Do GRE verbal scores matter for graduate admissions in mathematics?

Question

For the next 6 months or so, I plan to study hard for GRE. Since I don't have much time, I plan to ONLY study for the mathematics portion of the test as I would like to score at least in the 90th+ percentile for the Mathematics Subject Test. But in order to have the capability to achieve such a task within my time frame, I plan to spend no more than 3-4 days to study for the analytic writing and verbal reasoning portion of the general test.

So my question is: Do graduate schools focus or take into consideration the (possibly low score for the) analytic writing and verbal reasoning portion of the test if you were to score high on the mathematics part? Advice from those who are familiar with how selection committee of graduate schools (in Mathematics) pick applicants would be great!

Some clarification: I feel as if I was not too clear with my question. By "studying hard" I don't mean to sit down and just study "specifically" for the GRE test but rather to enhance my knowledge in Real Analysis, Abstract Algebra, Topology, and Differential Equations among other topics such as Set Theory and Logic. GRE Mathematics Subject Test will be something that I will be preparing for along the way. My reason to spend ~6 months has nothing to do with scoring high on GRE. I will perhaps spend no more than 1-2 weeks reviewing the Princeton Review. I just don't know if I will have much time to work on my writing skills if I'm deep-neck into learning Mathematics.

Answer 1

Yes, a decent verbal score is definitely desirable. I mean it: I have done graduate admissions for my math department, mentioned low verbal scores as a point against the applicant, and had the point taken by the rest of the committee. But I would think that spending three-plus years in a university would be the best possible way to study for the GRE verbal: if after all that you need separate study in reading comprehension, sentence completion, and so forth, something has gone wrong!

(I remember the GRE as the SAT all over again, to the extent that my GRE scores were each within 30 points of my corresponding SAT scores. The GRE has been retooled multiple times in recent years with the unfortunate effect that a lot of people doing graduate admissions, like me, are now a little vague on what it actually contains. We tend to assume that the changes are cosmetic and that the exam is more or less the same as the one we took years ago.)

The math subject GRE is not a hard test in the sense that you need to devote much (or any, necessarily) time to study for it. I remember contemplating studying for the Math GRE (circa 1997), learning that the study guide was really pretty miserable, and maybe spending a weekend flipping through a book on differential equations because I had never taken a course on that. More pertinently, I remember that at least one third of the test was rendered trivial by my deeper study of subjects like topology and analysis: e.g. I think there was a question like "Which of the following properties does the real line, as a topological space, not have? Hausdorff, compact..." To me this read more like a mathematical joke than anything else.

Here's the bottom line: please don't spend six months only, or even primarily, studying for any/all of the GRE. What a terrible waste of your time. If I am honest, the fact that you think this might be a good use of your time makes me concerned that you might not be a good candidate for a graduate program in mathematics. I would not say this if you provided any distinguishing or identifying information whatsoever, but as you are completely anonymous it is something to think about. The way you prepare for graduate school in mathematics is by learning math. That means absolutely mastering truly basic, undergraduate level topics like linear algebra and deepening your knowledge of algebra and analysis. This knowledge and mastery will serve you well at every step of the graduate process, starting with admissions.

Answer 2

As a graduate student you will have to write a thesis or dissertation. Your communication skills are important!

When I'm evaluating applicants to our graduate program in mathematics, it goes without saying that the applicant had better have a very high quantitative score and anyone who doesn't isn't going to be a desirable applicant. What separates the desirable students from the undesirable students are reasonably high scores on the other sections of the test.

Keep in mind that students in many disciplines take the GRE. I don't expect a mathematics student to be as good a writer as a student in English, but if the student has a ranking in the 10th percentile, I've got to wonder whether that student will be able to write an acceptable thesis without me doing a lot of editing.

Answer 3

My fairly extensive anecdotal experience is that "verbal" scores are harder to "cram" for, perhaps harder to "game", than the math subject test and "quantitative" exam... so that, perhaps oddly, a high verbal score seems to be a very positive indicator of future success, while subject test success has a more limited indication.

Lowish verbal scores, even with high Subject test scores, seem to be correlated (again, extensive-anecdotally) with a certain brittleness moving forward into PhD-level mathematics, ... perhaps after considerable success in lower-level mathematics. (We can observe the substantially non-verbal, nearly cryptic stylized form of much low-level mathematics.)

So, yes, some people on admissions committees (e.g., me...) pay attention to those other numbers. On the other hand, it's not at all clear that these things can be "studied-for", much less studied-for for several months.

As Pete Clark noted, learning more sophisticated mathematics is often far more effective at more-than-gaming the system, since quite a few of the prankish-shallow questions that would stump novices are transparent to a person with some experience, perspective, and context.

Even so, self-coaching about how to take a timed, multiple-choice test, and some prior samples, are surely helpful to best represent one's knowledge of the subject matter. And being rested, not over-caffeinated, not hung-over, etc. Despite the potential silliness of the last observation, I have observed many people disserve themselves by putting themselves into bad physical states for these and other exams... If diagnosable anxiety is an issue, it should be dealt with forthright-ly...