MCAT Equivalency Table

[Monkitty]

I recieved an email from my pre-health adviser yesterday with a table showing the equivalencies for old -> new MCAT. Figured some would find this useful.

Overall Score
New MCAT / %tile / Old MCAT
528 100 45
527 100 44
526 100 43
525 100 42
524 100 41
523 100 40
522 99 39
521 99 38
520 98 37
519 98 37
518 97 36
517 96 35
516 95 34
515 93 34
514 91 33
513 89 32
512 87 32
511 85 31
510 83 31
509 80 30
508 77 30
507 74 29
506 71 29
505 68 28
504 64 27
503 61 27
502 57 26
501 54 26
500 51 25
499 47 25
498 43 24
497 40 24
496 37 23
495 34 22
494 31 22
493 28 21

Biology Score
New MCAT / %tile / Old MCAT
132 100 15
131 99 14
130 97 13
129 93 12
128 87 11
127 77 10
126 67 10
125 54 9
124 44 8
123 32 7
122 21 6
121 14 5
120 7 4
119 3 3
118 1 1

Physical Science Score
New MCAT / %tile / Old MCAT
132 100 15
131 99 14
130 97 13
129 93 12
128 87 11
127 79 10
126 67 9
125 55 8
124 45 7
123 32 7
122 21 6
121 12 5
120 7 4
119 3 3
118 1 1 & 2

---

Members don't see this ad.

 

[Goro]

Interesting. The numbers here are slightly different from the chart posted by @efle .


What was your advisor's source?

 

[Monkitty]

Interesting. The numbers here are slightly different from the chart posted by @efle .


What was your advisor's source?

Unknown. She sent it as an Excel sheet which shows herself as the author. I don't know if it's something my institution compiled themselves or taken from a list-serve or just want.

If you'd like to know more specifics @Goro, I can private message you.

 

[efle]

This has some small errors

For examples, see here that 39 is 100th percentile not 99th, a 14 in PS is 100th percentile not 99th...the way it tries to display equivalency is also a little sloppier/less informative than my table in my (obviously biased) opinion.

All my data was taken off the two official/latest percentile releases, can compare your excel sheet to mine here

 

[GypsyHummus]

Interesting. The numbers here are slightly different from the chart posted by @efle .


What was your advisor's source?

I had a chance to talk with a DO school today, and they said while they were expecting ranges in the 503 502, they have had a lot of people with 498 applying. Apparently the test is a lot tougher than the old one, so they are bending on the bar a bit as a 502 is not quite the equivalent of a 26, and that even a 497 would be "good enough". Has your school experienced anything like this, Goro?

 

[Goro]

I'll have to ask the wily old Admissions dean!

I had a chance to talk with a DO school today, and they said while they were expecting ranges in the 503 502, they have had a lot of people with 498 applying. Apparently the test is a lot tougher than the old one, so they are bending on the bar a bit as a 502 is not quite the equivalent of a 26, and that even a 497 would be "good enough". Has your school experienced anything like this, Goro?

 

[Dreamstoo]

I had a chance to talk with a DO school today, and they said while they were expecting ranges in the 503 502, they have had a lot of people with 498 applying. Apparently the test is a lot tougher than the old one, so they are bending on the bar a bit as a 502 is not quite the equivalent of a 26, and that even a 497 would be "good enough". Has your school experienced anything like this, Goro?

Yeah, I think 502 = 26, 498 = 24, 503 = 27, 497 = 23-24 about.

Wait, sorry I may have misunderstood this. Do you mean they are considering a 502 to be higher than a 26?

 

[GypsyHummus]

Yeah, I think 502 = 26, 498 = 24, 503 = 27, 497 = 23-24 about.

Wait, sorry I may have misunderstood this. Do you mean they are considering a 502 to be higher than a 26?

No, you are correct that a 502 approx equals a 26, according to AAMC. What they were telling me is that there are not a lot of people with 502 applying, and they expect that they are gonna have to accept a lot of people in the 498 range. The list of it is that the new MCAT is much more difficult than the old one.

This is just one school though, who knows how things will actually pan out.

 

[Dreamstoo]

No, you are correct that a 502 approx equals a 26, according to AAMC. What they were telling me is that there are not a lot of people with 502 applying, and they expect that they are gonna have to accept a lot of people in the 498 range. The list of it is that the new MCAT is much more difficult than the old one.

This is just one school though, who knows how things will actually pan out.

Oh ok, thank you! Oddly enough, I heard something similar from a former adcom a few years ago. He predicted that med schools may be more lenient with the scores for the first two years the new scores come out. But I'm sure it varies from school to school.

