academic potential beyond a set of grades and scores. At the other end of the spectrum, it could include anything a college wants to include that year (and perhaps change what’s included the following year). As you said, few applicants are admitted to Oxbridge on exams alone. Faculty interviews or other types of deep faculty involvement will help ensure academic qualification and potential of those who are finally admitted.

205

Because the British system at both school and college level is built around “long, difficult exams” which test your knowledge of everything covered in one or two years of studying. Virtually nothing else matters, no participation grades, in most cases no continuous assessment either, just (in our case) four 3 hour exams over 2 days.

Revision and memorization is intrinsic to the system too, at high school and college level all classes stop a month before exams and you are expected to spend that time sitting at home learning everything 8 hours a day, 5 days a week, because your entire future depends on that.

When that’s the system then of course the top universities will want to see you excel in exams. If you don’t then you won’t be successful in your degree course. It isn’t easy for Americans who have no experience in doing that, compared to European and Asian students who are more familiar with high stakes, summative testing.

1 Like
206

What? To these schools full pay is 20-30K. So they have a budget and needs like any other school. To compare them to US schools is unwarranted. Their “FA” if you even want to call it that is totally different, mainly based on scholarships which are based on merit factors and less on income. Though there are “scholarships” for low income students it is nothing like the US system. Programs for low income students don’t work like the US where kids are getting $$ primarily based on parental income rather than “merit”.
Essentially, it would be very difficult if not impossible for a low income/middle income student from outside the EU to get enough money to attend Oxford/Cambridge, etc. Their parent is likely to be paying the entire freight. Whereas in the US, a low income student might get much more in the way of FA. Some US kids can bypass this FA quandary by attending a US school and going abroad where FA is covered.

Actually, in many/most cases full pay are paying 100% more. And to someone from a low/middle income family who receives no scholarship this can be significant. Just not apples to apples, it’s a totally different system.

207

This is true, and which is why I have previously said that kids that do well on these exams have demonstrated talent in competition math.

That’s really not the correct denominator. The stages are Darwinian by design, as the end goal is to identify the six kids that will represent the US in the IMO team. Therefore the AMC-8, meant for middle schoolers and below, is intended to expose as many students to competition math as possible, but also intended to identify those that are strong enough to do well in later competitions. Those that don’t do well on AMC-8 tend to stop and not take the AMC-10 or 12, but in a very real sense the AMC-8 did its job for those that attempted it.

I would estimate that roughly 20% of students have been exposed to the AMC exams either taking the exam for practice and deciding not to proceed further, or actually taking the exams and getting a score.

This confirms your point that there is a large majority that simply doesn’t know about these exams, and there are undoubtedly some very bright math kids in there. These students get distributed to many colleges across the USA, with a large percentage no doubt going to their state flagship. And after college, they likely shine wherever they end up working. So yes, people can successful careers from a large number of colleges.

But my point about the skew still remains. These exams identified some extremely strong students among the roughly 20% exposed to the exams, and a disproportionate number of them end up in a small number of colleges.

208

I’m not sure why math competitions are even relevant. Many kids try these competitions do very well and decide not to pursue them. It’s an event/EC/Competition like any other and kids need to spend their time on what interests them.
In the US, I don’t think they portend anything other than a specific interest in math contests and a skill in taking the tests and a likelihood they can do very high level math.

Since most US high school kids don’t have exposure to high level math or these competitions, you can’t really see the full picture.
I have a kid who did very well at an early age on these competitions and was “recruited” to a high level practice team. My kid thought it was boring and didn’t pursue. Many kids on that team (older) went on to very high level awards.

My kid is still very interested in math and doing advanced level things but those math competitions. Zero interest. Has found interests which are math related but not contests.

2 Likes
209

I have no idea. All I know is that people do prepare for IQ and improve their scores on IQ tests without gaining any new knowledge or understanding, just learning “tricks”. Ergo, it does not test anything innate.

No, it does not test for creativity or innovativeness, because they do not have the time or even the experts to check whether the person has solved the problem in an innovative way, not do they have the time or personnel to test the ability of contestants to find a a new probl

Exactly - MIT wants the best mathematicians, while Harvard is casting a wider net. So why were you making the claim that it’s because Harvard has lower standards?

Exactly - it is not testing for the skills and talents that a mathematician needs to do well in any of the many jobs in which they are engaged, nor in high level research.

Moreover, you still are assuming that math abilities = intelligence and that “AMC competition wins” = abilities to do everything.

I agree with you 100% on this one.

210

You probably haven’t looked at any problem on an IMO competition. Without creativity and innovativeness, not a single problem on an IMO competition can be solved. Some contestants may even come up with innovative ways to solve a problem that competition designers never thought of.

