To take GRE Math or not

[mindreader]

Hi all, I am a rising senior in a top 100 university in south east asia (according to topuniversities.com, upper half) concentrating in Applied Math. I hope to pursue a PhD in Applied Math in Brown or Cornell and both schools indicate that they recommend the GRE Math subject test but it is not required.

I am reluctant to take the test because I'd have to spend for it, prepare for it, and that as an Applied Math concentrator, I did not get to take courses such as Abstract Algebra, Topology, and Number Theory. Most of my coursework is on Scientific Computing (Numerical Analysis, Differential Equations), Algorithm-based Math (Optimization, Game Theory), and a few Statistics courses. I have, however, taken Real Analysis part 1 (Theory of differentiation, limits, series, etc., no integration) and I will be taking Complex Analysis in the fall, both of which are required for my concentration.

I'd like to know whether anyone has managed to get accepted into the above-mentioned programs without taking the GRE Math. If so, what did you think compensated for not taking the test? (OR is this unlikely? It seems that those who put their details on the tracker here took the GRE Math for Applied Math PhD at Brown.)

I am unsure about my profile but my GPA is in the top 5% of my cohort, I have 1 journal publication, 1 conference paper, and I've also done an attachment at Imperial College London. The 2 publications were borne from 2 1-year research programs in my local university.

Advice would be appreciated. And if you do have recommendations on review materials, that would be awesome. My target score would be 800 or so. But still, I am reluctant to take it.

(BTW, Is this post in the right forum?)

[kka34]

At most Canadian universities, for physics (not sure about math), we have a similar GRE requirement. It is only recommended for Canadians, and required for international students. I talked to a professor about this, and he said that they do this because they are not familiar with the physics programs in other countries. They don't know how good your institution is and being in top 5% at a bad institution doesn't mean anything. The GRE score gives them an indication of your skill without having to know where you are from. I wouldn't be surprised if Cornell and Brown did the same thing. Have you thought about emailing the universities about how "recommended" the GRE is for international students?

[mindreader]

Hi @kka34, thanks for your insights. I've read the same thing as to why GRE subject tests can be useful it's just that I've also read that "students should only take it when they are confident they'd do very very well." In my case, the pure math parts (Topology, Abstract Algebra, and Number Theory) might be detrimental to me. I can always review for the test but at the same time, I'd be occupied with the application process coupled with my upper division courses (which are becoming more time-consuming).

About emailing the universities, both Brown and Cornell indicate that it's recommended - no adjectives like "very" or "highly". Perhaps it might be worth a try to e-mail them. Do you think it's advisable to ask them roughly what proportion of their admitted students have taken the GRE math test?

[TakeruK]

You can try, but I have found they don't usually give information like that out. Also, the numbers will be skewed because most Americans applying to US schools would have taken the GRE Math test anyways since they probably need it for other schools.

I know studying and paying for the test can be a big pain! I would take the test once and see what happens. If you get a bad score, then don't report it to Cornell or Brown (when you report score, you can choose to report Subject, General, or Both -- just choose General).

I feel like if it says "recommended", then you should submit it unless it's really bad. And if you don't submit one, they might just assume it's bad (since the majority of Americans would probably submit one). I applied to Planetary Science programs for this fall and the Physics GRE was not required at some schools, and "recommended" at others. Like Applied Math, Planteary Science is related to Physics but a large part of the test is not my main area of interest and with my undergrad Astronomy major, I took fewer Physics courses. I still submitted my scores (44th (2009) and 51st (2011) percentile I think) and got in to all the places that didn't require or only "recommended" Subject test. I think at least if it says "recommended", you know that they don't care about the GRE as much as other things!

Basically -- if you take the test and submit your score, it's a show of "good faith" and even a mediocre score is better than no submission, in my opinion.

[mindreader]

@TakeruK Oh hey that's wicked - you got into Caltech! It oughta be competitive since it's a small size program. Thank God someone feels my pain of having to pay for loads of exams! Not to mention having to pay for them to be sent! and add to that, there's TOEFL for me...

