MCAT scores are sort of weird

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

Leer

Full Member
10+ Year Member
Joined
Nov 16, 2010
Messages
48
Reaction score
0
Members do not see this ad.
I haven't done enough research into the MCAT to figure out how the scoring works, so I'm wondering just how far apart (in terms of percentile) a score of say . . . 28 and 33 are.

Based on what I read on these boards, for US MD, a score (ROUGHLY) of 0 to say ... 27 or 28 is not favorable, a score of 29 to 35 is optimal, and a score of 36 to 45 is of course the best, although most people won't score into this range.

How much worse does the person who gets a 28 on the MCAT perform than the person who gets a 33?
 
Leer said:
I haven't done enough research into the MCAT to figure out how the scoring works, so I'm wondering just how far apart (in terms of percentile) a score of say . . . 28 and 33 are.

Based on what I read on these boards, for US MD, a score (ROUGHLY) of 0 to say ... 27 or 28 is not favorable, a score of 29 to 35 is optimal, and a score of 36 to 45 is of course the best, although most people won't score into this range.

How much worse does the person who gets a 28 on the MCAT perform than the person who gets a 33?
Click to expand...

While I don't have the percentile breakdown at my disposal (and it changes for every test day). Its a very tight curve with an asymtote at 45, so it becomes increasinly difficult to reach that top score. A 26 is ~50th percentile. I got a 31which was an 80-83.7th percentile on my test day. I believe a 35 is >90th and so on. Just to give you an idea.



Typed on my phone so sorry for grammar.
 
Leer said:
I haven't done enough research into the MCAT to figure out how the scoring works, so I'm wondering just how far apart (in terms of percentile) a score of say . . . 28 and 33 are.

Based on what I read on these boards, for US MD, a score (ROUGHLY) of 0 to say ... 27 or 28 is not favorable, a score of 29 to 35 is optimal, and a score of 36 to 45 is of course the best, although most people won't score into this range.

How much worse does the person who gets a 28 on the MCAT perform than the person who gets a 33?
Click to expand...

28 and 33 are VERY far apart. That's a 5-point spread. Of course, I have known people to make jumps like that. A friend got a 26 her first time around, followed by a 34 her second. I suspect her second time around involved some devoted studying, however. The first time, she ignored my suggestions and just went for it without time to study. The second time she knew she had to make time, so she did.
 
askamsky51 said:
While I don't have the percentile breakdown at my disposal (and it changes for every test day). Its a very tight curve with an asymtote at 45, so it becomes increasinly difficult to reach that top score. A 26 is ~50th percentile. I got a 31which was an 80-83.7th percentile on my test day. I believe a 35 is >90th and so on. Just to give you an idea.



Typed on my phone so sorry for grammar.
Click to expand...


Standard I've seen used a lot is:

50th -- 25
75th -- 30
90th -- 33
95th -- 35
99th -- 38

For MD applicants, median is ~3.5/28; MD matriculants is ~3.7/31
 
The difference between a 28 and 33 is pretty big for US MD. Since a 32 is average for accepted students, that's a jump from 4 points below average to a point above.

This thread I made may help you with part of your question http://forums.studentdoctor.net/showthread.php?t=888650

In particular, this chart where I added rough MCAT score percentiles to the bottom http://i44.tinypic.com/2r40pr6.jpg

You can see that if you have a 3.7 GPA, the historical acceptance percentage jumps from ~50% to ~78% when you move from a 28 to 33. The graph is only for US MD schools BTW. More information regarding the sources is contained in the thread

This page has detailed MCAT percentiles for a ton of test years https://www.aamc.org/students/applying/mcat/admissionsadvisors/mcat_stats/
 
Last edited:
28 ... 33 is really yuge. Divide by three, and you're essentially asking what's the difference between 9.3 and 11. It's a lot....

From eight to ten
To get from 8 to 10 is a big deal if you're starting at 8. To scale those two points require moving from 50% to the 80% percentile on a curved test. I imagine this is where the bulk of the B+/A- pre-med folks at a mid-range undergraduate school fight it out.

