Questions regarding GRE and Graduate Admission

[HungryChemist]

Here's what my situation is, I hope this will help to better answer below questions.
Citizen(Non-International)
Major:Physics and Applied Mathematics
Junior College - GPA 3.5
Large Public University- GPA 3.8
Research Experience - two summer REU and Honors Thesis Research

I just got my GRE general test scores back. I have 760 math 580 verbal 3.5 writing. I am super happy with 580 since English is not my native language. But 760 in math and 3.5 writing is disappointing and I am wondering now whether to retake the test or not. Any advice would be very helpful.

Also, I am planning on taking GRE Physics this coming October and prepare for it over the summer break. Can you teach me how to well-prepare for it? What scores would considered competitive based on my GPA and so on?

Havard-applied physics
Stanford-applied physics
Columbia-applied physics
yale-applied physics
cornell-applied physics

Do I stand any chance with above school or is it just going to be a waste of application fee?

Thanks

 

[Jacob Chestnut]

Hello,

It's never a waste to try and get into somewhere really great. On the other hand, if you can't afford it, that's the way it is. In any case, I can only tell you about my experience.

I'm just ending the grad. school application process. For the most part, my file looked strong, and I applied to only the top graduate programs in math. In the end, I got into only one of seven schools (I'm waiting to hear back from the last). However, I received the most prestigious fellowship this particular university offered.

Many people thought I was crazy to not apply to any backup schools, but I knew what was important to me and I made that choice. I was willing to sit out a year had I not gotten in toone of my schools.

With that said, I can tell you a little about one of my good undergraduate friends.

He had a 4.0 double major from UC Berkeley in physics and applied math, with something like 10 "A+'s". He did two REU's, like you, and worked in a lab on campus for at least two years. He took a bunch of graduate classes in physics and did well. If I remember correctly he got a 770 on the physics GRE, and did something like 7th percentile on the verbal (he was newly arrived from China, but a domestic student).

He applied to the schools you listed, and more, and was accepted into (off the top of my head):
Cornell
Yale (with fellowship), and
U of Illinois, UC

He ended up at Cornell as a TA. Take from his story whatever you like.

 

[HungryChemist]

Wow, thanks for replying me. I guess my question isn't too much of favorite kind for people to answer. Anyhow, hearing your story, I must bring down my expectation and apply more of safe side school. Thanks~

 

[mathwonk]

Well I am currently reviewing applications to math grad school and looking at math scores like 730, 740, 750 760, 780, 800 etc...and wondering what to make of them. so i went on the ETS? website and looked at a sample of the questions on the GRE.

I was astonished at how trivial they were. It looked like a junior high arithmetic test. Hiow are we supposed to conclude anything about a person's readiness for a math PhD program based on that?

So you do need to be in the upper 700's I geuss to avoid looking like a neanderthal, but it really is a pretty stupid measure. Why don't they make them harder so they begin to separate people out more at the top?

So we look far more at letters of recommendation, and grades in significant courses. We know these tests are nonsense, even if we still look at them.

Statements like: " this student compares favorably with our best past stduents over the past 10 years, who have been successful in PhD programs at schools like Penn, Illinois, and Princeton", count a zillion times more than a silly GRE.

Spend your time reading Spivaks Calculus on manifolds instead, or a good book in your field.

 

[Stingray]

I was astonished at how trivial they were. It looked like a junior high arithmetic test. Hiow are we supposed to conclude anything about a person's readiness for a math PhD program based on that?

That's why there's a separate math subject test. The one you're looking at is meant for all majors (though it's still ridiculous). Still, I agree that GRE scores don't mean much in general.

 

[Locrian]

HungryChemist,

I see you are looking into applied physics. Lab experience is likely to be far more valuable than a very high GRE test score. Most professors worth their salt know the GRE is just another entry fee that people are required to pay to get to apply, and that it has little bearing on the value of the student.