I tried the 2nd most recent GRE Math subject test posted by ETS on their site. Though this was published in 2007 and that the most recent was published in 08, the 2007 material was the actual exam way back in 2000. I tried it, without reviewing my courses and discounting the time, and I scored 40/66 which translates to the 68th percentile. Hmm, judging from the scores you submitted to Caltech, adding in some parallelism between Planetary Science and Physics and Applied Math and Pure Math, do you think a score in the 70th percentile would be fine? Fretting a little bit because I've seen on some forums that Chinese students could easily snag a score of 800+/900.

[mindreader]

and oh, how bad is "real bad" ?

[TakeruK]

@TakeruK Oh hey that's wicked - you got into Caltech! It oughta be competitive since it's a small size program. Thank God someone feels my pain of having to pay for loads of exams! Not to mention having to pay for them to be sent! and add to that, there's TOEFL for me...

I tried the 2nd most recent GRE Math subject test posted by ETS on their site. Though this was published in 2007 and that the most recent was published in 08, the 2007 material was the actual exam way back in 2000. I tried it, without reviewing my courses and discounting the time, and I scored 40/66 which translates to the 68th percentile. Hmm, judging from the scores you submitted to Caltech, adding in some parallelism between Planetary Science and Physics and Applied Math and Pure Math, do you think a score in the 70th percentile would be fine? Fretting a little bit because I've seen on some forums that Chinese students could easily snag a score of 800+/900.

I also had to travel 2-3 hours to get to a city that gives the GREs. With exams at 8am, I had to get there a night before so in total, I spent around 800 CAD to take the tests, travel to the tests, and submit my scores!

Caltech's Planetary Science program "strongly recommended" the subject GRE, just for full disclosure. They also told us they accepted 8 students out of 33 applicants, which was a higher fraction that I would expect (not complaining though!). I'm an international student as well. But I'm from Canada, so I think since the US and Canada school system is similar, they also don't have to use the GRE as much to "standardize" things.

It's really hard to tell what score is "good enough" or "really bad" because it depends on the rest of your profile. I also got into Cornell's Astronomy program (which is closer to Physics than Planetary Science) and they required the Physics GRE. I do have a lot of research experience, which I think was what helped me a lot. It also depends on a lot on the school you're applying to.

I would consider a "bad" score to be below whatever the minimum score for even applying. Some schools say this is a 500 / 800 (but not give a percentile). Scores around what I got aren't good, but if you think about it, I got the median score of test takers, and only people considering grad school are taking the test, so the median of that population isn't so bad. At least that's what I tell myself Personally, I would probably submit ~40th percentile and above for "strongly recommended" and maybe only 60th percentile if it is "optional".

One disadvantage of being international is that the departments in public schools pays up to 5x as much tuition for us as an American student so they take much fewer international students. University fo California schools is apparently very selective in its international student rate (only 10%), compared to east coast schools like Harvard (30%) [got rejected from all of the above anyways lol]. Another advice I got from my profs is to apply to the private schools in the US because all students have the same tuition (Public schools are funded by American tax dollars which people want to spend on American students, not internationals!).

I really can't say what score you will need to get in. I think your GPA sounds good (top 5% is probably pretty good; my undergrad GPA wasn't in the top 10% even) and your research experience is typical of most applicants, I think. The most help I could give is to tell you what I applied with and the results:

GPA: 3.88/4.00 (Undergrad); 4.0/4.0 (Masters)

GRE: Quant 800 (94%); Verbal 630 (91%); Writing 4.5/6 (66%?)

Physics GRE: 640 (44%) in 2009; 690 (51%?) in 2011

Research: ~3 years total full time research at 4 positions, 6 co-authored publications

Accepted: U Washington Astronomy, Caltech Planetary Science, U Arizona Planetary Science, Cornell Astronomy, Toronto Astro

Rejected: UC Berkeley Astronomy, UC Santa Cruz Astronomy, Harvard Astronomy

But you should also take a look at these forums: http://www.mathematicsgre.com/ if you haven't already. Just remember that people who score really highly are much more willing to post and share their scores than those who didn't. It's not a fair sample! So hopefully my less than stellar GRE scores will balance those posts out, haha. Good luck!

