MSAR 2019 is out

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
Members don't see this ad.
LadyGlen said:
MCAT
25%: 514 (accepted), 513 (matriculated)
75%: 520, 519
Median: 517, 516

GPA (total)
25%: 3.68 (accepted), 3.66 (matriculated)
75%: 3.92, 3.90
Median: 3.81, 3.80
Click to expand...

upload_2019-3-26_18-20-33.jpeg
 
efle said:
Interestingly enough, the predictive validity for step 1 stayed exactly the same after switching to the new MCAT (mid 0.60s), and the USMLE just took down all the old practice NBMEs and replaced them with a few new ones, since the old ones had become invalid as predictors. Sounds like the powers that be have recognized that memorization of minutiae is a lot less important than mastery of principles and good reasoning/application skills. Bodes well for the future of medical students, in my opinion, because the current approach a lot of students use is straight-up rote memorizing 24,000 flashcards ("Zanki") which has always struck me as an absurd way to approach medical education.
Click to expand...
The way I view it, if rote memorization takes 3x longer than general concept understanding but improves Step score from a 240 to a 260...that time commitment is just simply not worth it for the score difference.
 
baylafan said:
Does anyone else find the UNC numbers to be off?
In state: 1094 apps verified, 1110 interviewed
Out of state: 5605 apps verified, 3262 interviewed
Click to expand...
It is likely a typo that should say '362.' I know that their in state preference is only pre-interview and applicants are treated the same post II (so they should theoretically have the same rate of matriculation post II as In state).
 
samualjhatfield said:
The way I view it, if rote memorization takes 3x longer than general concept understanding but improves Step score from a 240 to a 260...that time commitment is just simply not worth it for the score difference.
Click to expand...
You say that now, but after you fall in love with ortho or plastics or whatever? 18-24 months of incessant flashcarding seems a small price to pay for decades of your dream career.

I suppose there's an ethical argument to be made in favor of study time directly adding to your score, too. If step1 truly becomes more like the MCAT, a lot of people with the best work ethics will be S.O.L. because no amount of flashcards can make you a better standardized test taker
 
efle said:
You say that now, but after you fall in love with ortho or plastics or whatever? 18-24 months of incessant flashcarding seems a small price to pay for decades of your dream career.

I suppose there's an ethical argument to be made in favor of study time directly adding to your score, too. If step1 truly becomes more like the MCAT, a lot of people with the best work ethics will be S.O.L. because no amount of flashcards can make you a better standardized test taker
Click to expand...
But does increased study time directly correlate with better work ethic? It likely has a causal relationship with the level of neuroticism which can lead to perfectionism and increased work ethic.

But to say that one is willing to put in twice the effort for a ten percent better outcome I do not think translates to "better work ethic." I would argue that the other guy that puts in half the effort lower numbers but near identical outcome is the better time manager and is less likely to be stressed in a workplace environment.
 
Last edited by a moderator:
puahate said:
Should i apply to 40 schools?
Click to expand...


I did and I have no regrets. If I had to reapply I would have cut out some low-yield east coast schools and added a few more top tiers even though I am super happy with where I will be attending school this summer.

Edit: It really depends on your stats though. I was a CA ORM with a severely unbalanced gpa/MCAT.
 
samualjhatfield said:
How do so many schools even have 520+ medians? With 80,000 MCAT takers, only 800 applicants should even have >520s...
Click to expand...
520 is 98th percentile. If we're rounding up, at most 2.5 percent of applicants have 520s or better. That's 2,000 people.
 
samualjhatfield said:
Just went through all of the schools with a median >519 matriculated. about 550 of the 10000 or so applicants with a 519 or higher end up matriculating to one of 10 schools.
Click to expand...
There aren't ten thousand applicants with a 519 or better. There are about 2500.
 
Walter Raleigh said:
520 is 98th percentile. If we're rounding up, at most 2.5 percent of applicants have 520s or better. That's 2,000 people.
Click to expand...
Yeah, the math still disfavors applicants.

For example, commonly hear that about how WashU is a stats ***** with their ridiculous 99th percentile median, and so the key to getting their attention is just to score insanely high on the MCAT.

But really there are ~1,000 people per year with a score that high, and WashU only needs to fill 60 seats with those people to have that as their median. As famous as they are for their high scores, at the end of the day they're only soaking up 6% of that pool.

If someone wants to do some back of the envelope math, would be interesting to add up the class sizes of all the 520+ median schools, divide by half, and see what proportion they're soaking up overall. My guess is that the majority of the top 2,000-2,500 scorers are ending up elsewhere.
 
Walter Raleigh said:
There are around 50,000 applicants/year; 3 percent of them have 519 or better, so that's ~1500 people.
Click to expand...
You gotta go by the population that takes the MCAT though, not the population that proceeds to apply. The MCAT had like 95k yearly test takers when I last checked. 519 is top ~3% so more like ~3,000 people.

