how are schools comparing old and new gre scores?

[applyingtograd]

---

Members don't see this ad.

I'm guessing that schools are going to be receiving some apps with old gre scores and some with new. How will schools compare the scores? will they convert new scores to old? I'm am taking the new gre in a couple of weeks and on the new scale my verbal and math scores are equivalent, but on the old scale my verbal score is much lower than math. I am trying to decide if I should spend the next couple of weeks concentrating on the verbal or if it would be more prudent to divide my time equally between quant and verbal.

thanks so much for the advice!

 

[mewtoo]

From what I've gathered I think they are basically just going to convert them until they get used to the new scale in a few years. Most schools seem to really just look at the combined scores. I wouldn't worry about it unless it was a huge difference, something like 800 Q and 400 V or something. If your combined score was where you want it, then don't worry about it. If your wanting to raise it, I've always heard to concentrate on the quant section because its much easier to raise that score than the verbal.

I wouldn't worry about the conversions. The scores I've gotten on practice tests the Quant is lower than my verbal by like 10 points, but when you convert it to old scale my quant is actually higher than my verbal by 30 points. The conversion is all sorts of wonky. Personally, I'd just worry about raising my score in general and not worry about trying to make the scores equal.

 

[psych4]

I am very curious about the comparison of scores as well. It is too bad that we cannot get a few responses from the individuals evaluating these applications. 🙁

 

[AcronymAllergy]

My guess is it's going to be a mix of using the conversion tables provided by ETS, as well as direct comparisons of the percentile ranks.

 

[thewesternsky]

I agree with AA... My impression is that most faculty are using percentiles to gauge performance on the new test.

 

[Ollie123]

Also fairly certain they will be relying on percentiles.

Its not surprising score comparisons between the quant and verbal are so discrepant. I believe they had to completely redo how they were scoring quant because the old one was wayyyyy too easy and had such massive ceiling effects that anyone going into a math-heavy field was pretty much expected to have an 800. You could have a lower score on the verbal and a much higher percentile since I think the quant percentile capped in the 80's even for an 800. Keep in mind that "easiness" didn't necessarily help people - you are compared to others so scores were inflated but percentiles were lower.

I wouldn't worry too much - schools realize what is going on and that the new test is meant to better align quant and verbal scores. A previously higher math score should be closer to the verbal on the new one. The percentiles will presumably be more directly comparable, though hopefully with less of a ceiling effect for the new one. If you are on the cusp of meeting cutoffs I guess it might be concerning since they might have department admin assistants doing the screening and they may not be thinking in that way, but I imagine departments will be keeping a close eye on it for the time being.

 

[EmotRegulation]

I am very curious about the comparison of scores as well. It is too bad that we cannot get a few responses from the individuals evaluating these applications. 🙁

As a faculty member who does review applications, I can say that our program does do the comparison of old scores to new and we only look at combined scores. We actually don't really look at percentages. We are aware of the differences in percentiles when converting scores, so some of that might get taken into consideration when reviewing files, but it's really more of a gestalt thing than a nitpicky examination of GRE scores.

 

[psych4]

As a faculty member who does review applications, I can say that our program does do the comparison of old scores to new and we only look at combined scores. We actually don't really look at percentages. We are aware of the differences in percentiles when converting scores, so some of that might get taken into consideration when reviewing files, but it's really more of a gestalt thing than a nitpicky examination of GRE scores.

Thank you very much for your insight.

 

[mixolyd]

Also fairly certain they will be relying on percentiles.

Its not surprising score comparisons between the quant and verbal are so discrepant. I believe they had to completely redo how they were scoring quant because the old one was wayyyyy too easy and had such massive ceiling effects that anyone going into a math-heavy field was pretty much expected to have an 800. You could have a lower score on the verbal and a much higher percentile since I think the quant percentile capped in the 80's even for an 800. Keep in mind that "easiness" didn't necessarily help people - you are compared to others so scores were inflated but percentiles were lower.