[mindreader]

@TakeruK man thanks a lot for the insights! Yeah I am mostly applying to private universities all of which are located in the east coast, mostly in new England and mid west. I just noticed that you did a master's degree first before embarking on a PhD. I've been pondering that as well - for all of my safety schools I'm only applying to their master's programs because I'd want to pursue my PhD in a school that I really really like. The school's location is a big factor in my decision as I'll be spending the 5-6 years of my early 20's over the course of the PhD

[TakeruK]

The main reason why I did a Masters degree first, then applied to PhD was because that's the way it works in Canada. Except for a few programs that are emulating the US system, you need to have a Masters degree to apply to a PhD program. The two programs are completely separate -- even if you are staying at the same place for PhD, you have to go through the whole application process again (including LORs from your Masters supervisor to your PhD supervisor, who may be the same person, heh!). In total, a Masters degree is 2 years and a PhD is 3-4 years, so in the end, a Canadian student will have the same amount of grad school years as an American! I had non-academic reasons to stay in Canada as well, but now I have to go to the US to continue Planetary Science research since it's a very very small field in Canada.

The nice thing about Masters programs in Canada is that they are fully funded because you are basically treated as a PhD student with fewer course requirements and your thesis research does not have to be novel. But most Masters-only programs in the US are NOT funded and will be very expensive .... so maybe this is not the best safety school idea? Also, most US PhD programs will NOT really count your masters degree (from Canada or otherwise) towards your PhD, so it's still another 5+ years at that stage. Anyways, maybe you want to apply to some Canadian masters programs too then, as safeties, since you will be funded if accepted to Canada.

[R Deckard]

@TakeruK Oh hey that's wicked - you got into Caltech! It oughta be competitive since it's a small size program. Thank God someone feels my pain of having to pay for loads of exams! Not to mention having to pay for them to be sent! and add to that, there's TOEFL for me...

I tried the 2nd most recent GRE Math subject test posted by ETS on their site. Though this was published in 2007 and that the most recent was published in 08, the 2007 material was the actual exam way back in 2000. I tried it, without reviewing my courses and discounting the time, and I scored 40/66 which translates to the 68th percentile. Hmm, judging from the scores you submitted to Caltech, adding in some parallelism between Planetary Science and Physics and Applied Math and Pure Math, do you think a score in the 70th percentile would be fine? Fretting a little bit because I've seen on some forums that Chinese students could easily snag a score of 800+/900.

mindreader, realistically you are not going to get into those schools without the Math GRE.

I know domestic students that have gotten in with scores in the 70's, but you will probably be held to higher standards being an international student. Also, expect to score quite a bit lower on the actual exam than you did on the practice exam(s).

What other schools are you considering? Maryland and Stony Brook might be good fits. I know students that got into Maryland without the Math GRE, by the way.

[mindreader]

@R Deckard - hey man thanks loads for the advice! Ok I have decided, bregrudgingly, to take the GRE Math and I'll start preparing for it after I take the TOEFL exam which is scheduled after I finish my summer research project.

Well, right now I'm slightly confused. On my list, I only plan to apply to 3 PhD programs - Cornell and Brown (Applied Math) and Columbia (IEOR). My research interests are in stochastic DE and stochastic control hence I am considering Brown. I chose Applied Math over ORIE at Cornell because the ORIE program places too much emphasis on optimization and statistics rather than probability. Columbia, on the other hand, does not focus on stochastic D.E.'s in its Applied Math Department.

In choosing the schools I want to pursue a PhD in, the location is a major consideration. An exception would be Cornell - it appealed to me because they mentioned on their site that they accept about 150 applications and that they admit 10-15% of the students (whereas for Columbia IEOR, they receive about 300 applicants for 10 places). Brown and Columbia are in phenomenal locations (with Providence being close to Boston) and I would not hesitate spending the next 5 or so years of my life there.

Other than that, I'm applying to Master programs for the safeties. I thought Maryland might be a good choice but they don't offer funding for Master programs. I have, however, scouted some schools which offer TAships to Master students (which entitles them the tuition waiver) and a reasonable stipend. I thought that doing a Masters degree would somehow strengthen my foundations and would better prepare me for the PhD program. That being said, the location of the school doesn't matter that much as it is only for 2 years but I prefer universities in the east coast. So far, I'm considering these schools for the Masters degree in applied math: Johns Hopkins, UIUC, UMich-Ann Arbor, and UMass Amherst.

What do you think?

@TakeruK - seems reasonable to do a masters program in Canada! But I'm not entirely sure that all of the master programs are funded. Let me check on that again.

I was thinking of applying to ETHZ and EPFL for their applied math masters too (not sure if you've heard of 'em) because they offer scholarships and that my school has partnerships with them...

[mindreader]

@TakeruK - do you think McGill is an ambitious goal? It seems that McGill's applied math program jibes with the focus of my undergraduate studies and yes, as you mentioned, master's students are normally funded.

I checked other schools such as UToronto and UBC but the applied math program for the latter does not focus on my research interests while the program for the former seems to imply that funding is not available for students in the terminal Master's. My end goal is to pursue a PhD in the states due to personal reasons and that's why I'm not interested in the doctoral stream master's...

[R Deckard]

In choosing the schools I want to pursue a PhD in, the location is a major consideration. An exception would be Cornell - it appealed to me because they mentioned on their site that they accept about 150 applications and that they admit 10-15% of the students (whereas for Columbia IEOR, they receive about 300 applicants for 10 places). Brown and Columbia are in phenomenal locations (with Providence being close to Boston) and I would not hesitate spending the next 5 or so years of my life there.

Other than that, I'm applying to Master programs for the safeties. I thought Maryland might be a good choice but they don't offer funding for Master programs. I have, however, scouted some schools which offer TAships to Master students (which entitles them the tuition waiver) and a reasonable stipend. I thought that doing a Masters degree would somehow strengthen my foundations and would better prepare me for the PhD program. That being said, the location of the school doesn't matter that much as it is only for 2 years but I prefer universities in the east coast. So far, I'm considering these schools for the Masters degree in applied math: Johns Hopkins, UIUC, UMich-Ann Arbor, and UMass Amherst.

Just an FYI, that statistic on the Cornell website is not accurate -- they accept far fewer candidates than suggested. I'm not sure that applying to masters programs is a great idea. I suppose it depends on your goals, though. It might be better to just apply to PhD programs if you think you can get into one that you would enjoy being at. Again, Maryland is an excellent school (better than Columbia in applied math) and also has a great location, along with quite a few people doing work you would probably find interesting. NYU is also a great school, although admission is ultra-competitive (much like Brown).

Of the master's programs you listed, JHU and Michigan would be good choices. UIUC does not even have an applied math program, so I wouldn't bother applying there. UMass a step down from the other programs you listed, but still be a solid back-up choice.

[mindreader]

Just an FYI, that statistic on the Cornell website is not accurate -- they accept far fewer candidates than suggested. I'm not sure that applying to masters programs is a great idea. I suppose it depends on your goals, though. It might be better to just apply to PhD programs if you think you can get into one that you would enjoy being at. Again, Maryland is an excellent school (better than Columbia in applied math) and also has a great location, along with quite a few people doing work you would probably find interesting. NYU is also a great school, although admission is ultra-competitive (much like Brown).

Of the master's programs you listed, JHU and Michigan would be good choices. UIUC does not even have an applied math program, so I wouldn't bother applying there. UMass a step down from the other programs you listed, but still be a solid back-up choice.