 

[kinzav]

No, you are correct that a 502 approx equals a 26, according to AAMC. What they were telling me is that there are not a lot of people with 502 applying, and they expect that they are gonna have to accept a lot of people in the 498 range. The list of it is that the new MCAT is much more difficult than the old one.

This is just one school though, who knows how things will actually pan out.

But why? It's a curved test. If everyone did worse you would still come out with the same amount of people with the same score equivalents (assuming the same number of test takers). if the application averages are going down it has to mean that worse scores are applying at higher end schools and will undoubtedly overflow their spots at their schools and get denied, and end up applying to lower end schools eventually. I mean, it doesn't make sense that a curved test would disrupt the average applicant numbers for any schools, it'd be something else at work.

 

[PKA<0]

^^ absolutely agree, it's a standardized exam where percentiles are based off of the group that took the test. I don't see why schools would start taking lower percentiles.

 

[GrapesofRath]

^^ absolutely agree, it's a standardized exam where percentiles are based off of the group that took the test. I don't see why schools would start taking lower percentiles.

Absolutely this and after really looking at both tests yes the new test being longer causes difficulty but beyond that I think the difficulty of the new one compared to the old one is overblown to a fair extent.

 

[efle]

Absolutely this and after really looking at both tests yes the new test being longer causes difficulty but beyond that I think the difficulty of the new one compared to the old one is overblown to a fair extent.

For the difficulty to change, the ability of the testing population would have to change. If anything this first round of tests will be easier to hit very high numbers, since a lot of gunners that wanted to study their way 38+ rushed to take the old one

 

[GrapesofRath]

For the difficulty to change, the ability of the testing population would have to change. If anything this first round of tests will be easier to hit very high numbers, since a lot of gunners that wanted to study their way 38+ rushed to take the old one

Speculating on little things like "oh I feel like alot of smart overachieving people wanted to take the old test and get it over with" isn't really worthwhile unless there is really something specific to ground those claims in. You could just as easily argue the competition is more fierce now than it was a couple years ago and hence as schools average stats go up and the students are smarter and smarter the competition to get top scores on the bell curve is tougher.

Both tests play to different peoples strengths.
a) The CARS on the new MCAT allows more time per question than the old MCAT. Now perhaps the passages are a little longer(although the AAMC didn't explicitly say so) but all in all many will say you have more time on verbal for this new test.
b) The psychology and soc section seems to integrate a ton of definitions into questions. Many will tell you if you put in the time to memorize all these definitions familiarize yourself with research in these fields and have a background in these areas it will be of advantage to you. Also there are skills in CARs that translate to this section.
c) The new biological section seems to emphasize interpreting data and research studies and designs. Those with extensive research experience and reading research literature are at an advantage here. As are those who have taken alot of biochem related classes.
d) The physical sciences integrates alot of chem and physics in biochem settings. All in all many such as Berkeley review have said the section really hasn't changed as much as people might have thought. This seems to be a section where learning material and practicing alot of problems pays huge dividends and the material might be more "learnable" than skills needed for bio or CARs.

 

[GrapesofRath]

Edit

 

[GrapesofRath]

For the difficulty to change, the ability of the testing population would have to change. If anything this first round of tests will be easier to hit very high numbers, since a lot of gunners that wanted to study their way 38+ rushed to take the old one

Speculating on little things like "oh I feel like alot of smart overachieving people wanted to take the old test and get it over with" isn't really worthwhile unless there is really something specific to ground those claims in. You could just as easily argue the competition is more fierce now than it was a couple years ago and hence as schools average stats go up and the students are smarter and smarter the competition to get top scores on the bell curve is tougher. Its all subjective.

Both tests play to different peoples strengths.
a) The CARS on the new MCAT allows more time per question than the old MCAT. Now perhaps the passages are a little longer(although the AAMC didn't explicitly say so) but all in all many will say you have more time on verbal for this new test.
b) The psychology and soc section seems to integrate a ton of definitions into questions. Many will tell you if you put in the time to memorize all these definitions familiarize yourself with research in these fields and have a background in these areas it will be of advantage to you. Also there are skills in CARs that translate to this section.
c) The new biological section seems to emphasize interpreting data and research studies and designs. Those with extensive research experience and reading research literature are at an advantage here. As are those who have taken alot of biochem related classes.
d) The physical sciences integrates alot of chem and physics in biochem settings. All in all many such as Berkeley review have said the section really hasn't changed as much as people might have thought. This seems to be a section where learning material and practicing alot of problems pays huge dividends and the material might be more "learnable" than skills needed for bio or CARs.

