GRE Concordance Table???

[NeuroBatman]

I have a question regarding the GRE concordance table.  My confusion stems from the fact that I have been repeatedly told that I have a low Quant GRE score (152), yet when I look it up on the concordance table (https://www.ets.org/s/gre/pdf/concordance_information.pdf) it is equivalent to a 670 on the old scale, which according to the 2012/2013 'Insiders Guide to Clinical Graduate Programs' is barely below the mean for scores of students accepted by top-tier schools ( x̅=676), and well within the standard deviation(SD=38).  So is my score really that bad?  I would appreciate any input.  Cheers!

Edited September 1, 2014 by justinhayes1982

[TakeruK]

I would personally advise against mixing new scores with old scores since the old scores have not been issued since July 2011, which is over 3 years ago now. This means that any document that discusses old score cutoffs are at least 3 years old. 

 

The concordance table says 152 in Q is 48th percentile. But what "low" means is subjectively. In the grand scheme of things, yes, your score is "low"--by definition, you are in the bottom half of all test takers (but just barely). However, the pool of "all GRE test takers" can be very different from the pool of "applicants to the same program as you". In my field, this would be a very low score that would raise a lot of flags in your application, but your field may not be this way.

 

Maybe someone from your field can comment on your score. Have the people who told you your score was low in the past been from your field (or is knowledgeable about your field)?

[NeuroBatman]

I understand what you are saying.  Yes some folks in my field have commented on my score being low, yet if you look at historical trends (mean scores on admissions sites, program guides, etc) , 1200-1250 has often been an acceptable comprehensive score (using the old scale) in my field and my scores are equivalent to 1350 (165V + 152Q).  It seems that people are saying that standards have risen significantly since the implementation of the new test, suggesting that the applicant pool has quickly become more qualified.  It seems possible but unlikely.  Do you see where my confusion lies?  Perhaps it's a matter of, 'are these concordance tables really accurate when comparing scores?'

[TakeruK]

If you didn't know, the Concordance Tables are updated every year so that the scaled score-to-percentile-rank-conversion changes each year by a little bit. ETS recomputes what a "152" means every year based on performance of test takers in the last 3 years. For example, the most recent concordance table uses test scores from April 2011 to April 2014. So, this is why I warned against mixing old standards (e.g. "1200-1250") with "translated" new scores. Even if the "1200-1250" standard was published in early 2011, they would likely be basing it off the 2008-2011 Concordance Table, which can be different enough from the current 2011-2014 Concordance Table. Not only there is a 3 year gap, but the most recent Concordance Table is the first table that is computed from *almost* entirely Revised GRE scores. 

 

In addition, there were many changes to the test after implementation of the Revised GRE, so it's not really an equal comparison any more. ETS will continue to publish the Concordance Tables with the old and Revised GRE scores until 2016 probably, since GRE scores are valid for 5 years so in theory people can still submit old GRE scores until July 31, 2016. Note also that the Concordance Table is really meant to convert "old scores" to "new scores", not the other way around! You can tell this by looking at where the "new" scores end--the highest GRE Q score listed is 166, not 170 (the actual maximum). 

 

Another difference in the old/new scoring is that the new score is more finely graded to allow for better distinction at the high scores end. For example, when the test was implemented, 6% of all test takers scored 800/800 on the old GRE Q (that is, the maximum score of 800 is 94th percentile). Now, you see that 800/800 is actually only 92nd percentile, which does show that the applicant pool in general is scoring higher than before (perhaps due to the new format? who knows for sure). However, you'll also see that 800/800 on the old test is only 166/170. This means the new test has 4 new scoring ranks (167,168,169,170) that are better able to distinguish top scorers (although this probably does not matter!). But basically, the "addition" of the extra ranks means you are not really comparing things on the same scale when you mix the two scoring systems.

 

So, in summary, I think you should not be using Concordance Tables to convert your new score to the old score system for all the reasons listed above. The Concordance Table are meant to go from old->new, not vice versa. Instead, this PDF shows the relevant stats for your score: https://www.ets.org/s/gre/pdf/gre_guide_table1a.pdf. As I said above, your percentile rank places you just below the median. The listed average GRE Q score is 152.14 with a standard deviation of 8.83 so I would say you are very close to the average score too. Finally, I would also strongly recommend you take the advice of people in your field if you trust them and their wisdom/experience (especially your supervisors and mentors or professors in your current department sitting on your school's admissions committee).

[NeuroBatman]

Cool.  Thank you for clearing that up.  Much appreciated.  : )