My kid is doing pretty badly on the math section of the SAT /ACT, even though she does well on other sections, and is doing well in math class.
IDK what's up. We've hired a math coach, like good liberal parents, but it's not helping. My current theory is that the math exams are witched.
Is it an SAT/ACT specific math coach, or one who teaches math generally? If it's the latter it might be less helpful if the problem for the kid isn't math overall but the very specific SAT/ACT-Math that they're testing, which is definitely its own thing. Someone who is used to coaching students for the standardized tests will also come with a whole set of useful suggestions regarding how to approach questions and stuff like that* that you won't get from practicing problems more, and having a lot of tricks in her back pocket will also probably help with confidence going in.
*At one point one of the review/coaching courses did a promotion where they had people go through an older set of SATs with the questions blacked out to show how well their tactics worked, and the answer was in the vicinity of "embarrassingly well". I'd link to it but I can't find anything about it in the mess of advertisements for those courses, ironically. Google has been unhelpful to me recently.
I'm not familiar with the math subject test, but for the physics GRE, the advice is generally that it's its own thing, and that in preparing you should use mainly GRE specific test prep books and your freshman intro physics textbook.
On the other hand when I tried Bing to see if that would be any better for me the second and third pages of results included this and this respectively. So I have no idea what's going on to be honest.
Shit, looking at this I realize I really should have taken this my last year of undergrad when I was tutoring undergraduates in calculus because it would have been excellent prep.
Tutoring undergrads had me really sharp at these sorts of questions. Instead I took the physics GRE when I was too rather far removed from undergrad physics classes and didn't do too great.
My (blue/pink) collar parents made me go to a math SAT tutor and I showed them by going to a hippie college that wouldn't accept SAT or AP scores.
7, amended: replace "hippie" with "weirdo"
Obviously I have no personal knowledge of the Math GRE but I have a strong belief that all College Board standardized tests can be fairly easily gamed if you get tutoring from someone who knows how to game them. That's certainly true for the SAT and the LSAT. My guess is that you could find a GRE specific-coach who could easily get this kid into the median or higher zone. What this says about our ludicrous higher ed system is left as an exercise for the reader.
Shorter 9: don't get a math tutor, get an exam-specific tutor.
I suspect 9 is not happening with future math grad students. It's probably a large part the tutoring-effect that Trivers mentions, which hadn't occurred to me. There must have been math tutors at Michigan but I wasn't really plugged in and didn't do so.
I also did badly on the math subject test!
I have always blamed that on the fact that I took it on the same day as and following the standard GRE, but perhaps that wasn't the issue (incidentally: I did at least a handful of practice tests, but the actual test had almost no overlap in terms of subject matter the the practice tests I had taken). I was never motivated to retake it, however.
12
did you end up going to grad school anyway?
Oh, also I suspect that it's probably hard to get a high percentile score in October since a large number of people will be retaking it from April and these people will be on-average better prepared than first-time takers in October.
What about doing some realistic practice exams under test-like circumstances? Being surprised by the structure of the exam (number of questions and amount of time) is the sort of thing that should be really easy to avoid. The trick there is getting actual realistic exams, my experience was that Princeton Review was very good at making realistic exams, but that some of the other prep that was more content-based and less game-the-system based had unrealistic practice exams (often unrealistically hard!).
I did the GRE again recently because I am considering going for a professional degree and just want to have a score in my back pocket. I'm a grown up and haven't done any math in years, though back in the day I did take a bunch of math courses.
So I took it twice and got essentially identical scores. 99th percentile verbal, 56th percentile math. BOTH TIMES.
The first time I took it completely cold, fully intending to take it again later on. Then I bought a GRE prep book and been actually drilling math during the months between time one and time two.
I angrily formulated some theories for why my score hadn't budged, of course. It felt like the questions on the exam were harder than the ones in the prep book, but I think it really comes down to the time pressure rather than the actual difficulty. A whole lot of the math section would go very quickly for someone who was comfortable about calculating in his or her head, or had an instinct about numbers. Not having any instinctual response just slowed me down-- enough that even when I knew a lot of the approaches to the problems, I didn't have the time to actually do it. I think there are people with that kind of comfort around numbers, but it requires practically just bathing in math all the time. The time limit catches up with you pretty fast in the exam setting.
The test is for every kind of student, including math majors, who of course actually ARE dealing with math every part of most days. If it wasn't hard enough to test their abilities too, it wouldn't be much of an exam.
15% is really low though (no offense)-- I think you could raise it at least a few points just by adopting some crass test taking strategies. Was it test anxiety or something?