 

[jammidactyl]

Well I am currently reviewing applications to math grad school and looking at math scores like 730, 740, 750 760, 780, 800 etc...and wondering what to make of them. so i went on the ETS? website and looked at a sample of the questions on the GRE.

I was astonished at how trivial they were. It looked like a junior high arithmetic test. Hiow are we supposed to conclude anything about a person's readiness for a math PhD program based on that?

Mathwonk, are you talk about the general GRE scores or the math subject scores? For grad school in the sciences, the general scores do seem pretty pointless; some schools I've looked at outright ignore them. But for the math subject exam, I wouldn't exactly call those questions trivial. They're obviously not at the depth of qualifying exams, but the exam covers enough ground to make up for that.

Looking at the statistics available in the practice booklet, the questions at the end of the test are measurably more difficult. For instance, while 90% of people get the first few questions correct that percentage drops to 30% and below for the later questions. Also the range of scores on the subject exam goes up to 900 versus 800 for the general exam. This means there's a pretty good distinction between a score of 700 (70th-percentile) and a 890 (99th-percentile) on the subject exam.

 

[jammidactyl]

To the original poster, with your background I'd say a physics GRE score of around 800, plus good letters of recommendation would make you competitive.

On the subject of preparation (for the math or physics exam), the first step is to look at the material available from ETS:

http://www.ets.org/Media/Tests/GRE/pdf/Physics.pdf
http://www.ets.org/Media/Tests/GRE/pdf/Math.pdf

Next go through all your standard reference books and practice problems. To further demonstrate your proficiency with the topics, you might want to look at qualifying exams from the schools you'd like to apply. Also in math we have this reference:

http://math.berkeley.edu/~desouza/pb.html

But individual schools usually post old quals as well.

 

[HungryChemist]

To the original poster, with your background I'd say a physics GRE score of around 800, plus good letters of recommendation would make you competitive.

Thanks for the input. I guess at this point, the only thing I can do is to working GRE Physics. I wasn't thinking of taking GRE Mathematics though.

 

[mathwonk]

well i will now look at the subject questions, but the scores that are reviewed more for admissions are the stupid and trivial general scores, in my experience. maybe i am admitting to my own stupidity but i am telling you what is going on right now in my own grad admissions process.

 

[mathwonk]

ok i looked at the subject test in math. it is much harder than the general test. i.e. instead of being a junior high arithmetic test, it is a sort of SOPHOMORE LEVEL MATH CONTEST TYPE OF teST. STILL IT SEEMS To have no relation at all to the kind of thinking i am interested in in a grad stduent.

there are no statements of significant theorems, there are no proofs, there are no hard problems.

this kind of stuff just distinguishes better undergrads from weaker ones, not real mathematicians in my opinion. stilol it may identify people with some math ability. forgive me it is so long since I looked at such trivial picky questions. these are games, not significant mathematics questions.

if you want to be a mathematician, think about deep results in math and try to understand and extend them, instead of worrying about this type of minor problem. this is ok for fun, but it seems pretty shallow to me.

to be explicit and personal, i was extremely good at this kind of nonsense in high school, but in college did not have a clue how to attack a significant homework problem even in freshman math from john tate. read courant and spivak and do their problems, and apostol. if you are any good you are hopefully far beyond the GRE level by the time you graduate from college.

 

[Jacob Chestnut]

I agree with the poster above me wrt the GRE subject exam. I've been doing research, and taking graduate/reading courses for the last two years. When it came time to do the GRE, I didn't do well at all. I think that kept me out of some schools, but if those are the kinds of students they're looking, then those aren't the places for me.

 

[bigplanet401]

Hi Jacob,

The GRE IMvHO is window-dressing at any top-20 school. Grades, letters, any previous research and papers, etc. have much more weight. Chew on this:

460 Verbal (and I am a native speaker ) + 780 Math + 760 physics + 3.5 writing didn't seem to matter too much in Caltech's eyes.

So don't sweat it, at least as far as test scores are concerned...