 

[GrapesofRath]

Edit

 

[GrapesofRath]

I recieved an email from my pre-health adviser yesterday with a table showing the equivalencies for old -> new MCAT. Figured some would find this useful.

Overall Score
New MCAT / %tile / Old MCAT
528 100 45
527 100 44
526 100 43
525 100 42
524 100 41
523 100 40
522 99 39
521 99 38
520 98 37
519 98 37
518 97 36
517 96 35
516 95 34
515 93 34
514 91 33
513 89 32
512 87 32
511 85 31
510 83 31
509 80 30
508 77 30
507 74 29
506 71 29
505 68 28
504 64 27
503 61 27
502 57 26
501 54 26
500 51 25
499 47 25
498 43 24
497 40 24
496 37 23
495 34 22
494 31 22
493 28 21

Biology Score
New MCAT / %tile / Old MCAT
132 100 15
131 99 14
130 97 13
129 93 12
128 87 11
127 77 10
126 67 10
125 54 9
124 44 8
123 32 7
122 21 6
121 14 5
120 7 4
119 3 3
118 1 1

Physical Science Score
New MCAT / %tile / Old MCAT
132 100 15
131 99 14
130 97 13
129 93 12
128 87 11
127 79 10
126 67 9
125 55 8
124 45 7
123 32 7
122 21 6
121 12 5
120 7 4
119 3 3
118 1 1 & 2

The percentages match what the AAMC released in April and May. The difference between this and elfe is interpreting what percentiles convert to on the old test.

74th percentile could be argued as a 30 on the old test. 85th percentile could be argued as a 32 on the old test. 89th percentile could be argued as a 33. 95th percentile could be argued as a 35. These tables listed by the AAMC for the old test tend to list the highest percentile that a score could be ie(88th being 32). Hence it goes against what your advisor said slightly

https://www.aamc.org/students/download/430684/data/finalpercentileranksfortheoldmcatexam.pdf

In general its good to be on the conservative side of things in these estimations I wonder how ADCOMs will look at these. Will they look at 511 85th percentile as a 31, 32 or hybrid 31-32 in between?

 

[efle]

Speculating on little things like "oh I feel like alot of smart overachieving people wanted to take the old test and get it over with" isn't really worthwhile unless there is really something specific to ground those claims in. You could just as easily argue the competition is more fierce now than it was a couple years ago and hence as schools average stats go up and the students are smarter and smarter the competition to get top scores on the bell curve is tougher. Its all subjective.

Both tests play to different peoples strengths.
a) The CARS on the new MCAT allows more time per question than the old MCAT. Now perhaps the passages are a little longer(although the AAMC didn't explicitly say so) but all in all many will say you have more time on verbal for this new test.
b) The psychology and soc section seems to integrate a ton of definitions into questions. Many will tell you if you put in the time to memorize all these definitions familiarize yourself with research in these fields and have a background in these areas it will be of advantage to you. Also there are skills in CARs that translate to this section.
c) The new biological section seems to emphasize interpreting data and research studies and designs. Those with extensive research experience and reading research literature are at an advantage here. As are those who have taken alot of biochem related classes.
d) The physical sciences integrates alot of chem and physics in biochem settings. All in all many such as Berkeley review have said the section really hasn't changed as much as people might have thought. This seems to be a section where learning material and practicing alot of problems pays huge dividends and the material might be more "learnable" than skills needed for bio or CARs.

Any increase in the demands of the test are faced by all testers. For something based on your percentile like the MCAT, the only thing that matters when comparing exams is the quality of the test taking population. My belief is that there is negligible difference in that regard, but I do know many people rushed to take the old test out of a desire for something that could be more reliably studied for. My university told all students interested in taking our paid prep course to take it before the change since they are not offering it this first year. The AAMC also mentioned in their materials about the new MCAT that they expect the "curve" may change after the first year. I don't think there's any significant difference in test populations like I said...but if there was going to be a difference it would be towards a lot of the more careful planners (slash number ***** gunners) avoiding this first round of the new test.

 

[efle]

The percentages match what the AAMC released in April and May. The difference between this and elfe is interpreting what percentiles convert to on the old test.

74th percentile could be argued as a 30 on the old test. 85th percentile could be argued as a 32 on the old test. 89th percentile could be argued as a 33. 95th percentile could be argued as a 35. These tables listed by the AAMC for the old test tend to list the highest percentile that a score could be ie(88th being 32). Hence it goes against what your advisor said slightly

https://www.aamc.org/students/download/430684/data/finalpercentileranksfortheoldmcatexam.pdf

In general its good to be on the conservative side of things in these estimations I wonder how ADCOMs will look at these. Will they look at 511 85th percentile as a 31, 32 or hybrid 31-32 in between?