Ha ha, whoops, I meant 14th percentile, not 15th. Told you I didn't have an instinct for numbers!
The test is for every kind of student, including math majors, who of course actually ARE dealing with math every part of most days. If it wasn't hard enough to test their abilities too, it wouldn't be much of an exam.
Right, this is why it's different from the subject exam, which Heebie is talking about in this post.
18:
So, just to get this straight, you didn't have anything to add to the actual discussion here, but are just so pissed that I misread a sentence in the original post that you had to pipe up?
20: Let's get this straight: you're so pissed off to be corrected in a less than fully obsequious way that you're going to be an asshole about it?
Since your mistake was itself fairly substantive, I think pointing it out counts as substantive. I'm surprised you're worked up enough about it to post something pissy and passive-aggressive.
My illustrious colleague has kindly demonstrated for you the pissiness of the mode of expression adopted.
Looking at the questions, I can see that I would have thought "66 questions in 150 minutes" and panicked because each question would have taken me 5-10 minutes to work through. I think that's exactly what happened - I answered maybe 20 questions.
I don't know about the math GRE, but that definitely happens with the physics one. You spend four years learning to be thorough and systematic and thoughtful and taking exams that consist of three or four complicated questions each, and suddenly you're faced with a hundred-odd problems and only 45 seconds to think about each one. But if someone tips you off ahead of time that dimensional analysis and things like "should this answer get really big when the magnetic field is really big, or really small?" are sufficient to answer most of the questions, you realize it's not scary. If no one tells you that, you're screwed.
Chemistry GRE scores were pretty much uncorrelated to undergrad GPA or to any measure of success in graduate school. The main problems that I could identify was that the subject matter was too broadly distributed and idiosyncratic in what they expected you to know. So generalists who are good at memorizing lots of things did well on the Chemistry GRE, but specialists and/or people who don't remember tons of random facts did poorly.
Luckily, most grad schools in chemistry recognize that the subject test is useless.
22:
That was actually me being aggressive-aggressive. I just can't crank my outrage up to the same level as you guys can, so it probably looks pretty pitiful from where you're standing.
That is why I characterized it as passive aggression and not outrage.
We've hired a math coach, like good liberal parents
Uh, good liberal parents hire a coach for an underprivileged kid from the ghetto they moved out of because mumblemumble.
To be more explicit, being a generalist is useless in graduate school. For instance, a surprisingly large part of the Chem GRE is qualitative analysis (i.e. memorizing which cations form insoluble salts with which anions and what things are colored) and stoichiometry/equilibria (i.e. simple arithmetic and algebra), neither or which is remotely connected to being able to reason your way through a complex problem in Chemistry, particularly if you want to be an organic chemist.
So I wouldn't even be surprised to find an inverse correlation between Chem GRE and grad school success.
29
I thought to myself many times while preparing (for physics) that they were really checking to see how well-qualified I was to be a TA for freshman physics.
I ended up taking the math GRE like a month into my first course in abstract algebra, having never taken point set topology. I did not know the definitions of some of the words in the questions. (I did amazingly well, considering, but not great.)
The issue was that I'd never taken my math major all that seriously, because I kept assuming it would get too hard for me at some point. So I double majored and basically did the bare minimum coursework for the math major. Even though I was acing all my math classes in a small program at a small school, no one really talked to me about grad school. It just didn't occur to me that it was an option at all until one of my professors invited me to participate in an REU he was running between my junior and senior years. So time elapsed between (1) the first glimmer of an idea that I might like to go to grad school and (2) time to take the GRE was about 2 months.
Of course, now it seems to be a requirement that admits to the math Ph.D. program have taken at least a couple graduate courses while they were in undergrad. Where does this leave prospective students from institutions that have no grad program? I guarantee none of the faculty members here give a fuck.
What I'm saying, heebie, is your student is probably screwed anyway, apart from the GRE thing. At least don't let her apply here.
Oh btw I am now entitled to comment under this pseud.
did you end up going to grad school anyway?
I did not, and am happy with that decision (though I have wondered more than once what my life would have been like had I gone).
At that point I was more interested in political philosophy than math anyway, and I'm very glad that I didn't go to grad school in the humanities. . .
Congrats Dr. L.! (Would you like any further advice?)
I think I may have met someone who took the history GRE but definitely haven't met anyone who did any history grad school who's said they'd taken it.
I must have grandkids around here somewhere.
Waves hand. I took the History subject test and did go to grad school in history. The subject test was easy, I thought. Even in European history, which wasn't my main area as an undergrad, it was easy. I did take it in the same sitting as the general exam. I never thought not to take it.
