Elite School Bias

[snowonkey]

Traveling out on the interview trail, I've been lucky enough to interview at a good number of top 10 and top 20 type MSTP programs. One interesting observation- among interviewees and students at these places, nearly all came from some top undergraduate institutions (Ivy leagues, Stanford, MIT, Caltech, WUSTL, Duke, Northwestern). While it's been said many times here that undergraduate institution doesn't matter much in terms of admission, the bias towards this type of school seems much heavier than just based on the type of students at these schools.

Anybody else notice the same thing? I know a lot of self selection and hard work also has made it this way, but I also think that name recognition for undergrad institution may mean more than schools (and people in these forums) let on.

Full disclosure- I come from one of these type of schools. Not making a judgement on whether this is bad or good, just interesting.

---

Members don't see this ad.

 

[carpediem22]

Traveling out on the interview trail, I've been lucky enough to interview at a good number of top 10 and top 20 type MSTP programs. One interesting observation- among interviewees and students at these places, nearly all came from some top undergraduate institutions (Ivy leagues, Stanford, MIT, Caltech, WUSTL, Duke, Northwestern). While it's been said many times here that undergraduate institution doesn't matter much in terms of admission, the bias towards this type of school seems much heavier than just based on the type of students at these schools.

Anybody else notice the same thing? I know a lot of self selection and hard work also has made it this way, but I also think that name recognition for undergrad institution may mean more than schools (and people in these forums) let on.

Full disclosure- I come from one of these type of schools. Not making a judgement on whether this is bad or good, just interesting.

I've noticed the same thing. Also top liberal arts colleges like Williams and Middlebury. Anecdotally, I think school name matters more for MD/PhD admissions than it might for MD admissions. Or maybe these "elite" schools just tend to offer more impressive research opportunities to undergrads, thereby producing more qualified MSTP applicants? Would be cool to hear some adcom's opinions on this.

 

[Sreltraln]

+1 to being interested on adcom opinion! Coming from a state school in the midwest I definitely felt the school bias at my last interview. I'd say at least 75% were from Ivy Leagues and/or other top undergrad institutions. I hope school name doesn't affect admissions but it does "seem" to affect interview invites. I'm wondering if school bias is more prominent at higher ranked institutions too?

 

[mTORC]

Since you can't change anything about it now, why worry about it?

it may be good to hear an opinion for people applying to undergrad though. Unfortunately what I've noticed at my school is that the majority of people interviewed come from extremely high tier schools. I find it hard to believe it's all merit-based or school-blind; they definitely seem to seek out people from certain schools. I'm not from a ridiculously high level undergrad and I felt my level of achievement had to be 2-3x as high to compete at the same level.

 

[neurotroph]

I think it's self-selection and the resources provided at these schools more than anything else. Higher tier schools tend to have more of the type of people who would be drawn to a career like this (probably part of why they chose to attend a top research university in the first place), and they're already in the habit of attaining the grades, standardized tests scores, etc. necessary to be a competitive applicant. Then there's the much easier access to quality research opportunities, the greater availability of physician-scientist mentors, and pre-med advising offices that actually know that the MD/PhD is a thing. I say this as someone who's attended both a no-name public uni and a top private uni. No doubt there are some unconscious (or probably just conscious) biases also at work. A degree from a top school automatically guarantees a higher level of qualification in people's minds, regardless of field. People from "non-elite" schools have to work harder to prove that they're at that level since it's not assumed like it is for elite school grads. But I don't think that's the major factor shaping the makeup of the interview pool by any means.

 

[QofQuimica]

I think it's self-selection and the resources provided at these schools more than anything else. Higher tier schools tend to have more of the type of people who would be drawn to a career like this (probably part of why they chose to attend a top research university in the first place), and they're already in the habit of attaining the grades, standardized tests scores, etc. necessary to be a competitive applicant. Then there's the much easier access to quality research opportunities, the greater availability of physician-scientist mentors, and pre-med advising offices that actually know that the MD/PhD is a thing. I say this as someone who's attended both a no-name public uni and a top private uni. No doubt there are some unconscious (or probably just conscious) biases also at work. A degree from a top school automatically guarantees a higher level of qualification in people's minds, regardless of field. People from "non-elite" schools have to work harder to prove that they're at that level since it's not assumed like it is for elite school grads. But I don't think that's the major factor shaping the makeup of the interview pool by any means.

Agree.

