|
Author |
Message |
|
Current Student
Joined: 19 Jan 2009
Posts: 21
Followers: 0
Kudos [?]:
3
[0], given: 0
|
Same component scores, different overall score: what gives? [#permalink]
 01 Feb 2009, 09:55
Just out of curiosity... Can anybody explain why the same component scores yield different overall scores sometimes? Or why lower component scores even yield a higher overall score sometimes? I assume that this is because you are compared to your peers at the time of the exam, but then why don't you get a component score to match? A few examples: I've seen a Q50, V41 get a 740. I've also seen a Q50, V40 get a 750. I've seen lots of variations on these scores. I also saw a Q50, V41 get a 760. I'm satisfied with my personal score and am not complaining. I just want to understand how this system works. Thanks.
_________________
http://www.MilitaryToBusiness.com - Veterans / HBS Alumni consulting other veterans
|
|
|
|
|
|
|
Manager
Joined: 21 Oct 2008
Posts: 131
Schools: Rady School of Management at UC San Diego GO TRITONS
Followers: 2
Kudos [?]:
12
[0], given: 0
|
Re: Same component scores, different overall score: what gives? [#permalink]
02 Feb 2009, 11:23
Hey matterhorn, my understanding is that the composite score takes percentile into account, and since percentiles fluctuate over time, so to the scores. So I could take the GMAT in March, score a 710, and then people might start doing better on the GMAT, so my score would go down to a 700 by June. If you request a new score report, your composite score will be recalculated based on what it is at that time (i.e. your score report will look different in June than in March). Schools know this, and won't care if there's a slight discrepancy as long as you report your raw scores accurately.
|
|
|
|
|
|
SVP
Joined: 05 Aug 2007
Posts: 1502
Schools: NYU Stern '11
Followers: 15
Kudos [?]:
202
[0], given: 22
|
Re: Same component scores, different overall score: what gives? [#permalink]
02 Feb 2009, 13:26
Your score remains the same. A 710 will NEVER go down to a 700. A Q47 might go from being 81st p.c. to 77th p.c., but you will still have a 710. Of course, a 710 might have been 93rd p.c. or something at a point in time, and might only be 91st p.c. now. Matterhorn, that's a question that gets raised pretty often. No one quite knows how GMAC's scoring algorithm for the GMAT really works, so it's all speculative. But rjacobs is right in that you're compared to a pool comprising of all test takers within a specific time frame, and if a Q50 V41 is 99th p.c. when compared to that pool, then you get a 760, and if it's 97th p.c. then a 740 et cetera. The percentiles fluctuate all the time, albeit only slightly. rjacobs wrote: Hey matterhorn, my understanding is that the composite score takes percentile into account, and since percentiles fluctuate over time, so to the scores. So I could take the GMAT in March, score a 710, and then people might start doing better on the GMAT, so my score would go down to a 700 by June. If you request a new score report, your composite score will be recalculated based on what it is at that time (i.e. your score report will look different in June than in March). Schools know this, and won't care if there's a slight discrepancy as long as you report your raw scores accurately.
|
|
|
|
|
|
Current Student
Joined: 19 Jan 2009
Posts: 21
Followers: 0
Kudos [?]:
3
[0], given: 0
|
Re: Same component scores, different overall score: what gives? [#permalink]
02 Feb 2009, 15:33
That's not exactly true: A 50Q, 44V might be a 760 ( 8-t67974) or it might be a 770 ( 8-t74857). More amazingly, a 750 might be a 50Q, 40V ( 8-t66785) but a 50Q, 41V is a 740! ( 8-t73035). So my question is how can somebody get a higher total score with lower composite scores? Based on the above examples, a 50Q, 40V is better than a 50Q, 41V. Or more generally, how come the same composite scores yield different results? I've seen up to a 20 point spread for the exact same composite scores. For example, if you scored a 50Q, 41V it might be 760 for you ( http://www.beatthegmat.com/760-in-gmat- ... 14208.html) but 740 for somebody else ( 8-t73035). None of this is a big deal. Just curious as to how to interpret this data.
_________________
http://www.MilitaryToBusiness.com - Veterans / HBS Alumni consulting other veterans
|
|
|
|
|
|
Manager
Joined: 15 Aug 2008
Posts: 94
Schools: NYU Stern Class of 2012
WE 1: GM, Diversified Financial Svcs.
WE 2: Sales & Trading
Followers: 1
Kudos [?]:
10
[0], given: 3
|
Re: Same component scores, different overall score: what gives? [#permalink]
03 Feb 2009, 05:47
I remember reading somewhere ( OG11?) that your final composite score is also based on the difficulty of the questions you answered during the test. Apparently, two people can take the test side by side and, despite getting the same number of Qs right/wrong, one can have a considerably more "difficult" exam overall and the test adjusts for that. Hence a very difficult Q50,V41 might be adjusted higher than a relatively easy Q50,V42? Pure speculation of course - but makes some sense to me.
|
|
|
|
|
|
|
Re: Same component scores, different overall score: what gives?
[#permalink]
03 Feb 2009, 05:47
|
|
|
|
|
|
|
|
|
|
|
close
