Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Gre scores

  1. Oct 11, 2005 #1
    I need someone to be honest with me. I just got back from taking the GRE and I got a 410/610 for verbal/quantiative. I have 3.79 gpa with a double major in EE & physics with some research experience. But I want to get into cornell or MIT. Do I definitely need to retake? I hate these stupid standardized test. They are such a waste of time to study for.
  2. jcsd
  3. Oct 11, 2005 #2

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Yes, I think you do need to retake the exam, and you should probably take a prep course. The GRE is very coachable, and it's more a matter of beating ETS* at its own game than anything else. I have taught GRE courses for The Princeton Review, and I've gotten people to improve by over 100 points on each section. Of course, the higher your score is the more difficult it is to improve. So in your case you can expect a prep course to bring you up by more than 100 points on the verbal part, and less than 100 points on the quantitative part.

    If you try to go to MIT or any Ivy League school you are going to be going up against a bunch of valedictorians. I would think that the fact that you didn't exactly break the bank on the math section is going to raise some eyebrows.

    *ETS=Educational Testing Service, the authors of the GRE, SAT, and GMAT
    Last edited: Oct 11, 2005
  4. Oct 11, 2005 #3
    I barely have enough money to take another GRE, let alone the $1000 plus classes. I went through barrons quantitative section, but I didn't really study for the verbal. Maybe I should just get an old book of past GREs and practice. The thing is, I can answer all the quantitative questions, eventually. I just tend to be a slow test taker. That probably just means I need to do more practice problems. Plus the physics GRE is coming up and I'm trying to prepare for that, but with classes and everything there just isn't enough time. I wish schools had December admission, then I wouldn't have to cram all this stuff in. Any advice? I could really use it.
  5. Oct 11, 2005 #4

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Here's what I would do then:

    * Call the admissions offices of each of your schools of interest and ask them to give you their opinion.

    * If you have to retake the exam (which I think will be likely) then go ahead
    with your plan to practice.

    * Regarding the verbal section: You should look into getting a Princeton Review book that covers the "Hit Parade", which are the 300 most frequently recycled vocabulary words on the GRE. The problem solving tricks that you find in any book will be useless to you unless you have a great vocabulary.

    And I have a question for you: Did you take a pencil-and-paper test, or the computer adaptive test (CAT)?
  6. Oct 11, 2005 #5
    That sounds like a good to just call and ask. I assume they would give a straight answer.

    I took CAT--I think you have to unless they don't have CAT within a two drive or so.

    I'll consider picking up the hit parade. Have you looked at Barron's 333 high frequency list?

    But what about the quantitative part. Would just working more problems be the best thing?
  7. Oct 11, 2005 #6

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    OK, then there's something else you need to know: Not all the questions are of equal value.

    The average GRE score on any section is 500 (this is not a coincidence, the test is engineered to produce that result). At the start of the exam the computer thinks that you're an average test-taker, and it gives you a 500 level question. If you get it right then the computer thinks more highly of you and gives you a more difficult question at, say, the 620 level (I'm just pulling this number out of the air). But if you get it wrong you are demoted by the same amount, so you would be at 380.

    These adjustments start out coarse and get finer as you move through the exam. If you get yourself in a hole early then it is very difficult to dig yourself out. By the time the exam is 2/3 over, you are fluctuating in a narrow band of scores, and by the time you get to the last few questions your score is basically determined (except for the finest adjustments).

    What this means is that the early questions are by far the most important, and that you should spend the most time on them. Getting the last 5 wrong doesn't hurt your score that much, whereas getting the first 5 wrong is a total disaster. This brings me to the toughest thing to get through to a GRE student: In most cases you can actually improve your score by slowing down.

    You should get the PowerPrep software with practice GRE CATs to get used to this system. It is authored by ETS.

    I haven't, but I do know that Barron's is a good name. I would assume that their list is well-researched.

    Yes, but there are tricks to be learned. Technical students are the hardest to coach for the GRE quantitative part, because they want to tackle all of the problems the way they learned to in their courses (set up equations and solve them, etc). ETS knows that the average test taker is going to try to solve the problems that way, and they deliberately put more problems on the exam that can reasonably be solved conventionally within the time limit.

    There are techniques for avoiding algebra, spotting "trap" answers, and using estimation that help cut through the minefield that ETS lays out for you.
  8. Oct 11, 2005 #7
    I knew everything that you told me. I used the powerprep software and did the two exams, which i averaged around 500/640. And I learned alot of the tricks from barron, but for some reason I was much slower on the real test. I still had 5 or 6 questions left.
  9. Oct 11, 2005 #8
    ETS cumulative score reporting

    Regardless, if you submit new scores MIT will see your old scores as well.

  10. Oct 11, 2005 #9
    do you think it really matters if they see my old scores?

    I figure at this point with classes and the physics GRE, there is no way I can go back and study for the general GRE and still do decent on the physics GRE. So if I have to retake it, then I will essentially have to push back applying for grad school for another semester. Also, since admissions for the top schools are only in the fall, then I would pushing my starting of grad school back a year. Which sucks, but I would be willing to sacrifice if it meant getting into a top school.
  11. Oct 11, 2005 #10
    I was just reading your post again Tom, and you said that you should slow down at the end. But how many questions can you leave blank and still get the same score?