More than enough to fill all the top schools with a majority left over
 
efle said:
Yeah, the math still disfavors applicants.

For example, commonly hear that about how WashU is a stats ***** with their ridiculous 99th percentile median, and so the key to getting their attention is just to score insanely high on the MCAT.

But really there are ~1,000 people per year with a score that high, and WashU only needs to fill 60 seats with those people to have that as their median. As famous as they are for their high scores, at the end of the day they're only soaking up 6% of that pool.

If someone wants to do some back of the envelope math, would be interesting to add up the class sizes of all the 520+ median schools, divide by half, and see what proportion they're soaking up overall. My guess is that the majority of the top 2,000-2,500 scorers are ending up elsewhere.
Click to expand...
I did, it is 550 students. So, if 1500 students are >520, then that is around a third of the 99th percentiles going to one of 7 schools.
 
efle said:
Interestingly enough, the predictive validity for step 1 stayed exactly the same after switching to the new MCAT (mid 0.60s), and the USMLE just took down all the old practice NBMEs and replaced them with a few new ones, since the old ones had become invalid as predictors. Sounds like the powers that be have recognized that memorization of minutiae is a lot less important than mastery of principles and good reasoning/application skills. Bodes well for the future of medical students, in my opinion, because the current approach a lot of students use is straight-up rote memorizing 24,000 flashcards ("Zanki") which has always struck me as an absurd way to approach medical education.
Click to expand...
Regarding perfect scores: yes, a 528 is easier to get than a 45 once was. Some people counted pixels for the 528 bar in the AAMC graph, and 1/30 of a percent of test takers get perfect scores. That's around 1 in 3,000, meaning that around 25-30 people earn perfect scores each year. Old SDN scuttlebutt was that a 45 was attained only once every several years.
 
samualjhatfield said:
I did, it is 550 students. So, if 1500 students are >520, then that is around a third of the 99th percentiles going to one of 7 schools.
Click to expand...
That's a fun fact then. Without even accounting for attrition for other reasons (like bad GPA or not applying to schools in that region), if you can score a 520+ and you apply to those programs, you're in a population that has an admit rate of 33% at minimum.
 
Walter Raleigh said:
Regarding perfect scores: yes, a 528 is easier to get than a 45 once was. Some people counted pixels for the 528 bar in the AAMC graph, and 1/30 of a percent of test takers get perfect scores. That's around 1 in 3,000, meaning that around 25-30 people earn perfect scores each year. Old SDN scuttlebutt was that a 45 was attained only once every several years.
Click to expand...
Oh it's way easier. A 41 on the old scale was already 1 in 1,000 (score report of Percentile: 99.9-99.9).

As an example, on the old test a subsection breakdown of 14/13/15 was a percentile of 100/100/100. On the new test a 100th percentile is always a 132, so someone with an old 42 that looked like that was on track for a perfect score, assuming psych wasn't any harder than the others.

Another historical artifact is that the Verbal section used to be reported as 10, 11, 12, 13-15. People up this high in the score range couldn't even give an exact composite score number, they'd just be a range like "41-43"
 
efle said:
That's a fun fact then. Without even accounting for attrition for other reasons (like bad GPA or not applying to schools in that region), if you can score a 520+ and you apply to those programs, you're in a population that has an admit rate of 33% at minimum.
Click to expand...
Overall MCAT greater than 517 admit rat is 83%. I will assume it is basically the same for 520+. So 0.83*1500=1245. So right at 44%.

That is a fun fact - the acceptance rate for 520+ at the highest 7 stats schools is the same as the overall acceptance rate for all students to all medical schools.
 
Yeah, a perfect score nowadays is probably what a 42 once was. Also, it's my opinion that there isn't any great difference in test-taking skill between a 524 and anything higher, although a perfect score might mean more than a 525.
 
samualjhatfield said:
Just goes to show even with phenomenal stats and a 'wow' factor, the process is just too random to bank on applying to only a few schools.
Click to expand...
This says nothing about the process. Good scores get your app read, compelling essays and interesting activities get you interviewed. Sometimes, its the luck of who reads your app and sometimes its the luck of when you submit the app.
 
Walter Raleigh said:
Yeah, a perfect score nowadays is probably what a 42 once was. Also, it's my opinion that there isn't any great difference in test-taking skill between a 524 and anything higher, although a perfect score might mean more than a 525.
Click to expand...
Honestly, above a 520 there is little difference. At that point it is one question wrong in each category between a 525 and a 520.
 
LukaDoncicLuk said:
I did and I have no regrets. If I had to reapply I would have cut out some low-yield east coast schools and added a few more top tiers even though I am super happy with where I will be attending school this summer.

Edit: It really depends on your stats though. I was a CA ORM with a severely unbalanced gpa/MCAT.
Click to expand...

