The 40+ Myth

[owen_osh]

Relatively Prime has sort of addressed this on the Hopkins rejection thread, but I want to throw in my two cents:

How many times on this board have we read some variant of the following: "Yeah, you may have above average MCAT scores, but it's nothing to get excited about because top schools could easily fill their entire classes with 4.0 and 40+ MCATs."

or this: "Don't worry to much about your below average MCAT scores. Top schools could easily fill their entire classes with 4.0 and 40+ MCATs, but we know the average stats for accepted students are lower than that so it must mean that top schools care more about the 'total package.'"

Judging from the 2003-2004 MSAR, there are not that many 40+ MCAT scorers. I used a ruler and the graph on page 27 titled MCAT Total Numeric Score Distribution, Year 2001 Applicants to estimate the number of people with each score. I think there are only about 184 people with scores of 40 or greater, a number which would barely fill the class at Harvard.

Here's a list with the score, my estimate of the number of applicants with that score, and my estimate of the number of applicants with that score who are accepted.

Score/ # Applicants/ # Accepted
42 / 26 / 26
41 / 53 / 53
40 / 105 / 105
39 / 211 / 211
38 / 395 / 316
37 / 579 / 474
36 / 711 / 579
35 / 974 / 789
34 / 1368 / 1105
33 / 1684 / 1237
32 / 2026 / 1421
31 / 2395 / 1658
30 / 2474 / 1632

For comparison, the top 5 schools based on US News enroll 648 first year students. If they filled their classes based solely on MCAT scores, the lowest MCAT accepted would be 38. The top 10 schools take 1388 students, and the lowest MCAT they would have to take to fill their classes would be 36. For the top 20 schools, which enroll 2697 first years, they would have to take people with at least a 35 and some with a 34. Finally, for the top 30 schools, which enroll a total of 3984 students, they would have to take some students with 33's.

The point is that the 40+ myth grossly exagerates the number of people with very high scores.

Obviously, schools don't completely base admissions on MCAT and there are a lot of factors that complicate any meaningful theory of MCAT cutoffs for top schools. Some things that might be important...1) The average MCAT score for accepted students at every top school is below what would be predicted by my analysis above. 2) The average MCAT for accepted student includes applicants with multiple acceptances. I think this scews the averages up. For example, a person with a 40+ might be accepted at 10 out of the top 10 schools, thus bumping up the averages at each of those schools. On the other hand, someone with a lower score that could balance out the 40+ guy/gal would probably be accepted at fewer of the top 10 schools. 3) I imagine that many people with high MCAT scores would choose to go to their state school which might no be highly ranked. These applicants probably have a good chance at getting large scholarships at their state school. 4) From what I've read, the scores for accepted URM's, at least at the top schools, are very close to those for non-URM's.

 

[jot]

you're cut out for a research career

 

[none]

Yeah...it is an exaggeration. The message is that schools look for things other than high MCAT scores.

 

[CaNEM]

Nice...

Got a link to the MSAR?

 

[Jumpu]

Wow! That really puts things into perspective! It also seems that SDN attracts the best and brightest of all the applicants out there too.

 

[limit]

Here's what I want to know:
What's up with those that scored 30-38 and don't get in?

 

[DW]

lack of ECs.
bad interviewing.
shoddy personal statement.
bad LORs.
just bad luck
among other potential factors...........

 

[UCLAMAN]

People with 40+ and 3.8+ shouldn't fret. Those people have plenty of opportunities. I've always thought of medical school classes as being analogous to football teams. On a football team there are many different positions to be played. Lets say for the sake of this example that people with 40+ and 3.8+ are quarterbacks. But on a football team there is only room for 3 maybe 4 quarterbacks on the roster. So lets say a football team picks a team composed solely of quarterbacks(40+s and 3.8+s). How do you think they would fare? I would rather have a football team with players filling the need for every position. Imagine a football team composed Joe Montana, Brett Favre, Troy Aikman, Steve Young, Warren Moon, etc. etc.(a roster solely filled with quarterbacks). Now if I were a quarterback(ie a 40+ 3.8+) I dunno about you guys but I would want to make sure that I had a bunch 6'5" 350lb linemen blocking in front of me rather than a bunch of Joe Montanas in the trenches. Now Joe Montana was great no doubt. But a team solely of Joe Montanas would suck. Medical schools don't want a team full of quarterbacks. They want linemen, runningbacks, wide receivers, free safeties, etc. etc.

