Study and Strategy questions relating to the GMAT.
abhishek.gautam01
Course Students
 
Posts: 8
Joined: Fri Jul 24, 2009 7:37 am
 

Scores - Trend

by abhishek.gautam01 Sun Oct 11, 2009 2:55 pm

Hi,

I am taking the GMAT on 14th Oct'09. I have given some mock GMAT tests. The scores are listed below:

MGMAT 1: 680
MGMAT 2: 660
MGMAT 3: 700
MGMAT 4: 690

GMAT Powerprep 1: 680 (2months ago...)
GMAT Powerprep 2: 690 (taken today) Q49,V34

I was struggling with the timing on the Verbal section and had to guess the last 3 questions blindly... though I was able to complete the test. 4 incorrect in the last 5 :( Can something be done this late now?

Is there a chance that I can get past 720 on the actual exam on the D day??

Thanks

Abhishek
StaceyKoprince
ManhattanGMAT Staff
 
Posts: 9361
Joined: Wed Oct 19, 2005 9:05 am
Location: Montreal
 

Re: Scores - Trend

by StaceyKoprince Sat Oct 17, 2009 1:59 pm

I'm sorry I didn't get to this until you took your test. I'll still answer for the benefit of any others who find themselves in a similar situation. (FYI for future: it typically takes 3-7 days to get a response in this folder.)

You're scoring in the high-600 range right now. The standard deviation on GMATPrep tests is about 30 points and the SD on our tests is about 50 points, so technically, you are in the general range of mid-600 to low-700 right now.

If you are guessing blindly at the end and answering four of the last five incorrectly, that's easily costing you 20 points right there. So you do need to fix this problem before you take the test.

Here's the thing: you are ALWAYS going to have to guess on some problems, no matter how good you get. The ONLY choice you have is: on which questions are you going to guess? If you don't cut yourself off when you hit really hard questions earlier in the test, then you are going to be forced to guess at the end.

So, which would you rather do? Cut yourself off as you see especially hard questions on the test (which you're probably not going to get right anyway)? Or spend too much time when you see especially hard questions and then be forced to guess at the end (and you would probably be able to do at least some of those problems at the end, if you had time)?

No-brainer. Identify the hardest questions as you see them, spend normal (NOT EXTRA) time, then guess if you have to and keep going!
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep
tomslawsky
Students
 
Posts: 137
Joined: Wed Jul 15, 2009 12:07 pm
 

Re: Scores - Trend

by tomslawsky Sat Nov 07, 2009 10:41 am

"The standard deviation on GMATPrep tests is about 30 points and the SD on our tests is about 50 points, so technically, you are in the general range of mid-600 to low-700 right now."

Can you please explain the difference between SD and standard error of the mean, as you are using SD here. I'm a tad confused. Than you.
StaceyKoprince
ManhattanGMAT Staff
 
Posts: 9361
Joined: Wed Oct 19, 2005 9:05 am
Location: Montreal
 

Re: Scores - Trend

by StaceyKoprince Sat Nov 07, 2009 7:01 pm

There are a couple of different measurements that we can talk about.

First, there is something called a "within-student standard deviation." If one student could somehow take the test 100 times in a row, with no changes in ability - and without his/her head exploding :) - then we can calculate a standard deviation. One SD is about 50 points on our test. (On the resl test,

There is also a another measure that can be called one of two things (they're mathematically equivalent): "standard error of measurement" or "standard deviation of the error of the measurement." This is the SD of the difference between a student's last practice test score and the same student's score on the official test. That also happens to be about 50 points on our test. (Note: our tests are also "unbiased," meaning that over the entire population of test takers, the standard difference is zero. Our tests don't consistently skew high or low.)

When I wrote the last post, I was thinking about the "within-student standard deviation" - meaning that that student's "ability"* is in that general range, as best we can tell according to the test results.

* (An aside: I always feel as though I have to put "ability" in quotes because, really, what are standardized tests actually testing? :)
Stacey Koprince
Instructor
Director, Content & Curriculum
ManhattanPrep