Hi, do you mind sharing your stats? Thanks
 
samualjhatfield said:
I did, it is 550 students. So, if 1500 students are >520, then that is around a third of the 99th percentiles going to one of 7 schools.
Click to expand...

Did you consider that each cycle contains applicants from different years (juniors, seniors, post-grads)? So there might be more students with high MCATs. Also, those with high MCATs are more likely to apply to med school than people with low MCATs, so the distribution of applicants would be different from the distribution of test scores.
 
wfire said:
Did you consider that each cycle contains applicants from different years (juniors, seniors, post-grads)? So there might be more students with high MCATs. Also, those with high MCATs are more likely to apply to med school than people with low MCATs, so the distribution of applicants would be different from the distribution of test scores.
Click to expand...
I assumed that everyone with high test scores will apply. That is why I used the test taker number rather than the applicant percentage.

Additionally, I just used raw percentage because the portion of people who apply the year of the test and who wait a year will be the same year-on-year. No one year is more or less likely to have a different number of applicants with high stats. If someone applying this year took their MCAT last year, there is likely someone from this year not applying until next year.
 
@samueljhatfield, I respectfully disagree; after a 520 I think it's more like two questions wrong in each category. I think there's a significant difference between a 520 and a 525, as regards content knowledge and reasoning skills. I think the demarcation's a good deal higher than 520.
 
Walter Raleigh said:
Yeah, a perfect score nowadays is probably what a 42 once was. Also, it's my opinion that there isn't any great difference in test-taking skill between a 524 and anything higher, although a perfect score might mean more than a 525.
Click to expand...
yeah it's all splitting fractions of a hair. Once you get close to perfect it comes down to a few questions because each question is worth a full composite point. I don't think you'd ever be able to tell the differences between the top 5% in a classroom setting, let alone the top fraction of a percentile.

Like the idea that getting a 520+ (with everything else adequate) gives you a 44% chance of success for this cohort of top schools is nuts. They are collectively admitting half of this tiny pool of people. They have to be sacrificing a lot of extraordinary candidates to maintain these kinds of numbers.
 
I am an MCAT tutor; I do think there is a difference between a 519 or 520 and someone that gets a 524 or 525. The 524 has more solid content knowledge and slightly fewer reasoning gaps.
 
Walter Raleigh said:
I am an MCAT tutor; I do think there is a difference between a 519 or 520 and someone that gets a 524 or 525. The 524 has more solid content knowledge and slightly fewer reasoning gaps.
Click to expand...
This feels like a very high over generalization.

It could very well be that they mixed up the Hawthorne and Thorndike effects. Or the stupidly forgot to convert to seconds from minutes on one question. Or they thought it was 3Ks and 2Nas in the ATP anti porter.

Literally one small mistake like that is the difference between those two upper level scores. I would disagree that there is any difference between those scores qualitatively.

Source: All three were mistakes I made on FLs 2 and Sample test lol
 
Walter Raleigh said:
@samueljhatfield, I respectfully disagree; after a 520 I think it's more like two questions wrong in each category. I think there's a significant difference between a 520 and a 525, as regards content knowledge and reasoning skills. I think the demarcation's a good deal higher than 520.
Click to expand...
Walter Raleigh said:
I am an MCAT tutor; I do think there is a difference between a 519 or 520 and someone that gets a 524 or 525. The 524 has more solid content knowledge and slightly fewer reasoning gaps.
Click to expand...
If I recall, they no longer disclose the exact scoring grid, but I can tell you on the old test they used to show you exactly how many wrong answers equated to each scaled score. Once you got up to 13+ in a section it was usually a single question = a point. It was crazy, if you had narrowed it down to 50-50 on a couple questions in each section, you could get anything from 39 to 45 depending how a half dozen coin flips turned out
 
Sure, and as discussed previously, a 528 is approximately a 42 on the old test. Figure 524s are as common as 39s. Wouldn't you say there's a difference between a 36 or 37 and a 39 or 40?
 
efle said:
yeah it's all splitting fractions of a hair. Once you get close to perfect it comes down to a few questions because each question is worth a full composite point. I don't think you'd ever be able to tell the differences between the top 5% in a classroom setting, let alone the top fraction of a percentile.

Like the idea that getting a 520+ (with everything else adequate) gives you a 44% chance of success for this cohort of top schools is nuts. They are collectively admitting half of this tiny pool of people. They have to be sacrificing a lot of extraordinary candidates to maintain these kinds of numbers.
Click to expand...

Interestingly enough, this document suggests that schools have some motivation for splitting hairs.

https://www.aamc.org/download/493340/data/lsl2018validity.pdf

A 524+ scorer’s likely Step 1 floor is significantly higher than a 520 scorer, in general. At the same time, the Step 1 floor (bottom 10th percentile) barely budged from 515-520. If your goal is to maximize your class’ Step average, you stack your class with as many 524+ as you can get your hands on. It’s a marginal difference (10 pt increase approx in Step floor) but it’s there.
 