R deckard, I sense that you are staunchly vouching for UMaryland. lol Are you attending there this fall? Where are you heading to?

To me, I haven't been to Maryland before but comparing Baltimore and College Park from whatever I can muster from Wikipedia, Baltimore seems like the ideal city. Not sure so perhaps you can correct me on that. Note: I've spent 6 weeks in London and I pretty much enjoyed the city vibe. It wasn't as crazy as NY but it was certainly not a quiet town in the midst of nowhere...

Last time I checked, UIUC has an applied math master program, aside from the actuarial science masters program in the math department. Although I might be dropping this option since I've started considering other safety schools. JHU is certainly the master program of my choice as the courses and tracks they offer appeal to me. The teaching experience I could gain from there would be no doubt invaluable.

With regards to UMass, I am applying to it because after sifting through the profiles of Brown's PhD students, I came across one student who did her master's here before embarking on the PhD at Brown. The intention is not really to emulate her academic path but rather a gauge of what schools Brown deems as "acceptable."

Would you mind expounding how doing a master's might not be a great idea? What aspects are you talking about? I'll be pretty much young once I finish my undergrad (21) and so an extra 2 years won't really hurt that much.

[TakeruK]

@TakeruK - do you think McGill is an ambitious goal? It seems that McGill's applied math program jibes with the focus of my undergraduate studies and yes, as you mentioned, master's students are normally funded.

I checked other schools such as UToronto and UBC but the applied math program for the latter does not focus on my research interests while the program for the former seems to imply that funding is not available for students in the terminal Master's. My end goal is to pursue a PhD in the states due to personal reasons and that's why I'm not interested in the doctoral stream master's...

Yeah, UToronto is "Americanizing" its graduate program and is the exception I usually think about when I say "most programs in Canada....[do something that is different than US]...."

Most people would view McGill, UBC, and Toronto as the top 3 schools in Canada, broadly speaking. They are the most well known in the US and would be more helpful if you want to go to the US for PhD. There are other good schools for specific fields though (e.g. Engineering at Waterloo is strong). I'm not at one of those 3 right now, but I did my undergrad at UBC so that helped I think. So, McGill is a great school to apply to, especially since it does sound like you find a good fit there.

I don't mean to pressure you into looking at Canadian programs -- it might not turn out to be the best choice for you, obviously! Just giving you another option.

By the way, some people might suggest that you can enter a doctoral stream masters at Toronto and then graduate with a Masters and not enroll in their PhD program. It's technically possible but everyone I've met who said they were going to do just that ended up being pressured into the doctoral stream somehow.

[R Deckard]

R deckard, I sense that you are staunchly vouching for UMaryland. lol Are you attending there this fall? Where are you heading to?

To me, I haven't been to Maryland before but comparing Baltimore and College Park from whatever I can muster from Wikipedia, Baltimore seems like the ideal city. Not sure so perhaps you can correct me on that. Note: I've spent 6 weeks in London and I pretty much enjoyed the city vibe. It wasn't as crazy as NY but it was certainly not a quiet town in the midst of nowhere...

Last time I checked, UIUC has an applied math master program, aside from the actuarial science masters program in the math department. Although I might be dropping this option since I've started considering other safety schools. JHU is certainly the master program of my choice as the courses and tracks they offer appeal to me. The teaching experience I could gain from there would be no doubt invaluable.

With regards to UMass, I am applying to it because after sifting through the profiles of Brown's PhD students, I came across one student who did her master's here before embarking on the PhD at Brown. The intention is not really to emulate her academic path but rather a gauge of what schools Brown deems as "acceptable."

Would you mind expounding how doing a master's might not be a great idea? What aspects are you talking about? I'll be pretty much young once I finish my undergrad (21) and so an extra 2 years won't really hurt that much.

I am not attending Maryland, although I was admitted and visited the program. I was really impressed, but it just wasn't my top choice (although it was my favorite in the Mid-Atlantic region). It's close to DC, which is an awesome city, and also pretty close to Baltimore if you want to go there.