    Thanks for all the advice. I think if I memorize the word list and work math problems till they all look the same (is this possible?), then I should be alright. I've studied quite a few of the math tricks, and If I see that a problem takes more than a minute then I'll look that sort of problem up in a guidebook.

    I just don't see why graduate schools take this test seriosly. I am essentially studying how to take a test.
  12. Oct 11, 2005 #11

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    NO NO NO

    Slow down at the beginning!

    Don't leave any blank. The penalty is pretty stiff. Instead, if you see yourself running out of time just pick a letter (A through E) and click it until the questions are all used up.

    The math problems won't all look the same because of the different question types (problem solving, charts, etc). But practice can only help. If you are going to forego a prep course, then I strongly recommend that you get the PowerPrep software. If I am not mistaken you can order it from ETS when you register for the CAT.

    That is exactly right. In fact right at the beginning of the Princeton Review's GRE Manual, it says in bold font: The GRE tests how well you take the GRE. That that's all it does. But ETS has managed to hype up their exams as "intelligence tests" because they produce nice bell curves. After all, everyone knows that a small fraction of the population is very intelligent, a small fraction is very....not intelligent, and most people are unspectacular. ETS exam results reflect that, but it is because the exams are engineered to reflect it. On every ETS exam there are unscored "experimental" questions that may or may not appear on a future exam, depending on whether or not the correct number of people answer it correctly.

    Get a load of that: You pay them to do their R&D. :grumpy:
  13. Oct 11, 2005 #12

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    My mistake: Pick a letter from A through D. Not all questions have 5 answer choices, but they all have at least 4.
  14. Oct 11, 2005 #13
    I've used the powerprep software. I averaged a 500/650 on verbal/quantitative when I took the two practice test. I went down a bit on the real thing--stress, I guess. So what is the best way to study quantitative, because I know anything less than 700 is going to be looked down upon by the top schools. And I suppose 300 or so vocabulary is enough for the verbal.

    I know your giving your advice away for free, but I do appreciate it. If I could afford prep classes I would.
  15. Oct 11, 2005 #14
    I just took the test and they had an entire *section* that was experimental. After sitting there for 3 hours taking the real test, I wasn't much inclined to participate. :devil:

    Brentd: You mentioned Barron's. Have you got their Passkey to the GRE Book? It's a pretty small (and cheap :tongue: ) My situation was unusual in that I had already been accepted into Grad School, so I didn't put a whole lot of time into studying, but I found that a little time spent with this book really paid off. The tips and tricks in there are really helpful, but there's a limit to how much they can help. If you don't know how to solve the problem (or at least reduce the number of possible choices) all the tips in the world won't help you. They just help you *not* miss questions you already know the answer to.

    In your case, if you're finishing up a degree in physics, you've got all the math you'll ever need (and then some!) for the quantitative section of the general test. If you're missing a lot of problems here, it's most likely because you're running out of time -or- getting fooled by an answer that looks right but isn't. If you can get yourself familiar with all the *types* of problems in that section and the best way to solve them, you should be able to raise your quantitative score quite a bit. I can't get a whole lot of advice for the verbal section other than just learn some good vocabulary words (the Barron's book has a decent list of maybe 300 - might be the same one you're talking about.)
  16. Oct 12, 2005 #15
    The ETS unscored sections


    ETS does not label its experimental sections. Once you get to the second verbal or quantitative section, there is no way for you to know if you have already taken the experimental section or not.
  17. Oct 12, 2005 #16
    Typical MIT applicants and their perfect-800 GRE-Q scores


    University of Virginia
    MS in Computer Sc.
    The current nominal scores for successful applicants for admission are 750 for the quantitative part and 550 for the verbal part. The average scores of accepted applicants are generally higher than those scores.

  18. Oct 12, 2005 #17
    It didn't say experimental, per se, but it had a disclaimer at the beginning that said 'Your participation in this section is optional and it in no way affects your final score' or words to that effect. I thought it was some type of survey actually, but it seemed to be testing a new format, i.e. different shaped boxes, questions that required more than 1 answer, etc. Since it was 45 minutes long I stopped after about 2 questions and left. It wasn't reflected on the final score report that popped up on the screen at the very end.
  19. Oct 12, 2005 #18
    I don't understand that statement. How can you average the scores from "successful applicants" and be lower than the average of "accepted applicants"? Are successful applicants and accepted applicants different?

    Berkeley has minimums suggested as 540/580, but they also have 560 for analytical so those scores might be from the old kind of tests. And cornell says 80%/90% for minimum suggested.
  20. Oct 12, 2005 #19

    Tom Mattson

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I don't know about this "successful" vs "accepted issue, but:

    You should talk to your specific departement of interest as well. School-wide averages aren't going to be useful to everyone, especially technical students. For instance Berkeley has one of the top physics departments in the US. I'm sure they wouldn't even consider a 580 in math. but that's the average you get after factoring in a few thousand communications majors.
  21. Oct 12, 2005 #20
    Given that the GRE has quite a bit of correlation with the SAT, I would venture to guess that you didn't do that well on the SAT and are probably not in a top-ranked undergrad program?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook