How to interpret this bottom 10th percentile business?

[blackcadillacs]

Hi!

I'm confused about how to determine matches based on percentile info, and what 10th percentile info tells you about a school.

For example, Harvard's bottom 10th is PS 10 / BS 11, but BU's is PS 9 / BS 10 - Harvard (I'm just guessing) is going to be much harder to get into than BU but that's not really reflected in the scores, which are only a point apart.

So are 10th percentile scores just a very rough guideline for your chances at a school, since from one school to another there doesn't seem to be a huge amount of variation in these stats?

---

Members don't see this ad.

 

[TheWeeIceMan]

I'd much rather use average scores to determine matches, not 10th percentile. Personally, I use 10th percentile scores to determine if there is a potential to be screened.

 

[FrkyBgStok]

And there isn't going to be. Remember averages. If a school has an average MCAT of 36 and another 33, looking at that it is only a few points apart, but with averages that becomes a huge difference. And adding in MD schools who basically all have an average over 30 their bottom percentiles aren't going to be that much different. It is more to help you guage. If the average is 33 and the bottom percentile is high, it means out of 100 applicants, the were mostly in a range of 32-34 as opposed to mostly a range of 29-37.

 

[Praefectus]

I'd much rather use average scores to determine matches, not 10th percentile. Personally, I use 10th percentile scores to determine if there is a potential to be screened.

I would second this. There are a lot more applicants with 10th percentile stats than there are for the average, so it'll be a lot harder to stand out with such low stats.

 

[deleted103644]

To be annoying, you really want median stats, not average.

Not sure what you guys actually get in your available data.

 

[blackcadillacs]

Thank you!

MSAR provides percentile and median info, no averages - so for median, how much of a divergence do you think is acceptable? If a school's median PS is 11 and mine is 9, should that be grounds for eliminating the school?

 

[deleted103644]

Thank you!

MSAR provides percentile and median info, no averages - so for median, how much of a divergence do you think is acceptable? If a school's median PS is 11 and mine is 9, should that be grounds for eliminating the school?

For that kind of info, you really need the full distribution (which you won't get).

The other side of the coin you really need are applicant stats, not only matriculant. Not all schools receive the same caliber of applicants.

As far as advice goes, I would not rule out any school that you would really like to go to.

For more ambiguous schools, 10th percentile stats might be helpful. Unless you are above those or have a spectacular story/research/etc, you probably won't get in to that school.

 

[235788]

Don't count on an interview if you are in it (unless URM, or have some heroics ECs/LORs).

 

[OCDOCDOCD]

10th percentile isn't a good place to be. It just means you at least have a chance, but not a very good one. For example, let's say a school has an acceptance rate of 2% (pretty common). 10th percentile of 2% gives you a statistical chance of 0.2% of getting in, which means there's a 99.8% chance you'll be rejected.

Typically people in the 10th percentile had something else going for them that made up for their numbers that isn't evident in statistics. URM is the one everyone seems to scream about first, but there are other factors, like maybe they had connections to the school/ad com. Less jaded explanations would also be that they had a research background that rivaled most post-doc's (in the case of MD/PhD applicants at least) or had some amazing ECs like spearheading a vaccination campaign in sub-Saharan Africa. Also consider that the guy who got in with a 3.3 GPA may have also been a BME/physics double major at MIT and had a 42T.

In other words, if you fall into the 10th percentile the rest of your app better be AMAZING if you're to have any realistic chance of getting in.

 

[Ironmandoc]

There are no average scores published...instead median scores which mean a whole different thing in these small sample sizes

 

[TheWeeIceMan]

To be annoying, you really want median stats, not average.

Not sure what you guys actually get in your available data.

😳Yep. I was just looking at a U.S News publication, which gives average stats. MSAR, as I should have known, gives medians.

 

[gettheleadout]

10th percentile isn't a good place to be. It just means you at least have a chance, but not a very good one. For example, let's say a school has an acceptance rate of 2% (pretty common). 10th percentile of 2% gives you a statistical chance of 0.2% of getting in, which means there's a 99.8% chance you'll be rejected.

Is this actually that common? I believe Mayo's acceptance rate is about 2% and I also believe they have the lowest acceptance rate by the numbers.