From ten to twelve
I crushed my pre-reqs (as in setting the curve in most classes) and the ACS organic exam, and I found myself struggling all summer with TBR/EK/TPRH/AAMC practice tests to see where I would land in the 10 - 12 range. The very best students at a mid-range school will probably land here, as will the average HYPS student. Personally, I feel that I could have cut 75% of my MCAT preparation time and still land a 10. I put in 3/4's of the effort just to go from 10 to 11. Some top students at my mid-range school studied for a whole year to get a 10; others got a 9.x. I knew a guy with a 3.7+ cGPA/sGPA who did not study, and he got a 7.3 (22 composite).

13+
The bulk will come from the 10 - 12 crew that prepared exceptionally well, got lucky, or have a deep science/research background (helps a lot for BS). I have to imagine a very small proportion (< 1%) of test-takers can consistently bank on 37+ that doesn't factor in luck or background - like the top students at the elite schools.

I blew out my entire summer just to go from a 10 to 11 average sub-section score. That's how hard it was for me.

Source
Actual: 33S (12 PS, 10 BS, 11 VR)
AAMC Practice # 3: 34 (13 PS, 12 BS, 9 VR)
AAMC Practice # 10: 33 (11 PS, 11 BS, 11 VR)
AAMC Practice # 8: 36 (12 PS, 12 BS, 12 VR)
AAAMC Practice # 11: 34 (12 PS, 11 BS, 11 VR)
ACS Organic: 96% percentile
 
Last edited:
solitarius said:
28 ... 33 is really yuge. Divide by three, and you're essentially asking what's the difference between 9.3 and 11. It's a lot....

From eight to ten
To get from 8 to 10 is a big deal if you're starting at 8. To scale those two points require moving from 50% to the 80% percentile on a curved test. I imagine this is where the bulk of the B+/A- pre-med folks at a mid-range undergraduate school fight it out.

From ten to twelve
I crushed my pre-reqs (as in setting the curve in most classes) and the ACS organic exam, and I found myself struggling all summer with TBR/EK/TPRH/AAMC practice tests to see where I would land in the 10 - 12 range. The very best students at a mid-range school will probably land here, as will the average HYPS student. Personally, I feel that if I could have cut 75% of my MCAT preparation time and land a 10. I put in 3/4's of the effort just to go from 10 to 11.

13+
The bulk will come from the 10 - 12 crew that prepared exceptionally well, got lucky, or have a deep science/research background (helps a lot for BS). I have to imagine a very small proportion (< 1%) of test-takers can consistently bank on 37+ that doesn't factor in luck or background - like the top students at the elite schools.

I blew out my entire summer just to go from a 10 to 11 average sub-section score. That's how hard it was for me.
Click to expand...

The thing about very high scores it that they can be thrown off quite easily and there is a lot of luck to it. I personally was scoring in the 13-14 range on both science sections during all practice tests. I got an 11 and a 10 on the real deal. A lot is up to chance.
 
music2doc said:
The thing about very high scores it that they can be thrown off quite easily and there is a lot of luck to it. I personally was scoring in the 13-14 range on both science sections during all practice tests. I got an 11 and a 10 on the real deal. A lot is up to chance.
Click to expand...

This can be true, but not always. If you've prepped accordingly and don't screw up on the real test, you should be pretty set to score within your range. This includes those scoring 35-45 (the 'high' range pertaining to the 95th percentile).

I was scoring 14-15s on Bio and Phy Sci, and was getting 10-12s on verbal, and I ended up scoring nearly exactly what I expected, high ranges on them all. I attribute this entirely to be so comfortable with the environment by the time I took it. The only thing I regret is not going in the center beforehand and getting frustrated by the typing of other students (which is no longer an issue without the writing section...). Even though I was in the extremely high percentiles, I did not vary more than 1/2 points.

However, it can be pretty unpredictable. Far back in the post history of this account you'll find another person who created this account, a friend of mine. He was expecting 80th percentile range and ended up with a 95th percentile score. He knew he'd done well, as he cruised through the sections and felt like he was making good time and good choices.