Sorry if I was long winded.

Just my $0.02

 

[San_Juan_Sun]

I like the football analogy, although it may be good to know that offensive linemen always have the highest Wonderlic scores at the combines.

But, how many great O linemen can we name? I can't think of too many! Good post, and good info. Thanks.

 

[dr kevin40]

great post. had no idea i was in suchselect compnay

 

[CaNEM]

Originally posted by dr kevin40
great post. had no idea i was in suchselect compnay

 

[mws99]

Originally posted by dr kevin40
great post. had no idea i was in suchselect compnay

Well, maybe if those other 700 people with a 36 retook the test, there would be more 40's.

 

[mws99]

Score/ # Applicants/ # Accepted/ Acceptance Percentage
42 / 26 / 26 / 100
41 / 53 / 53 / 100
40 / 105 / 105 / 100
39 / 211 / 211 / 100
38 / 395 / 316 / 80
37 / 579 / 474 / 82
36 / 711 / 579 / 81
35 / 974 / 789 / 81
34 / 1368 / 1105 / 81
33 / 1684 / 1237 / 73
32 / 2026 / 1421 / 70
31 / 2395 / 1658 / 69
30 / 2474 / 1632 / 66


I thought this additional info would also be helpful. It seems acceptance rate is directly related to MCAT score, but you could make the argument that a higher MCAT probably means a higher gpa.

 

[cabruen]

Notice the distinct breaks in the data.

From 42-39, there is essentially a 100% acceptance rate.

From 34-38, the acceptance rate is a constant 80%.

From 30-33, the acceptance rate is essentially 70%.

Interesting.

 

[mongoose]

I bet there are less than you all have estimated at 40 or above....try a statisitical breakdown using standard deviations to estimate percentage wise the number a certain level of standard deviations over the average.....should be a better way to look at it. I always thought there were quite a few who exaggerated based on the standard deviations listed on my MCAT score report....(SD was 2.5 for my year with average a little over 24....do the math.)

 

[Random Access]

Originally posted by mws99
Well, maybe if those other 700 people with a 36 retook the test, there would be more 40's.

While I'll give Kevin some credit for improving that much, yeah...


-RA

 

[kito]

football teams are nothing without the men up front

 

[Random Access]

Originally posted by mongoose
I bet there are less than you all have estimated at 40 or above....try a statisitical breakdown using standard deviations to estimate percentage wise the number a certain level of standard deviations over the average.....should be a better way to look at it. I always thought there were quite a few who exaggerated based on the standard deviations listed on my MCAT score report....(SD was 2.5 for my year with average a little over 24....do the math.)

Hmm...let's see...68/95/99.7 rule... 2 SDs, or 29 and above should be only 2.5% of people. 3 SDs or 31.5 and above should only be 0.15% of people. 40 is 6.4 SDs...don't have a table handy, but very very very few folks...

-RA

 

[mws99]

Originally posted by Random Access
Hmm...let's see...68/95/99.7 rule... 2 SDs, or 29 and above should be only 2.5% of people. 3 SDs or 31.5 and above should only be 0.15% of people. 40 is 6.4 SDs...don't have a table handy, but very very very few folks...

-RA

I checked the AAMC website and they have nothing about the standard deviations of the composite score but they do have SD's for each section, which have been about 2.1-2.3 consistently. By your calculations, that would greatly increase the number of people with high scores.

http://www.aamc.org/data/facts/famg62001a.htm

 

[Random Access]

Originally posted by mws99
I checked the AAMC website and they have nothing about the standard deviations of the composite score but they do have SD's for each section, which have been about 2.1-2.3 consistently. By your calculations, that would greatly increase the number of people with high scores.

http://www.aamc.org/data/facts/famg62001a.htm

Well in that case, it's totally different. I was pretty sure those numbers I wrote above made little sense, since it looked like few folks would have scored higher than a 30...

Let's say 2.2 as the SD for each section. So a 10.2 on a section or higher is 1SD, or 16%. 12.4 or higher is 2SDs, or 2.5%. 14.6 or higher is 3SDs or 0.15%.

It's hard to figure out a composite score percentile, since some people do better in sciences than verbal, and to a lesser extent vice versa.

-RA

 

[CoffeeCat]

Originally posted by dr kevin40
great post. had no idea i was in suchselect compnay

Oh no no, drkevin, believe me you are part of a group in the premed world that is not small, unfortunately

 

[limit]

Originally posted by cabruen
Notice the distinct breaks in the data.

From 42-39, there is essentially a 100% acceptance rate.

From 34-38, the acceptance rate is a constant 80%.

From 30-33, the acceptance rate is essentially 70%.

Interesting.

Indeed. How, now, do we explain the overall 45-50% acceptance rate of everyone who applies?

<30 MCAT scorers then must have an acceptance rate of 30% or lower. Now, factor in all the people with <30 who are coming from IVY's and with high GPA's and sons of physicians who donate regularly and heavily... this lowers the percentage for the average joe down to 20-25% or so.

Next, factor in those that get accepted with <25 MCAT. This puts the 25-29 range into a very unfavorable situation -- odds are very much against them. Am I making any sense at all?

 

[mws99]

Originally posted by limit
Indeed. How, now, do we explain the overall 45-50% acceptance rate of everyone who applies?

<30 MCAT scorers then must have an acceptance rate of 30% or lower. Now, factor in all the people with <30 who are coming from IVY's and with high GPA's and sons of physicians who donate regularly and heavily... this lowers the percentage for the average joe down to 20-25% or so.

Next, factor in those that get accepted with <25 MCAT. This puts the 25-29 range into a very unfavorable situation -- odds are very much against them. Am I making any sense at all?

I was thinking along the same lines as you. I think this is a bigger numbers game than people always say.

 

[Jonkst]

Someone has too much time on their hands...

Probably time to start drinking.

 

[OneStrongBro]

Jonkst,

YOU ARE 100% RIGHT.

 

[Spiderman [RNA Ladder 2003]]

Originally posted by limit
Indeed. How, now, do we explain the overall 45-50% acceptance rate of everyone who applies?

<30 MCAT scorers then must have an acceptance rate of 30% or lower. Now, factor in all the people with <30 who are coming from IVY's and with high GPA's and sons of physicians who donate regularly and heavily... this lowers the percentage for the average joe down to 20-25% or so.

Next, factor in those that get accepted with <25 MCAT. This puts the 25-29 range into a very unfavorable situation -- odds are very much against them. Am I making any sense at all?

It means that there a lot of <30 applicants/acceptees

 

[limit]

Originally posted by Spiderman [RNA Ladder 2003]
It means that there a lot of <30 applicants/acceptees

Quite the contrary, many more <30 applicants than acceptees. Just look at the numbers.

To me, what these numbers are saying is that if you're a <30 overrepresented male with nothing else extraordinary -- don't apply, statistically you're not favored.

 

[Blitzkrieg]

Originally posted by limit
Quite the contrary, many more <30 applicants than acceptees. Just look at the numbers.

To me, what these numbers are saying is that if you're a <30 overrepresented male with nothing else extraordinary -- don't apply, statistically you're not favored.

well, I wouldn't say "dont apply"....You never know the current demographic of applicants...the pool is dynamic.

Also, never live by statistics =D

However, yes, I would say that one has a low chance of gaining admission if he is a <4.0 <30 applicant with hardly any EC's.

Reminds me of a study buddy from undergrad..he had a 4.0, and was gunner in every sense....took the MCAT..29..took again, 28.

No EC's, cos he was always studying...brilliant guy..but no acceptances.

 

[Jalby]

Lies, damn lies, and statistics.