 

[schwabejr]

10th percentile isn't a good place to be. It just means you at least have a chance, but not a very good one. For example, let's say a school has an acceptance rate of 2% (pretty common). 10th percentile of 2% gives you a statistical chance of 0.2% of getting in, which means there's a 99.8% chance you'll be rejected.

I know alot of people are aware of this but I feel the need to say it anyways:

Folks, remember that admissions committees aren't drawing names out of a hat to pick applicants. There's more at play than what percentile you fall in. Not a personal stab OCD, just using your post to illustrate a point.


2% seems really, really low. Are you sure ya'll aren't talking about matriculation rate? The MSAR, from what I remember, ends the table with this info using the number that matriculated. This can be really misleading, especially for state schools which typically have a low yield of OOS acceptances matriculating.

 

[TheWeeIceMan]

Is this actually that common? I believe Mayo's acceptance rate is about 2% and I also believe they have the lowest acceptance rate by the numbers.

U.S News has an acceptance rate section, I believe. Most of the schools that I saw were somewhere in the 5-9% range. There were a few obvious outliers on the high end due to some extremely IS friendly schools.

 

[Ironmandoc]

Georgetown, GWU, Drexel, and other 10,000-12,000 app schools have comparable acceptance rates to Mayo...less than 2%

 

[deleted103644]

Georgetown, GWU, Drexel, and other 10,000-12,000 app schools have comparable acceptance rates to Mayo...less than 2%

I wouldn't pay too close attention to Mayo.

Many people don't apply at all due to location (one of the worst unless you're local), so those stats won't extrapolate well.

If you want a mainstream example of good + competitive, stick to UCSF or Harvard.

 

[gettheleadout]

U.S News has an acceptance rate section, I believe. Most of the schools that I saw were somewhere in the 5-9% range. There were a few obvious outliers on the high end due to some extremely IS friendly schools.

Ah yes, I forgot about that. Just logged into my account there.

Georgetown, GWU, Drexel, and other 10,000-12,000 app schools have comparable acceptance rates to Mayo...less than 2%

Wrong. As of 2012 USNews data, only three schools have acceptance rates below 3.5%, Mayo, GWU, and Stanford. Of those only Mayo is below 2.5%.

Georgetown's acceptance rate is slightly above 3.5% and Drexel's is slightly above 5%.

For Johnny, UCSF and Harvard are both slightly above 4%.

 

[GagaOOlala]

OP i've been wondering this as well. I interpret it literally, ie in a class of 200 there are at least 20 people with mcat scores below or at the 10th percentile. of course, the circumstances surrounding these "low" scores are impossible to know, and a range of accepted mcat scores would be soooo nice to have.

 

[deleted103644]

OP i've been wondering this as well. I interpret it literally, ie in a class of 200 there are at least 20 people with mcat scores below or at the 10th percentile. of course, the circumstances surrounding these "low" scores are impossible to know, and a range of accepted mcat scores would be soooo nice to have.

Um, that's basically the definition of percentiles.

The issue is that med school admissions stats are probably at least bimodal.

There are people in the lower distribution with amazing extracurriculars/research/donations/diversity/etc who will overcome weaker stats.

Unless you have something that stands out though, you probably can't afford to be in the lower region.

 

[1289]

Um, that's basically the definition of percentiles.

The issue is that med school admissions stats are probably at least bimodal.

There are people in the lower distribution with amazing extracurriculars/research/donations/diversity/etc who will overcome weaker stats.

Unless you have something that stands out though, you probably can't afford to be in the lower region.

I think GaGa was saying that they don't try to read too much into it. if there are ~20 people in a 200 class who scored in the 10th percentile, he/she could get potentially get accepted if they score in that range. but yes, as you've said, they probably had something extraordinary in their app to help overcome it.

 

[gettheleadout]

Um, that's basically the definition of percentiles.

The issue is that med school admissions stats are probably at least bimodal.

Any data to support a bimodal distribution? If we're going to postulate distribution curves, we should probably have some data...

 

[GagaOOlala]

I think GaGa was saying that they don't try to read too much into it. if there are ~20 people in a 200 class who scored in the 10th percentile, he/she could get potentially get accepted if they score in that range. but yes, as you've said, they probably had something extraordinary in their app to help overcome it.