Congrats Dr. L!
I think most (or at least many) of our domestic graduate students come from schools with no graduate program. (Though I doubt we take very many students who came from schools that have graduate programs, but who didn't take any of those classes. Unless there's mitigating circumstances.)
Oh btw I am now entitled to comment under this pseud.
Congrats!
Yay dr l!
I took the big standard gre very soon after taking the lsat but did zilch to prepare (for the lsat I bought some old tests and practiced). Result: absurdly good scores on the non math bits and ludicrously bad on math, I'm pretty sure I just got everything wrong on the math bit. I was mildly worried they would I don't know investigate me for fiddling or something? But this is probably a totally normal recurring pattern.
[as my beloved co parent is currently out of the country and I'm deep in a giant work avalanche hence few opportunities to vent or natter to other adults just want to say that while I found the first few emails from the school re increased security absurdly overreacting I am now thinking that the local assholes who sprayed graffitti on a playground wall are well I don't know assholes. I have no point, just as I said venting.]
bog not big. It was on a computer I have no idea how big it was.
41: how long ago was that? My impression was the subject exam was universally recommended against in the late 90s as something that probably wouldn't hurt you but never would help you.
Let me be the first to congratulate Dr. L.
How is it even possible to start with this blog in high school, get college advice from here, and then have any accomplishments whatsoever? L is superhuman.
...and of course, greetings & well wishes to Dr. L.'s mom.
well s/b good. But wishing on wells is not deprecated.
I took the CS subject test 15 years ago and did only moderately-OK on it. I remember that it covered a lot of stuff that I never learned in undergrad (not because I skipped the reading; it wasn't on the syllabus). I feel like it was tilted toward some theoretical curriculum that probably matches what you learn at Stanford/MIT and gets gradually less relevant as you slide down the prestige scale (I went to a mid-tier university, both grad and undergrad).
I imagine a Math subject test would be impossibly broad. Are you sure your student has actually learned more than 50% of what's being tested?
I, for one, would like to thank L. for making many of us question the life choices and lack of accomplishments we've made over the past decade or so.
And then there is the English subject GRE, the Parnassus of dumb. Someone had tipped me off about the particular out-of-print prep book that gave the only accurate available description of the damn thing, so I tracked it down and was forewarned:
- One question will present a sample of Middle English and ask you to distinguish yogh from three;
- One-tenth of the test will cover allusions to Gray's "Ode in a Country Churchyard";
- The rest is association of authors and titles at the level you might require to get through a mixer at the faculty club.
And damn if that book wasn't right about everything. I can't believe that such an incoherent mixture of Oxbridge dowdiness and scan-tron quantification fever stayed stable long enough to generate a document that took two hours to fill out.
Elegy, not ode. Christ, isn't Mr. Oxbridge looking down his node at me now.
61 is totally right but taking it (and the six-week prep group the English department offered) was maybe the most fun I had in college.
How is it even possible to start with this blog in high school, get college advice from here, and then have any accomplishments whatsoever? L is superhuman.
Well, her college experience was disrupted by a hurricane, so there's that.
The notion of expecting people to have graduate classes to get into graduate study unless they went somewhere without graduate programmes is fascinatingly alien - in our system (a) there aren't any classes as such in post-graduate education and (b) no one who wasn't a graduate student could possibly take such a class and (b) I don't think there are any institutions that offer undergraduate degrees suitable for entry into postgraduate study that don't also offer post-graduate study themselves.
I did help a friend prep for the English GRE (I was much, much better at it than he was) but then he got a Rhodes to study English lit so it was a bit of a waste of time but on the other hand it was surprisingly fun to be better at it than he was.
62: I think you'll find it's Gray's Anatomy in a Country Churchyard.
The physics GRE is garbage, for all the reasons essear gave above, and because it has large known biases along gender and racial lines. This article gives a good run-down of why at least astronomy departments are moving away from it.
65: Also, if I recall correctly, she didn't listen to the advice she got. At which point we should have known she'd go far. Congratulations!
65: So the hurricane gave her superpowers? It explains the facts.
67: because, you know, there would be lots of bodies and skeletons lying around. It's the obvious place.
I can't find anything comparable to the analysis in 68 for math. Granted, the gender bias is not as pronounced as it is in physics. I'm not sure how minority representation differs between math and physics, or at least underrepresented minorities.
The article linked in 68 doesn't look at the subject test, I think, but rather the quantitative section of the general test--and parses scores by test taker characteristics (to include area of study).