 

[Random Access]

Originally posted by Jalby
Lies, damn lies, and statistics.

Well yeah, as you know, 79.5 percent of all statistics are completely made up.

-RA

 

[DALABROKA]

To me, what these numbers are saying is that if you're a <30 overrepresented male with nothing else extraordinary -- don't apply, statistically you're not favored.

I find this interesting considering that the average medical student has a ~29.3 MCAT score (see AAMC matriculant data for 2001). With this in mind, how can anyone say that people with <30 should not apply? It seems that many sub 30 applicants get accepted.

 

[limit]

Originally posted by DALABROKA
I find this interesting considering that the average medical student has a ~29.3 MCAT score (see AAMC matriculant data for 2001). With this in mind, how can anyone say that people with <30 should not apply? It seems that many sub 30 applicants get accepted.

That's the ugly beauty of statistics. The mid-upper 30s group is a minority by comparison, so there is enough <30 applicants to even out the average. So, for example, if 1 million out of 4 million get accepted, you can argue that "yes, a whole million got accepted," but you must also consider the 3 million that didn't. Plus you must consider those that get in with extremely low number for whatever reasons.

In the end, the numbers make perfect sense, and your "numerical" qualities are ostensibly very important. Anyone who says this is a "crapshoot" either hasn't done their homework (looked at the numbers), or somehow fell through the crack despite being statistically favored.

I'm not saying not to apply... I'm just pointing out, according the numbers posted earlier in this thread, that you are statistically at a disadvantage with an MCAT <30.

 

[Random Access]

Originally posted by DALABROKA
I find this interesting considering that the average medical student has a ~29.3 MCAT score (see AAMC matriculant data for 2001). With this in mind, how can anyone say that people with <30 should not apply? It seems that many sub 30 applicants get accepted.

Yeah, I thought the <30 claim was ridiculous. I mean, of course people with scores <30 should apply. The odds are against them for the top schools, but there are 125 or so med schools, so not all can require a 35 MCAT!

In any case, it seems some schools with lower average stats than the top schools seem to give interviews more to candidates with closer to their average stats. I mean, it makes sense. Every school wants a high yield, and they don't want to waste resources on people who will probably go to a higher ranked school...

-RA

 

[mws99]

To estimate your chances with a sub 30 score:

17000 spots - #of people above 30 accepted=7394 spots left

35000 applicants - # of applicants with scores above 30 = 21999

7394/21999 = 33% chance of acceptance.

Of course this is an average because a 29 has a better shot than a 27, etc.

 

[NYGirl12]

just bumping this up..hopefully, it will help some applicants

 

[Rendar5]

WHOO!!!!!!!!!!!!!!

That means I'm getting in. But only because of a number on a test and not because of who I am
I'd rather get in for the right reasons, but getting in for the wrong reasons is better than not getting in at all.

 

[robotdancing]

aamc website does list std devs for comp scores. for the april 2003 test, the mean was 24.8, std dev was 6.3

 

[LaurieB]

Originally posted by limit
That's the ugly beauty of statistics. The mid-upper 30s group is a minority by comparison, so there is enough <30 applicants to even out the average. So, for example, if 1 million out of 4 million get accepted, you can argue that "yes, a whole million got accepted," but you must also consider the 3 million that didn't. Plus you must consider those that get in with extremely low number for whatever reasons.

In the end, the numbers make perfect sense, and your "numerical" qualities are ostensibly very important. Anyone who says this is a "crapshoot" either hasn't done their homework (looked at the numbers), or somehow fell through the crack despite being statistically favored.

I'm not saying not to apply... I'm just pointing out, according the numbers posted earlier in this thread, that you are statistically at a disadvantage with an MCAT <30.

I think it is a crapshoot in the sense that if you have middle of the road MCAT scores (50-70th percentile), you never know where you'll get accepted. It begins to be based more on the outside factors. My 2 cents.