Meanwhile, I was expecting another friend of mine to score higher than me on the MCAT, he ended up scoring around 95th percentile even though he'd been in the 99th percentile range all along ('bombed' a section, went from scoring 15s to scoring an 11). He was well aware he did poorly on it, so it wasn't like he'd just unluckily gotten those questions wrong. He performed poorly and knew it. Unfortunately for him, he knows that retaking it is foolish since all scores must be reported.

All you can do is perform your best. It's a very well-designed standardized test and you can prepare for it very well for the time being. The changes will make it more interesting, though. In the end, the medical school you attend (as long as you get in) won't matter. So if you're shooting for the 95th+ percentiles to get into a top tier, it really isn't that important. If you crush the step 1 and perform well in any medical school, you can get into a residency of your choice.
 
PreMedOrDead said:
This can be true, but not always. If you've prepped accordingly and don't screw up on the real test, you should be pretty set to score within your range. This includes those scoring 35-45 (the 'high' range pertaining to the 95th percentile).

I was scoring 14-15s on Bio and Phy Sci, and was getting 10-12s on verbal, and I ended up scoring nearly exactly what I expected, high ranges on them all. I attribute this entirely to be so comfortable with the environment by the time I took it. The only thing I regret is not going in the center beforehand and getting frustrated by the typing of other students (which is no longer an issue without the writing section...). Even though I was in the extremely high percentiles, I did not vary more than 1/2 points.
Click to expand...

I'm not sure I follow your logic here. I don't think he said it was not possible to score the same as your practice score; rather, he said there was some luck to it for very high scores. How do you know you didn't get lucky with the content covered by your test date? Sample size n = 1.

The topics for the BS portion is so vast; do you feel you knew all of them cold for the stand-alone questions or even the standalone-like questions grouped in the passages? On the PS section of my MCAT, I got thrown off by a simple atomic number calculation that ate up too many minutes that I could have used for an extremely difficult oxygen passage. Those minutes easily cost me 3 - 4 questions in that passage and a 13+ score. Not complaining, but just making a point.

As you know, MCAT is not just a knowledge test. It is that + mental/emotional endurance test + time management test + analysis/interpretation test + reconfiguring awkwardly worded question test + reading through poorly-written passage test. I feel very fortunate to have had the writing sample as a breather, and now that's gone.

My point was that a very few number of people can consistently score 37+ on the actual on repeated attempts. AMCAS data on repeated test takers provides overwhelming evidence that I am right:

https://www.aamc.org/students/download/271680/data/retestertotalscorechange.pdf

If you believe AAMC's data above and if you treat the average of all your AAMC practice tests as one virtual MCAT sitting, then his original argument about luck playing a role in practice vs. actual results is probably valid as well.

PreMedOrDead said:
However, it can be pretty unpredictable. Far back in the post history of this account you'll find another person who created this account, a friend of mine. He was expecting 80th percentile range and ended up with a 95th percentile score. He knew he'd done well, as he cruised through the sections and felt like he was making good time and good choices.

Meanwhile, I was expecting another friend of mine to score higher than me on the MCAT, he ended up scoring around 95th percentile even though he'd been in the 99th percentile range all along ('bombed' a section, went from scoring 15s to scoring an 11). He was well aware he did poorly on it, so it wasn't like he'd just unluckily gotten those questions wrong. He performed poorly and knew it. Unfortunately for him, he knows that retaking it is foolish since all scores must be reported.
Click to expand...

Sample size increased from n=1 to n = 3. In doing so, it provides further evidence that luck is a major factor for many of the 36+ scorers. Subject #1 met his practice score; subject #2 exceeded it; subject #3 underperformed his practice score.
 
Last edited:
Might have misunderstood my intention. I was simply implying that comfortability and consistency in practices is likely to lead to a low standard of deviation on your real test. Of course, there is variation, and it is increased in higher scores. There is also going to be more variation around the ends of the spectrum (very low, very high) because that is how bell curves work.
 
Venom5 said:
34 = 50th percentile on SDN
Click to expand...

Typical SDN poster: Struggling with a B- average in Bio 1, but is sure she's gonna get a 39. Hence, 32 is "way too low." Oh wait, 32 is the average MCAT score for Harvard and Princeton undergrads and higher than the average allopathic matriculant.