4 Likes
211

Yet another thread extolling the virtues of MIT because it successfully recruits math contest winners?

Seems a bit like extolling the virtues of the University of Alabama because it successfully recruits top linebackers.

1 Like
212

It successfully recruits national merit winners too.

1 Like
213

That is very true. Some questions even require that someone follow an unusual path ( creatively) to get to an unusual answer. My kids used to talk about this a lot after taking math competitions. They’d laugh about how it seems so strange/counter intuitive but that was the only way to solve it.
Most Teams review old tests and teach off that. But even that approach only teaches so much. You cannot memorize math beyond a certain point. You have to have the intuitive sense. That’s what makes someone a contest winner.

214

There are a wide variety of real world jobs for mathematicians which place a premium on being able to rapidly solve complex calculations in your head or require intuition about different ways to look at problems. Plenty of quantitative finance firms are eager to recruit people with the “skills and talents” to do well in high level math competitions.

1 Like
215

I still am waiting for somebody to explain why they are using “winning at Math competitions” as a primary and central indicator of the wide array of academic skills and abilities that one finds in academically talented kids.

Even if this is true, and I have your word versus the word of a person who is, arguably, one of the greatest mathematician of his generation, if not the greatest. I also have the word of my wife, who did pretty well in the Math Olympiads in the USSR, and also did a PhD in theoretical CS. But even if this is true, why should that be considered anything but as being an indicator that a person is good at math?

Since when are Harvard, Duke, Berkeley, etc., “Math Institutions”? Since when have skills and talent in math been the only, or even the primary factor in accepting students who are going to study biology, chemistry, Humanities, social sciences? Even for engineering the requirement is the ability to perform that high level math required to solve engineering issues. Beyond that, abilities in theoretical math do not help all that much.

Even for CS, only a specific set of math abilities are helpful, and a genius in other math areas will have no advantage in CS.

Yet people here are going on and on as though Winners in Math Competitions should be the top targets for recruitment by all universities and the kids who all colleges should be running to recruit.

Do people think that they would also be the best in research in Life Sciences? Maybe they think that Winners of Math competitions would beat out English majors in creative writing? Or maybe that they would be the best surgeons, lawyers, environmental scientists? Do people somehow think that winning a gold medal at AMC indicates a capability of knowing everything else?

In 2019, 37 of the 1,610 students who graduated from Harvard college were math majors. Another 91 were applied math, and another 171 were engineering majors. Even assuming that AMC competitions test for math abilities, this is only an indicator that they will excel at math, of which, as I stated, there were only 37 graduates.

So, there were 37 for which the AMC could be a great indicator that they have the skills and abilities required to succeed, and another 261 which require very high math levels, but also require abilities and skills for which the AMC certainly does not test.

A gold medal in the AMC would not provide an indication that any of the other 1,312 students would succeed in their choice of major. Yet people seem to think that these students should be accepted on the strength of winning at Math Competitions, or perhaps they believe that Harvard should not accept any students to any of these other majors and focus only on Math? I’m not really sure.

What I am sure is that higher level math is not relevant to 82% of the students attending Harvard, and is one of a set of skills required by another 16%, so that Harvard does not need more than some 30-50 admitted students who are winners at top math competitions.

There are far more USMO Gold winners, perfect scores in AMC, etc, than that. Since there are actually other ways by which to identify kids with amazing math skills, that means that math competition winners will not fill even the places that are reserved for math majors.

Not because Harvard somehow has lower standards (though they probably do), but because Harvard is not trying to fill its class with 1,600 mathematicians, or even 300 mathematicians. They have other skills and talents, including multiple academic skills and talents, that are not math talents.

So they will be rejecting a lot of AMC winners even if they are maintaining as high standards as they claim that they are.

5 Likes
216

Interesting point. I’ve observed considerable diversity at several of the top UK universities I am familiar with, all of which do not employ holistic admission.

That’s also interesting, and somewhat concerning.

Again interesting, and concerning!

I am not against looking beyond grades and test scores but I wonder how well AOs are really able to evaluate “(good) character” and whether their assessment really matter at the end of the day to achieve their institution’s educational mission.

Yes, the Oxbridge interviews for STEM majors do look for originality of thought/creativity, particularly when an applicant inevitably gets stuck on a problem.

Yes, a lot of international students apply to Oxbridge and, pre-Covid (and today), those who cannot travel to the UK for interviews can do them remotely.

(Too many posts to read this morning so I’ll stop here.)

3 Likes
217

I’m not sure anyone has said “winning at Math competitions” is “primary and central”. There’re more discussions of math competitions because they have a longer history, are better organized, and more widely available/accessible than other academic contests for pre-college students. Advanced HS students also tend to have much less exposure to sciences than to math. Moreover, nearly all sciences and other quantitative disciplines require different degrees of proficiency with math and a student who has demonstrated great facility in math is more likely to succeed in those other disciplines.