I did not make any direct conversions like this chart does, I merely listed the percentiles given per each score. People can decide how to go about collapsing my raw data / perfectly accurate table however they wish. The table listed here has some factual inaccuracies regarding the percentiles however, regardless of the attempt at conversion.

 

[GrapesofRath]

I did not make any direct conversions like this chart does, I merely listed the percentiles given per each score. People can decide how to go about collapsing my raw data / perfectly accurate table however they wish. The table listed here has some factual inaccuracies regarding the percentiles however, regardless of the attempt at conversion.

It's interpreting the percentiles where the difference is. Like I said you might have 95th percentile as a 35 while the OP has it as a 34. But the 516 is still what's the 95th percentile regardless; it's just how you interpret it under converting it to the old testing scale.

 

[efle]

It's interpreting the percentiles where the difference is. Like I said you might have 95th percentile as a 35 while the OP has it as a 34. But the 516 is still what's the 95th percentile regardless; it's just how you interpret it under converting it to the old testing scale.

But there is no ambiguity in what the percentiles per each score are on either test. There is no way for both 34 and 35 to have a bin ceiling at the 95th percentile. The method I used for the other table is the only way to keep all relevant info clear - just list each of the descending percentiles with the scores in their respective spots. Shows exactly how much scores overlap / whether things are exactly equivalent or have ceilings a few apart

 

[GrapesofRath]

But there is no ambiguity in what the percentiles per each score are on either test. There is no way for both 34 and 35 to have a bin ceiling at the 95th percentile. The method I used for the other table is the only way to keep all relevant info clear - just list each of the descending percentiles with the scores in their respective spots. Shows exactly how much scores overlap / whether things are exactly equivalent or have ceilings a few apart

It's not ambiguous but part of the problem is that for recent MCATs from 2012-2014 they didn't list the range of percentiles for each score. They just listed it as number. The last AAMC stats I could find on range of percentiles for each score is 2011

https://www.aamc.org/students/download/264234/data/combined11.pdf

If we follow this data then clearly the OP's post has some minor flaws to it.

 

[efle]

It's not ambiguous but part of the problem is that for recent MCATs from 2012-2014 they didn't list the range of percentiles for each score. They just listed it as number. The last AAMC stats I could find on range of percentiles for each score is 2011

https://www.aamc.org/students/download/264234/data/combined11.pdf

If we follow this data then clearly the OP's post has some minor flaws to it.

The listed values are the bin ceilings. The range is from the percentile ceiling listed one score below up to the percentile listed beside. They state it's such / that they are listing ceilinfs with their definition of percentile as "proportion scoring lower or equal to"

 

[GrapesofRath]

The listed values are the bin ceilings. The range is from the percentile ceiling listed one score below up to the percentile listed beside. They state it's such / that they are listing ceilinfs with their definition of percentile as "proportion scoring lower or equal to"

Yeah I'm aware. It's just it seems to have confused people when they don't list ranges anymore like the OP's adviser.

 

[efle]

Yeah I'm aware. It's just it seems to have confused people when they don't list ranges anymore like the OP's adviser.

Yeah they also were off at the high end, looking at the bars in the histograms it's really 43-45 = 528 (nothing visible) 42 = 527 (barely visible baby bump if you zoom and enhance like CSI) 41 = 526 (first visible bump). Saying 523 = 40 waaay too generous, far more 523+ now than 40+ back then

Overall gotta say it is a pretty lazy chart that needlessly confuses and is pretty sloppy with some approximations. Much prefer the full dataset style. OP send my **** to your adviser 😛

 

[Monkitty]

I'm sure they would chuck your **** into the trashbin with all the other pre-med **** they get 🙂

I didn't realize you (or anyone) had posted a table before (considering you posted it in pre-Allo and I'm pretty exclusively a DO kid, that explains why I didn't find it). I just thought this might be beneficial for folks. Again I don't how they calculated this, maybe they somehow based it off those from my university? I dunno.

Just the messenger 🙂

 

[jd989898]

Yea I agree, the chart is lazily made as it's obvious the person exclusively based it simply off of equivalent whole number percentiles. Which, as @efle said, leads to significant misrepresentation of the top score's %iles in particular. Far more data and scores are grouped into the "100th" percentile on the new test than on the old one.