UIUC technically offers a MS in Applied Math, but to say that applied math is not popular there would be an understatement. If you want to consider UIUC, I would also join the CSE program (an excellent program) and have a second adviser outside the math department.

There are a lot of other good programs in interesting cities for applied math as well...

[mindreader]

I am not attending Maryland, although I was admitted and visited the program. I was really impressed, but it just wasn't my top choice (although it was my favorite in the Mid-Atlantic region). It's close to DC, which is an awesome city, and also pretty close to Baltimore if you want to go there.

UIUC technically offers a MS in Applied Math, but to say that applied math is not popular there would be an understatement. If you want to consider UIUC, I would also join the CSE program (an excellent program) and have a second adviser outside the math department.

There are a lot of other good programs in interesting cities for applied math as well...

Hmm, thanks R Deckard for your advice on UIUC. The location feels a bit off to me as well, judging, again, from whatever I find on Wikipedia. That's a more compelling reason for me to drop it off my list.

Do you mind telling me where you're heading to? And when you mention "There are a lot of other good programs in interesting cities for applied math as well", do you have other recommendations? Preferably schools in the east coast. I am kind of allured to the New England feel that's why Providence (and its proximity to Boston) appealed to me a lot...

@TakeruK - hey its fine! a master's in canada doesn't sound so bad... I do appreciate the advice though! And I'm still debating whether it's worth it to apply to European master's programs since most of them have scholarships or great funding.

[R Deckard]

You can't really learn anything about a city from Wikipedia. I would go ahead and throw out any opinions about places you have based on that.

Some additional applied math programs you might be interested in based on being in interesting cities:

ICES or Math @ UT Austin

Applied Math @ Washington (Seattle)

ACO or CSE @ Georgia Tech (Atlanta)

Math @ Minnesota (Minneapolis--Saint Paul)

Math @ UChicago

ESAM @ Northwestern (near Chicago)

Math @ MIT (near Boston)

Math, ACO, or PAL @ CMU (Pittsburgh)

ALL of these programs are excellent (for the record, I'll be attending one of them).

Edited June 7, 2012 by R Deckard

[TakeruK]

You can't really learn anything about a city from Wikipedia. I would go ahead and throw out any opinions about places you have based on that.

I think you can get some useful objective (i.e. statistical data) from Wikipedia, but with some caveats! I like how almost every city's entry in Wikipedia has a very useful table of average, high, and low temperatures for each month. I found that the numbers tend to contradict my preconceptions about what the weather is like in certain cities. Students at places I visited gave my numbers that agreed with Wikipedia data so I guess that's good enough for me! Other useful things are population size and a breakdown of demographics, but I guess you would have to be careful about their sources.

However, if you are going to be spending 90%+ of your time with people from your school, there is a high chance that most of them aren't from that city so the numbers won't apply! But if you are moving with someone who will be looking for work, then other useful things like the largest employers etc. could be of interest as well. But I agree that while the rest of the article may be true, depending on who has been editing it, it may not really represent a balanced enough view that you can infer what the city is like, in my opinion. Also, as a graduate student, we will probably have a different experience than someone who is there permanently!

[mindreader]

@R deckard - that's a great comprehensive list! Yes since I'm taking the GRE Math, I'm applying to CMU too since the application's free anyway. Btw, what do you think of Boston University? I mean for one, it's located in Boston beside Harvard and MIT! Since they're requiring GRE Math might as well ponder about it.

[ANDS!]

You can't really learn anything about a city from Wikipedia.

The hell? This entirely depends on what kind of information you need. Wiki is a good "starting point"; no one shouldn't base their entire impression of a city on Wiki, but it's doubtful anyone smart would.

To the OP, much like the general test, subject GRE is one piece of a larger puzzle. I've seen folks (from my institution) get into great schools with a score in the 30%. I would say as an applied guy, if you are grabbing above a 50% (since half the exam is Calculus) you should be pretty set (which of course carries the caveat - "depending on where you are applying").