Page 1 of 1

Convert old scores to new scale?

Posted: Sun Oct 25, 2009 8:09 pm
by joey
I took both practice exams with the rescaled scores at the end, and now I'm onto tests from before the switch. Since I'm shooting for a particular score, I'd like to know how I'm doing after each practice test. Does anybody know a good way to convert old scores to the new scale? My instinct is to compare percentiles, but that assumes the applicant pool hasn't changed since they rescaled. I'm not sure how the pool now compares to the pool when 8767 was given, for instance. Any ideas?

Re: Convert old scores to new scale?

Posted: Sun Oct 25, 2009 10:27 pm
by diogenes
Hard to say, but it seems to me that the test has changed.
So, increase those error bounds. :shock:

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 12:43 pm
by mtey
diogenes wrote:Hard to say, but it seems to me that the test has changed.
What exactly has changed in the test? I'm asking because I have more scaled math practice tests than rescaled ones... so I won't know what to expect on my actual exam.

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 1:15 pm
by joey
mtey wrote:
diogenes wrote:Hard to say, but it seems to me that the test has changed.
What exactly has changed in the test? I'm asking because I have more scaled math practice tests than rescaled ones... so I won't know what to expect on my actual exam.
I seem to remember reading that the contents hadn't changed, just the scale. It's a dim memory, though, and I can't seem to find where I read it.

Incidentally, where did you find all the old tests? I've only found two from before they rescaled.

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 1:26 pm
by diogenes
mtey wrote: What exactly has changed in the test? I'm asking because I have more scaled math practice tests than rescaled ones... so I won't know what to expect on my actual exam.
If I knew exactly, I would know the test. :wink:

But just look at the most recent test and compare it to the older ones; the trend appears to be away from mechanical calculations towards more conceptual problems.
I seem to remember reading that the contents hadn't changed, just the scale. It's a dim memory, though, and I can't seem to find where I read it.
I think you might be referring to one of the older released tests that was re-released under the newer scaling.

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 1:35 pm
by mtey
joey wrote:Incidentally, where did you find all the old tests? I've only found two from before they rescaled.
Well, I have the book of Morris Bramson where there are 5 tests which resemble the old scaled math test. In fact if you are interested in the book pm me ;)
diogenes wrote:But just look at the most recent test and compare it to the older ones; the trend appears to be away from mechanical calculations towards more conceptual problems.
diogenes, thanks for the comment, I hope that the theory in "Cracking the GRE" will be enough for my target score on the new test, because I do not have the time for reviewing all of the topics thoroughly.

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 7:45 pm
by joey
diogenes wrote:
Joey wrote:I seem to remember reading that the contents hadn't changed, just the scale. It's a dim memory, though, and I can't seem to find where I read it.
I think you might be referring to one of the older released tests that was re-released under the newer scaling.
Ah! It turns out I really did read it. See the bottom right of page 5 here. Despite this, I think you may be right about the tests changing, diogenes. The old tests strike me as easier, and ETS talks about revitalizing exams annually in its Practice Book.
mtey wrote:diogenes, thanks for the comment, I hope that the theory in "Cracking the GRE" will be enough for my target score on the new test, because I do not have the time for reviewing all of the topics thoroughly.
I'd be wary of that book. It contains many errors that have been caught, and an untold number of errors that haven't. Worse, it doesn't seem to be at the level of the real exam. For instance, linear transformations are always treated as endomorphisms of Euclidean space, while their analogues across generic vector spaces are more important on the actual test.

Personally, I won't study from Cracking the GRE at all. It's too unpredictable. I can't risk learning something false from the book and having it bring down my score. Instead, I work with old exams and study up on the questions I missed--using my undergrad texts. This focuses my studying on topics I'm likely to see and miss on the real exam, and it keeps me constantly aware of the level of rigor required. Since you don't have a lot of time, you might want to consider such a concentrated approach yourself.

Re: Convert old scores to new scale?

Posted: Tue Oct 27, 2009 9:01 pm
by diogenes
joey wrote:
Personally, I won't study from Cracking the GRE at all. It's too unpredictable. I can't risk learning something false from the book and having it bring down my score. Instead, I work with old exams and study up on the questions I missed--using my undergrad texts. This focuses my studying on topics I'm likely to see and miss on the real exam, and it keeps me constantly aware of the level of rigor required. Since you don't have a lot of time, you might want to consider such a concentrated approach yourself.
I think Cracking is good place to start, but not to finish.
There are a few errors, but you can get an errata from the publisher.

I would also look at "The Art and Craft of Problem Solving" by Zeitz if you have a lot of time.