 

[Mad Jack]

But why? It's a curved test. If everyone did worse you would still come out with the same amount of people with the same score equivalents (assuming the same number of test takers). if the application averages are going down it has to mean that worse scores are applying at higher end schools and will undoubtedly overflow their spots at their schools and get denied, and end up applying to lower end schools eventually. I mean, it doesn't make sense that a curved test would disrupt the average applicant numbers for any schools, it'd be something else at work.

The MCAT is NOT A CURVED TEST. For the love of God, how many times must that be repeated?

The MCAT is normalized. That means that they use a predictive algorithm to build a test that will come out naturally curved, in theory. But in practice, if you had a particularly unprepared batch of test takers, everyone could do more poorly than the algorithm predicted and end up with a score of, say, 502 or less. After this point, they will go back and look at the test and see if there were any algorithmic errors or if any questions were unfair (this is the reason it takes so long to get your computerized test back- they are looking for errors and anomalies in the test and ensuring you get the most fair results possible). It could, however, be determined that the test was, in fact, fair and that no corrections need to be made to the scoring formula, thus leaving an entire administration falling below the curve care of their lack of preparation.

You are not scored against those actually taking the test. You are scored against a hypothetical pool of candidates derived from all prior administrations of similar questions and material, as well as some ungraded groups that tested the new material so that the AAMC could gain data prior to the 2015 test starting. It isn't a curve. It's a data-derived algorithm that relies on people to fall within a certain bell curve of knowledge. The people taking a given test don't create the curve, the curve already exists and your score determines your place within it.

 

[efle]

The MCAT is NOT A CURVED TEST. For the love of God, how many times must that be repeated?

The MCAT is normalized. That means that they use a predictive algorithm to build a test that will come out naturally curved, in theory. But in practice, if you had a particularly unprepared batch of test takers, everyone could do more poorly than the algorithm predicted and end up with a score of, say, 502 or less. After this point, they will go back and look at the test and see if there were any algorithmic errors or if any questions were unfair (this is the reason it takes so long to get your computerized test back- they are looking for errors and anomalies in the test and ensuring you get the most fair results possible). It could, however, be determined that the test was, in fact, fair and that no corrections need to be made to the scoring formula, thus leaving an entire administration falling below the curve care of their lack of preparation.

You are not scored against those actually taking the test. You are scored against a hypothetical pool of candidates derived from all prior administrations of similar questions and material, as well as some ungraded groups that tested the new material so that the AAMC could gain data prior to the 2015 test starting. It isn't a curve. It's a data-derived algorithm that relies on people to fall within a certain bell curve of knowledge. The people taking a given test don't create the curve, the curve already exists and your score determines your place within it.

We don't know this to be the case for the first testing pool, only that what you speak of was the old method. Personally I don't think they could possibly have a predictive algorithm to make such a nice bell, and waited to get the April/may data and then fit a curve to it.

 

[efle]

I'm sure they would chuck your **** into the trashbin with all the other pre-med **** they get 🙂

I didn't realize you (or anyone) had posted a table before (considering you posted it in pre-Allo and I'm pretty exclusively a DO kid, that explains why I didn't find it). I just thought this might be beneficial for folks. Again I don't how they calculated this, maybe they somehow based it off those from my university? I dunno.

Just the messenger 🙂

Well you don't have to tell them a premed made it! Just say here's a table of the AAMC percentile and score values that a UC adcom said is identical to what they built to use this cycle...

 

[Mad Jack]

We don't know this to be the case for the first testing pool, only that what you speak of was the old method. Personally I don't think they could possibly have a predictive algorithm to make such a nice bell, and waited to get the April/may data and then fit a curve to it.

Regardless, the test has been in place long enough for them to get enough data from this point on. They also performed substantial testing on the new questions prior to their introduction in the MCAT, so they knew roughly how people would perform. There might have been some little things to be fixed here and there, but nothing amounting to a curve. For anyone moving forward, the test will no doubt be normalized, not curved.

 

[efle]

Regardless, the test has been in place long enough for them to get enough data from this point on. They also performed substantial testing on the new questions prior to their introduction in the MCAT, so they knew roughly how people would perform. There might have been some little things to be fixed here and there, but nothing amounting to a curve. For anyone moving forward, the test will no doubt be normalized, not curved.

Arguments made about it being curved are sometimes still valid though. Their method of normalization is essentially a curve assuming the test taking populations vary insignificantly between the two years. It makes people wrong to argue about times of the years or days being easier, but argument that being top X% is a static difficulty seems right to me , so it is weird for the expected percentiles to have dropped or risen