Or they just need to troll/lie better... like this guy.

askamsky51 said:
While I don't have the percentile breakdown at my disposal (and it changes for every test day). Its a very tight curve with an asymtote at 45, so it becomes increasinly difficult to reach that top score. A 26 is ~50th percentile. I got a 31which was an 80-83.7th percentile on my test day. I believe a 35 is >90th and so on. Just to give you an idea.
Click to expand...
 
solitarius said:
Typical SDN poster: Struggling with a B- average in Bio 1, but is sure she's gonna get a 39. Hence, 32 is "way too low." Oh wait, 32 is the average MCAT score for Harvard and Princeton undergrads and higher than the average allopathic matriculant.
Click to expand...

Where'd you get the Harvard/Princeton stats?

Still, considering a 32 is in the mid 80th percentile, that's a pretty damn good average.

solitarius said:
Or they just need to troll/lie better... like this guy.
Click to expand...

I think you read that wrong. He said 80-84th percentile, which would be right. Not 7th percentile that you bolded.
 
solitarius said:
I googled it. It's online. You just have to look.
Click to expand...

Ah, I see. Considering a 32 is ~85th percentile, that's a pretty amazing average. I don't even want to know their 75/90th percentile scores compared to their 50th, I would assume they're around 36/40, respectively.

solitarius said:
I seem to be doing that a lot tonight. Does AMCAS publish percentile scores for an exact test date?
Click to expand...

You're totally fine, it happens to the best of us.

I believe they only do yearly publications of that data, but you can use the MCAT THx to look at your personal percentiles that get sent to the medical schools. For example, on my test day a 12 on VR was worth 95.5-98.2 percentile. Normally, test-to-test these don't vary by more than a percentile in either direction for a score.
 
PreMedOrDead said:
Ah, I see. Considering a 32 is ~85th percentile, that's a pretty amazing average. I don't even want to know their 75/90th percentile scores compared to their 50th, I would assume they're around 36/40, respectively.


You're totally fine, it happens to the best of us.

I believe they only do yearly publications of that data, but you can use the MCAT THx to look at your personal percentiles that get sent to the medical schools. For example, on my test day a 12 on VR was worth 95.5-98.2 percentile. Normally, test-to-test these don't vary by more than a percentile in either direction for a score.
Click to expand...

I know an ORM who got into an MD school with a 23 mcat and standard life experiences and ECs. Granted that was in Texas, but numbers don't mean everything (or in his case, anything).
 
There's definitely some luck involved in the MCAT. I prepped in just over 2 wks and scored a 33 (9 in verbal and 12's in the sciences). I thought I was good for 13+ in the sciences and 8+ on verbal. I was stunned to only get a 12 in BS. I thought I scored at least a 14, so scoring lower means I made some mistakes because I didn't find that section difficult at all. I think scoring 13+ requires some luck and not making careless mistakes, and this is where practice plays a role. If I prepped HARD and took all the AAMC tests, it's possible I could've scored a 36+, but there's a slight possibility I could've only made a 31 after all that effort. Pray for luck on test day.
 
Once you're 12+ the scoring is ridiculously close, almost arbitrary.
 
And I think the MCAT should be phased out. We need a more rote memory test that mirrors medical school exams because I feel all I do everyday is memorize minutiae and think little 👎
 
You must log in or register to reply here.

Similar threads

O
  • Question Question
How do top Medical Schools view top MCAT scores
Replies
5
Views
2K
EEqualsRx2
EEqualsRx2
chthonicYODEL
Expired MCAT scores
Replies
4
Views
1K
chthonicYODEL
chthonicYODEL
J
Is a low MCAT score enough to “auto-waitlist” me? Or should I assume my stats are fine if I got IIs?
Replies
16
Views
913
Jjdoctor18
J
FutureDOCTORR123456789
Is it weird to call a medschool in my area to ask what they look for?
Replies
4
Views
1K
FutureDOCTORR123456789
FutureDOCTORR123456789
T
Secondary Essays Weird EC Question
Replies
2
Views
692
LizzyM
LizzyM
Top