I wouldn't worry too much - schools realize what is going on and that the new test is meant to better align quant and verbal scores. A previously higher math score should be closer to the verbal on the new one. The percentiles will presumably be more directly comparable, though hopefully with less of a ceiling effect for the new one. If you are on the cusp of meeting cutoffs I guess it might be concerning since they might have department admin assistants doing the screening and they may not be thinking in that way, but I imagine departments will be keeping a close eye on it for the time being.

I just took the GRE today and got 159Q (750, 77th) and 166V (700, 96th). It does seem like verbal is still the more difficult subject for most. I personally found the math to be somewhat difficult, and I was incredulous when I saw my score.

What's strange is that the Quantitative scores only go up to 166, which is converted as an 800. So it's not possible to get a perfect math score on the new scale?

http://www.ets.org/s/gre/pdf/concordance_information.pdf

 

[AcronymAllergy]

I had thought both scales went up to 170 on the new test. Perhaps the table only lists up to 166 because that's where it tops out in comparison to the old test (i.e., at the 94th percentile), whereas anything above 166 wouldn't necessarily coincide with an old test score (essentially being an 800+), as the percentile might climb all the way to 99th.

Just my guess.

 

[applyingtograd]

I just took the GRE today and got 159Q (750, 77th) and 166V (700, 96th). It does seem like verbal is still the more difficult subject for most. I personally found the math to be somewhat difficult, and I was incredulous when I saw my score.

What's strange is that the Quantitative scores only go up to 166, which is converted as an 800. So it's not possible to get a perfect math score on the new scale?

http://www.ets.org/s/gre/pdf/concordance_information.pdf

they made the math section harder in the new gre so it is possible to get a perfect score of 170 on the new gre, but if someone had gotten a perfect score (800) on the old gre, it would be converted to a 166 on the new scale. does that make sense? Your overall score is really good though. An old score of 1450 is above every single average for any school...so I don't think you should worry. 🙂

 

[phillips101]

As a faculty member who does review applications, I can say that our program does do the comparison of old scores to new and we only look at combined scores. We actually don't really look at percentages. We are aware of the differences in percentiles when converting scores, so some of that might get taken into consideration when reviewing files, but it's really more of a gestalt thing than a nitpicky examination of GRE scores.

I find it interesting that some schools don't compare the percentages. I thought the whole point of having percentiles is to be able to compare across test-takers. To me (and I may be wrong), that's like saying 2 people earning a 4.0 are viewed equally even though one person got a 4.0 on a 4.0 scale, and another got a 4.0 on a 6.0 scale. This is just for discussion purposes only, I'm not trying to incite anything.

 

[AcronymAllergy]

I find it interesting that some schools don't compare the percentages. I thought the whole point of having percentiles is to be able to compare across test-takers. To me (and I may be wrong), that's like saying 2 people earning a 4.0 are viewed equally even though one person got a 4.0 on a 4.0 scale, and another got a 4.0 on a 6.0 scale. This is just for discussion purposes only, I'm not trying to incite anything.

I agree with being surprised that programs might not be comparing percentiles. However, just to clarify a bit--it's actually the standardization of the test that predominantly allows you to compare scores on the same test across individuals, which is the concept you're getting at with the GPA example (e.g., that one 4.0 doesn't necessarily = another 4.0). The percentiles allow you to better compare the new test scores to old test scores if the underlying distributions of the test scores are different

 

[Pragma]

FWIW, my mentor told me that they only look at percentiles at my old program. The scores themselves don't mean a lot depending on your form.

 

[phillips101]

One big discrepancy that stands out to me between comparing percentiles and standardized scores (the 166 or 800 itself) is between the verbal and quant on the old exam. I don't know a lot about the new exam, so can't say much for that. But for example, on the old exam, a 700 verbal's percentile is approximately (+/- 3 percentiles) the same as the quant 800. So if you look at the number only, there's a huge difference between 700 vs. 800, but if you look at percentile, they are practically the same. But that is within subjects comparison. As far as between subjects, yes you can say that an 800 on one exam standardized to be approximately an 800 on the next. I don't know much about the new vs. old exam.