There’re other contests besides math competitions. Some of them are equally challenging and highly valued. Colleges may also have different preferences. Caltech, for example, may value IPhO medals more, because it views physics as fundamental to the understanding of all sciences and technologies. The sciences Olympiads also have “experimental” sections, where contestants would have to design experiments to solve scientific problems.

Relative popularity of math competitions may be tested by increasingly popular CS competitions (such as USACO/IOI), if the current trend continues. These CS competitions are also much more accessible (anyone with a computer and internet access can participate).

Winning any of these contests shouldn’t be viewed as a ticket to a tippy-top college. Instead, a student should pursue it only if s/he finds the challenge (a big challenge!) exciting and rewarding in its own right.

2 Likes
218

Oxbridge administers entrance tests that are sufficiently challenging that a perfect score is very rare and, in some subjects, where a 50% score would be enough to secure an interview invitation. This creates the dispersion needed to differentiate otherwise strong candidates (as indicated by grades, etc.). The number of students invited for interviews is typically 2.5-3.0x the available seats (they factor in SES when deciding interview invitations and ultimate offers).

My D22 applied for a STEM (not medicine) course at Manchester and Birmingham and both required interviews - but not as stress-inducing as her Oxford interviews. Agree, though, that UK universities generally have less resources.

2 Likes
219

I like the idea of tests which are purposefully difficult to ace. The downside to some schools is that they are testing the actual knowledge rather than potential. I understand that the system is different. For American students you have to: be advanced enough to have taken enough APs/ or IB ( if available, rare in our area) and have a fully formed interest and expertise in a specific subject. For our kid who was strong in every subject but one, this was a strange mismatch. Kids can still get in, but the hurdles are definitely different.
I like the idea of interviews. It allows for questions and being able to see how students think on their feet.
Yes, many European schools have less resources by far. The cost reflects this :slight_smile: In some case, we thought a particular course might have issues around funding ( research, size of group, etc).

220

Again, if you don’t know how these things are graded, you shouldn’t comment about them. As I explained earlier, everything from USAMO onwards are proofs, and these problems have many valid solutions. For example, there are easily dozens of valid proofs for the Pythagorean Theorem.

It’s not that the students get more credit for solving the problem in an innovative way, it’s that there can be innovative ways to solve the problem, and for solving it, they get credit. This also means that the testing organizations do in fact invest the personnel and time to check these answers carefully. They can do this because they have whittled down the participants to about 500 people total for USAMO and USAJMO. And it’s why earlier tests are either multiple choice (AMC 8/10/12) or fill in the blank (AIME).

That was a nice setup of a strawman, and I must say that you did a very good job attacking this strawman through the rest of your post. As the parent of a current Harvard student, I could quibble about other inaccuracies in the rest of your post re Harvard, but will refrain as this strawman you setup certainly doesn’t reflect what I said, nor anyone else as far as I can tell.

They way we got here this time is the way we often get here.

  • @eyemgh says yet again that all colleges are the same (don’t get me wrong @eyemgh, I like reading your posts, just as I like reading posts by @MWolf, but I just disagree with some of them).
  • I point out that they are not really all the same, and show some evidence to indicate that using an area I understand well.
  • Someone interprets that to suggest that I or others believe that these tests are the only things that are important in determining who should get admitted. And off we go.

I have repeatedly said that talented people can end up anywhere, and succeed from anywhere. On the flip side, I do believe that attending an elite college makes a significant difference in terms of certain opportunities.

I have an anecdote that I think reflects both situations. My D works at a major investment bank, and almost every one there in her cohort of analysts is from an elite college. Well everyone except one person from … the University of Alabama. My D reports this is one of the stronger analysts at her bank.

This is a person who likely could have gotten into many elite colleges but decided instead on the free ride at Alabama from being a National Merit Finalist. And furthermore got into an investment bank that does not recruit at Alabama, and therefore had to seek out those opportunities through extra effort. In contrast, the college where my D attended was a recruiting target, and there is in effect an ecosystem that makes it easier to get these opportunities.

To summarize, the highly selective employers (or highly selective groups within larger employers) that may only need to hire 100 people or less each year don’t recruit everywhere because it is not effective to do so. They focus their efforts where they think they will get the best yields, which is often the elite colleges.

It’s not that they will intentionally ignore talented people from other places, but the first difficulty is that people from other colleges have is that they are often not even aware these opportunities exist. I attended a state flagship for undergrad and had never heard of an investment bank until long after I graduated college. The second problem is that since they are not a recruiting target, they may not have the alumni base at the employer to tap into for advice. Third, they often don’t have the clubs at their college that teach the skills that are useful for interviewing.

5 Likes