The physics test was weird. I wish I'd gotten the advice essaer recommended, so I'd have done better. I guess it didn't matter anyway, as I got in everywhere I applied, which ranged from outstanding to good.
Since this is the grad school application thread, can I just vent about how frustrating it is to watch undergrads kneecap themselves for no reason?
An undergrad who worked in my lab for 2 summers and most of 2 academic years, did a great job, presented her research at 2 national meetings, is part of a prestigious underrepresented minorities in science program, and has (I'm pretty sure) a strong academic record, just got caught plagiarizing on her application essay to professional school (not grad school, so it would actually have led to good job prospects).
Why? Half the reason these essays even exist is to weed out the people who plagiarize. It's such a waste and was so completely unnecessary.
Oh, I hadn't clicked through on 68. I assumed it was a link that a friend of mine shared elsewhere, let me rummage that up.
Nope, I misread my friend's link, which was also general GRE, not subject.
Commiseration! I did pretty horribly on the math subject exam, too. I did ace the math part of the general, but that and four fifty will get you a coffee.
My current theory is that the math subject test measures quickness. Which certainly correlates with general math ability, but is going to select against those groups which are traditionally not groomed with clever math puzzle while growing up.
I was fortunate in that I was groomed on clever math problems. My problem was I was swamped with schoolwork and going through some bad relationship issues at the time so I mostly blew off preparing for it. (Yes, I am bad at prioritizing.) I also wasn't very interested in relearning how to do certain arcane integral techniques that might or might not actually appear on the test. I guess where the speed issue gets me is that it doesn't give any time to derive things that will help you with the test, which is something a working mathematician should be comfortable with doing.
66: At least from what I've seen at a lot of largish State universities probably have a category of classes which are kinda/sorta overlapping between graduate student classes and undergraduate classes. How they actually get classified can vary but there are bound to be classes where the two groups overlap* and I wouldn't be surprised if at least some get counted as being graduate level courses on undergraduate transcripts.
*Usually falling into the categories of: "boring survey class for graduate students"; "interesting course for undergraduate majors + added on "seminar" meeting for graduate students where the aging professor tries to stuff you with granola bars and tells you stories about his neighbor's cat"; or "HAHA UNDERGRADUATES IT WAS A TRAP!".
I agree - speed math seems semi-irrelevant to the ability to think deeply about a topic and be driven to prove for yourself that every last detail is true. (It seems great for generating questions that produce normal distributions, however, which is how I'm sure the test is constructed.)
It's not that common on non-professional/vocational graduate level courses here to have much in the way of taught classes. Or at least not classical 'prof gives a lecture' type classes.
On the B.Phil,* which is a partly taught philosophy graduate degree, pretty much all of the 'taught' component consists of tutorials and seminars, and in the seminars, it's the graduate students who are writing and presenting papers. The only lectures anyone would attend would be the same lectures that faculty are going to, i.e. 'visiting famous/important person, presents their latest thing'.
* which is the taught grad degree that I know.
66: Remember that students specialize much earlier in the rest of the world than they do in the US. The intro level graduate classes that I'd expect a student to take here would be undergraduate classes at a good university in the rest of the world.
RWM didn't get into Astronomy grad school because of the physics GRE.
Yeah, it's infuriating. Though I was oversimplifying a bit, it's only partly the exams fault. My understanding is that although she did great in Astronomy classes her Physics grades weren't as good. One root of that problem was that she started in Engineering (in part because in her home town no one understood you could do science that wasn't engineering) and so her lower-level math and physics classes were for engineers, and so weren't really adequate for scientists. Plus, of course, some super sexist Physics professors.
Looked at the math gre example test in the OP. I could figure out most all of that stuff after undergrad, but not necessarily 3 minutes per question. Some things I can't recall now I think I probably would have then, but the speediness, unlikely.
My vague memory of the general GRE was that it was easier than the SAT.
The subject tests are all over the place. Here is Michael Berube's account of retaking the English subject GRE 25 years after starting grad school.
My memory is that the second half of the test gets more esoteric. I can do most of these, even now, but definitely not quickly enough.
But not all of them. Scanning through it, I think ones like 53 tripped me up.
73: Right, sorry, the article is about the general GRE. Last time I did graduate admissions in the US (10 years ago, for an astronomy department), it was conventional wisdom that there was a 20 percentile point difference between men's and women's scores on the physics GRE given equal background and preparation. The racial disparities in the general GRE are a pretty generic feature of standardised testing, so the growing feeling is that the physics GRE is a pretty poor indicator of student knowledge.