Yes. that's what i meant.

just to share a personal story, a friend of mine who was a second-time applicant to med school this year was accepted to a 50-75 rank school with an mcat one point below the 10th percentile. he's asian, and he used his year off after getting rejected the first time to work as a cna and do some volunteering, he has no research experience. Solid ECs and experience, but nothing truly "extraordinary." of course this could be an anomaly, but it happened!

 

[NickNaylor]

Yes. that's what i meant.

just to share a personal story, a friend of mine who was a second-time applicant to med school this year was accepted to a 50-75 rank school with an mcat one point below the 10th percentile. he's asian, and he used his year off after getting rejected the first time to work as a cna and do some volunteering, he has no research experience. Solid ECs and experience, but nothing truly "extraordinary." of course this could be an anomaly, but it happened!

Yes, everyone knows a "friend" or "someone" that gets accepted despite ________. However, applicants would do well to fully understand that those are extremely exceptions and more than likely won't happen to them.

(sent from my phone - please forgive typos and brevity)

 

[deleted103644]

Any data to support a bimodal distribution? If we're going to postulate distribution curves, we should probably have some data...

Just because of the selection process. It's probably much more complex, but if you treat the decision process as being pulled from x*(GPA + MCAT) + y*N where N is either 0 or 1 if the student has great research, faculty parent, etc, you'd get a bimodal distribution.

 

[GagaOOlala]

Yes, everyone knows a "friend" or "someone" that gets accepted despite ________. However, applicants would do well to fully understand that those are extremely exceptions and more than likely won't happen to them.

(sent from my phone - please forgive typos and brevity)

Yes hence my disclaimer of "this could be an anomaly"

 

[tantacles]

10th percentile isn't a good place to be. It just means you at least have a chance, but not a very good one. For example, let's say a school has an acceptance rate of 2% (pretty common). 10th percentile of 2% gives you a statistical chance of 0.2% of getting in, which means there's a 99.8% chance you'll be rejected.

This is some fuzzy, fuzzy math. The sentiment is on the ball, but the math offends me.

 

[camshark]

Probably a dumb question...is the 10% cgpa the lowest gpa they accepted or is there any lower that got in? I know I have like a .1% if Im lower but I would like to add a couple more MD schools to apply to (im going mostly DO). I would have much higher cgpa if my grades from 6 years ago (in high school) didnt count so I would like to shoot for the moon.

 

[tantacles]

Probably a dumb question...is the 10% cgpa the lowest gpa they accepted or is there any lower that got in? I know I have like a .1% if Im lower but I would like to add a couple more MD schools to apply to (im going mostly DO). I would have much higher cgpa if my grades from 6 years ago (in high school) didnt count so I would like to shoot for the moon.

10th percentile simply means that 10% of the students who got in had a GPA at or lower than the number stated.

 

[deleted103644]

10th percentile simply means that 10% of the students who got in had a GPA at or lower than the number stated.

Precisely.

Although they may be giving the average of the bottom 10 percent, which is different and less than the 10th percentile cutoff.

 

[MiracleforMD]

I would say if you are anywhere near 10th percentile on MCAT or GPA you better be well above average and even as high as 90th percentile on the other to have a chance. Assuming average ECs.

 

[vayntraubinator]

You also need to take into consideration the combination of stats. A bottom 10% at UMiami may be an 8, but the other sections could be well above 90%, like 12-15. Someone with a 9 in PS may still get into UPenn with a 36 overall. The process is so chaotic, so you need to judge where to apply very logically but put forth your best effort. I could see a great PS/ECs easily differentiating you if you fall in these categories.

 

[Dbate]

how do you view the median stats on the online version of the MSAR? Because all I see are the 10th and 90th percentiles.

 

[1289]

how do you view the median stats on the online version of the MSAR? Because all I see are the 10th and 90th percentiles.

Huh??? Are you scrolling down the page w all the schools or looking at "Selection Factors" within an individual school?

 

[Dbate]

Huh??? Are you scrolling down the page w all the schools or looking at "Selection Factors" within an individual school?

I see something that looks exactly like this, but with the numbers included because I bought the MSAR:

https://services.aamc.org/30/msar/home

 

[1289]

yeah you have to click on the individual schools, there is a world of information you're missing out on 🙂