How do admissions compare the two types of MCAT scores?

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

kanda1o

Full Member
7+ Year Member
Joined
Aug 5, 2015
Messages
120
Reaction score
58
So with more and more applicants applying with their new MCAT score (myself included), how do admissions compare? Is it solely based on percentile ranks?

I'm applying this June (16/17 cycle) and am curious. I feel as if a good portion of the applicant pool for this next cycle will still have the old score, on top of all the students who took the new MCAT. So if admissions see applicant A with a 28 and applicant B with a 501. What do these numbers truly mean to them.


I'm mainly asking out of curiosity, as this has been on my mind for a while now. But the physician I work under was asking me about it yesterday. I figured SDN may have the answers.
 
As some of the admissions ADCOMs have mentioned they are basing it off of percentiles for now. I speculate they will get a better idea when faculty of each school sees the performance of both cohorts of student on medical school exams. This may determine whether they will still be strictly based off of percentiles in the next 2 years (or year not sure) or how it correlates to performance in medical school.
 
When applying, my suggestion would be to go with the safe bet and assume they are comparing the old vs new with percentiles.

However, if you're wondering what the reasonable and fairest alternatives would be I'll share my thoughts (mind you, this is speculation, I'm just an applicant that was recently accepted and still have no clue how things actually go).

1. Take the first 3 sections of the new exam and omit the last one (Psych/Sociology) and use percentiles to individually convert them to an old scale score. For example a 128 in any of these sections would be about an 11 and a 125 would be an 8. Then combine the scores to get an old scale composite score (and its corresponding percentile). For instance someone who get a 128/126/129/127 on the new exam would have an 11 in Phys/Chem, a 9 in verbal (or CARS), and a 12 in Bio/Biochem so their old scale composite score would be a 32. This is, objectively speaking, probably the fairest way to compare old and new test scores (because it eliminates the biggest difference between the two exams), but it has the caveat of needing to be done manually. Its doubtful that admissions at every school, let alone at most schools has the time to do something like this. Nonetheless, it is the fairest because it does the best job at controlling for the differences between the exams. There are other factors that are completely neglected by directly comparing composite percentiles like the fact that the testing time/length is nearly doubled for the new exam (3 hours and 20 minutes vs 6 hours and 15 minutes). Note this only has relevance if med schools actually do compare applicant test scores side-by-side when making decisions.

2. Designate a number of seats for takers of the old exam and a number of seats for takers of the new exam. Only compare applicants who took the same exam(s). This is equally as fair as the first option, but has the issue of limiting the number of seats available to applicants based on which exam they took which in most cases is entirely out of their control.

And of course I'm probably stating the obvious here, but all of this will become irrelevant (if it isn't already) once schools stop accepting old scores.
 
When applying, my suggestion would be to go with the safe bet and assume they are comparing the old vs new with percentiles.

However, if you're wondering what the reasonable and fairest alternatives would be I'll share my thoughts (mind you, this is speculation, I'm just an applicant that was recently accepted and still have no clue how things actually go).

1. Take the first 3 sections of the new exam and omit the last one (Psych/Sociology) and use percentiles to individually convert them to an old scale score. For example a 128 in any of these sections would be about an 11 and a 125 would be an 8. Then combine the scores to get an old scale composite score (and its corresponding percentile). For instance someone who get a 128/126/129/127 on the new exam would have an 11 in Phys/Chem, a 9 in verbal (or CARS), and a 12 in Bio/Biochem so their old scale composite score would be a 32. This is, objectively speaking, probably the fairest way to compare old and new test scores (because it eliminates the biggest difference between the two exams), but it has the caveat of needing to be done manually. Its doubtful that admissions at every school, let alone at most schools has the time to do something like this. Nonetheless, it is the fairest because it does the best job at controlling for the differences between the exams. There are other factors that are completely neglected by directly comparing composite percentiles like the fact that the testing time/length is nearly doubled for the new exam (3 hours and 20 minutes vs 6 hours and 15 minutes). Note this only has relevance if med schools actually do compare applicant test scores side-by-side when making decisions.

2. Designate a number of seats for takers of the old exam and a number of seats for takers of the new exam. Only compare applicants who took the same exam(s). This is equally as fair as the first option, but has the issue of limiting the number of seats available to applicants based on which exam they took which in most cases is entirely out of their control.

And of course I'm probably stating the obvious here, but all of this will become irrelevant (if it isn't already) once schools stop accepting old scores.
I wouldn't find the first option fair at all. If an applicant who took the new mcat focused more on Psych/soc and happens to do better on that than he/she did on verbal then it'd be unfair to completely omit that section even though it's a section ON the exam.

The fairest way to judge two applicants with different versions of the MCAT is to look at their percentile imo.
 
I wouldn't find the first option fair at all. If an applicant who took the new mcat focused more on Psych/soc and happens to do better on that than he/she did on verbal then it'd be unfair to completely omit that section even though it's a section ON the exam.

The fairest way to judge two applicants with different versions of the MCAT is to look at their percentile imo.

The first option wouldn't be fair if you were comparing new test takers to new test takers, but this cycle and the few after it are a mixed bag. Using your logic its unfair that someone who did bad in verbal on the old exam didn't have a psych section which they could make up for it on. So on the contrary it would seem the fairest thing to do would be to control for the variability of the two exams. There's no use in elaborating more on this because I'm not on an ADCOM nor do I know how ADCOMS go about comparing applicants, I'm simply stating the fairest way to handle a situation like the one OP describes.

This is a going to sound a bit hyperbolic, but by saying the composite score percentiles of old exams should be equated with those of the new exams you might as well be saying that med schools should let someone sub in a DAT or GRE score for an MCAT score. The point being just because they're both called the MCAT doesn't mean they're the same exam. I'm certain that the reason ADCOMS resort to comparing the total score percentiles generally is for the sake of efficiency considering we're talking about a couple thousand applicants. When its necessary to get down to the nitty-gritty though it only makes sense that they will start to compare subsection percentiles.
 
Top