 

[ehd327]

I don't know if this affects the stats, but we are all considering the overall score and not the breakdown of each section. I say this because I was interviewed at a particular school mainly because of a discrepency between my VS score and my non-science LOR stating my exceptional writing and comprehension skills (this is not speculation, he actually talked about the topic.) i.e, A 35 on the MCAT can be broken down in several ways, i'll take two approaches: 1.) VS-7, PS-14, BS-14 ; 2.) VS-11, PS-12, BS-12. Obviously, considering only the numbers, the 1st applicant seems skewed whereas the 2nd applicant seems very balanced across many fields. Both, however, have managed to get a 35 which is a great score. I would think that top 5 (maybe top 10) schools would definately see this as a factor. So, responding to a previous post, there needs to be an extra dimension for the MCAT score for the breakdown across all sections.

 

[Optimist]

Originally posted by mws99

39 / 211 / 211 / 100
38 / 395 / 316 / 80

It's strange how it drops like a rock at 38. What does a 39 have that a 38 doesn't?

I'd love to meet one of the 20% of the people with a 38 who didn't get in so that I can smack them upside the head.

 

[sinnah83]

has anyone checked the detailed score report (one that they send out by mail) I could have sworn it had some of that info, SD and number of ppl of who the test this april adn then they give you the percent of applicants who have a certain score.

 

[Rendar5]

I just know karma's going to bite me in the ass for that remark and I won't get in anyplace.

 

[Optimist]

Originally posted by Rendar5
I just know karma's going to bite me in the ass for that remark and I won't get in anyplace.

Hmm...

Predictions for Class of 2008:

39 / 211 / 210 / 99.5 %

 

[Gleevec]

Also the MCAT is the first thing to work against you for a school in which you are way above the MCAT range. I imagine that one of the first things that makes an applicant over-qualifed for a school is his or her MCAT.

 

[ehd327]

Originally posted by Optimist
Hmm...

Predictions for Class of 2008:

39 / 211 / 210 / 99.5 %

Tsk. Tsk. Optimist has become quite the pessimist!

 

[R_C_Hutchinson]

Originally posted by Gleevec
Also the MCAT is the first thing to work against you for a school in which you are way above the MCAT range. I imagine that one of the first things that makes an applicant over-qualifed for a school is his or her MCAT.

interesting point, kinda worries me though...
what happens if your MCAT and GPA dont "match"? do you get rejected from schools with lower mcats on the basis of being "over-qualified" and from higher-stat schools because your GPA isnt up to par?

if so i may be "proper-fu*ked"
(before "zee germans" get there)

 

[Gleevec]

Originally posted by R_C_Hutchinson
interesting point, kinda worries me though...
what happens if your MCAT and GPA dont "match"? do you get rejected from schools with lower mcats on the basis of being "over-qualified" and from higher-stat schools because your GPA isnt up to par?

if so i may be "proper-fu*ked"
(before "zee germans" get there)

I dont think its possible for me to speak for adcoms on this with even relative certainty, since most schools differ greatly.

From what I understand, and take this with a huge grain of salt, is that aside from the ideal combination of high GPA and high MCAT, its better to have a high MCAT and low GPA than vice versa, since the MCAT is a standardized measure.

But its not the end all be all of the process, just one of the 5 or so major aspects of it.

 

[jlee9531]

Originally posted by Gleevec
I dont think its possible for me to speak for adcoms on this with even relative certainty, since most schools differ greatly.

From what I understand, and take this with a huge grain of salt, is that aside from the ideal combination of high GPA and high MCAT, its better to have a high MCAT and low GPA than vice versa, since the MCAT is a standardized measure.

But its not the end all be all of the process, just one of the 5 or so major aspects of it.

i agree with what gleevec is saying here.

i think it is better to have the lower gpa higher mcat rather than higher gpa lower mcat because the lower gpa might just be a reflection of a more challenging undergraduate program. the mcat is like an equalizer... to be able to compare students that come from different schools.

 

[Optimist]

Originally posted by ehd327
Tsk. Tsk. Optimist has become quite the pessimist!

It's a joke. Besides, I'm not a big fan of arrogance.

 

[Rendar5]

me neither, not sure what got into me last night.