Walter Raleigh said:
Sure, and as discussed previously, a 528 is approximately a 42 on the old test. Figure 524s are as common as 39s. Wouldn't you say there's a difference between a 36 or 37 and a 39 or 40?
Click to expand...
I dunno man 520 vs 524 is what, 37 vs 39? If I recall even the AAMC said their 66% retest band was 2 points. And that's for the well populated parts of the curve, up in the tail the retest band will be even wider. And even if one does impress an MCAT tutor more, should it be worth any brownie points with an admissions committee?

I should just count my blessings that for residency matching, the most competitive specialties only have their medians at the 85th percentile (250 on Step1). I cannot imagine what it would be like to have to score in the top ~2-5% to have all your options open
 
LukaDoncicLuk said:
Wow USF Morsani with 3.7/517 medians. I guess they weren't kidding when they said each class gets more competitive.
Click to expand...
The new dean of the med school said 2 years ago they would be top 25 in less than 5 years (at a forum for Jeff Vinik, Bill Gates, et al that are building the downtown Tampa area)... he came from CO, I think, where he raised that school's ranking...
 
Yes; that bears out the idea that there is a difference between 520 and 524, just like there's a difference between 516 and 520. Should it be worth any brownie points with an admissions committee? I could see that they might think the 524's a smidgen better than the 520, and when you're comparing lots of stellar candidates, that might be the tiny nudge that gets you over the edge. 520 vs. 524 is 36 vs. 39, something like that.
 
Walter Raleigh said:
Yes; that bears out the idea that there is a difference between 520 and 524, just like there's a difference between 516 and 520. Should it be worth any brownie points with an admissions committee? I could see that they might think the 524's a smidgen better than the 520, and when you're comparing lots of stellar candidates, that might be the tiny nudge that gets you over the edge. 520 vs. 524 is 36 vs. 39, something like that.
Click to expand...
There is likely also statistical bias as a 524 or higher represents what, like 100 people in total for those step 1 data?
 
I don't have the data but my Kaplan instructor said they did a study and found that there was no statistically significant difference in scores in relation to content knowledge and correctly answering a question once you got above a 518. That being said I am sure ADCOMs at top schools don't care about this and will admit the person with the higher score just to boost their media MCAT score
 
Lucca said:
Interestingly enough, this document suggests that schools have some motivation for splitting hairs.

https://www.aamc.org/download/493340/data/lsl2018validity.pdf

A 524+ scorer’s likely Step 1 floor is significantly higher than a 520 scorer, in general. At the same time, the Step 1 floor (bottom 10th percentile) barely budged from 515-520. If your goal is to maximize your class’ Step average, you stack your class with as many 524+ as you can get your hands on. It’s a marginal difference (10 pt increase approx in Step floor) but it’s there.
Click to expand...
Unless my stats basics are completely failing me right now, doesn't this chart also show that half of 524-528 scorers would be contained within the 80% CIs of values 10 points lower? Like if most people up there wouldn't even be outliers in a 514 avg classroom, let alone 518 or 520, what the hell are they in such high demand for?
 
Wow I have no chance at competitive specialties based on my mcat score. I studied for a while for it too. I’m just glad it’s step median is above 230, but I’m not sure if that’ll be affected by having only 4-6 weeks to study.

How do I do well on step if my MCAT score wasn’t that great? Get a head start with memorizing?
 
  • Like
Reactions: M&L
efle said:
Unless my stats basics are completely failing me right now, doesn't this chart also show that half of 524-528 scorers would be contained within the 80% CIs of values 10 points lower? Like if most people up there wouldn't even be outliers in a 514 avg classroom, let alone 518 or 520, what the hell are they in such high demand for?
Click to expand...

Those are not confidence intervals. The bars represent percentiles.
 
Can anyone be a pal and provide the 25%, median, 75% MCAT and GPA for Einstein? For acceptees and matriculants, please. I will give you The biggest pre med love hug ever.
 
You must log in or register to reply here.

Similar threads

0
How Accurate is MSAR for Coursework Requirements?
Replies
9
Views
1K
Med Ed
Med Ed
Mr.Smile12
2025-2026 AMCAS MSAR Reports / Discussions
Replies
0
Views
853
Mr.Smile12
Mr.Smile12
W
MSAR Letter of Intent/Interest Description
Replies
2
Views
829
Waitlistno2
W
Mayajae
  • Locked
Why are schools not accepting urm's in the 2024 msar data?
Replies
24
Views
1K
wysdoc
wysdoc
S
MSAR Waitlist Data
Replies
6
Views
3K
gyngyn
gyngyn
Top