OP, while name recognition may make some difference in some cases, I'd agree with the person I quoted above that the main factor at work here is the high concentration of top students that you get at the big name institutions. Just about all the students at those schools are top students. Whereas, while the very best handful of students at other schools are just as accomplished, there is a much wider range of academic achievement among the students attending your local state U. Hey, all those B and C students from your HS classes had to go somewhere, right?

 

[Lucca]

Agree.

OP, while name recognition may make some difference in some cases, I'd agree with the person I quoted above that the main factor at work here is the high concentration of top students that you get at the big name institutions. Just about all the students at those schools are top students. Whereas, while the very best handful of students at other schools are just as accomplished, there is a much wider range of academic achievement among the students attending your local state U. Hey, all those B and C students from your HS classes had to go somewhere, right?

People always say things like this repeatedly, but if you look at the class profiles for the top MSTP programs usually about 3-5 schools are represented per class of 10-20 and MAYBE one of them will be a public school and even then it is probably Berkeley. As you go down in "ranking" you see a lot more public school graduates. I don't buy the resources argument either. There are far more tier-1 public research institutions than elite private ones. Academics don't scoff at research that comes from Illinois or the UCs or UVa or UWisconsin etc etc because these schools definitely have significant resources for undergraduates interested in research. Maybe it's true that more "top school" grads are interested in academics than state school grads but I also think that is untrue. Private school classes are tiny next to state school classes and 50% of those top school grads just shove themselves into the Ibanking / consulting / tech pipeline anyway and most will still probably apply MD only. I'm absolutely certain my public school sends way more students to both medical school, MD/PhD and PhD programs than Harvard or Stanford every year, the only difference is how prestigious those programs might be.

I'm not trying to be antagonistic Q, you are one of my favorite posters, I would just like a stronger argument or show of evidence against what I feel to be the simple fact that academics recruiting future academics want their applicants to have a shiny pedigree than "well those schools are just better and recruit better students so those students do better". I just don't think it is true. Maybe it was true when Harvard and Co. had a 20% acceptance rate for undergrad and they were actually doing something that resembled "choosing" for undergrad admissions but now basically nobody besides the Super Privileged and the Super Inspirational and the Super Lucky (I.e went to the right prep school for high school) get into those schools on a regular basis and most other students who would have been admitted to the top tier 5 years ago just say **** it and go to a school that offers them money so at least they can graduate without debt.


Rant over and I might be projecting but I think there's a better explanation out there for this or OPs original assumption is just correct.

 

[Fencer]

Admission officers have access to a handbook of GPA MCAT data arranged by undergraduate schools by MD (includes MD/PhD) applicants and matriculants. The success rate (defined as matriculants/applicants) for Ivy/top 40 USN&WR private schools is greater than for public Ivy or public schools. However, the number of matriculants from the public schools is much greater than private. This is not public data but it is widely available to counselors and admission officers. Unfortunately, there is no easily available data about the undergraduate colleges of MD/PhD applicants/matriculants. Some undergraduate schools track this very closely, for example, every year, I get an email (inviting us to a Fair) from U. of Michigan-Ann Arbor indicating that they produce 25-30 MD/PhD matriculants.

 

[downthedrain]

The admission's committee at my program uses a point system for selecting candidates. If you come from an elite school, you get an extra 10 pts added to your score. On the flip side, I think it might just be a self-selecting bias for MD/PhD candidates. If you go to a school with a lot more resources, then there is a greater chance you'll do research or work with mentors who are MD/PhDs which will then influence your decision to apply to MD/PhD programs. Also, I feel like there is a certain amount of hubris among students at elite schools (I went to a "top 20" private research university for undergrad), so when you see that you can get two degrees at the same time, who wouldn't want to gun for that?

 

[QofQuimica]

People always say things like this repeatedly, but if you look at the class profiles for the top MSTP programs usually about 3-5 schools are represented per class of 10-20 and MAYBE one of them will be a public school and even then it is probably Berkeley. As you go down in "ranking" you see a lot more public school graduates.

My point though is that you are not comparing apples to apples when you compare the students attending Large State U (LSU) to the students attending Prestigious U (PU). Take a sample of 100 premed students from each school, and what do you get? At LSU, you get a wide range of students, from C students all the way up to A students. A few of the very top LSU students will have the credentials to be competitive for top medical schools, and those will be the few that you see making it. In contrast, the majority of the LSU students (and/or their advisors) will be smart enough to self-select, realizing that their grades and MCAT scores aren't competitive for top medical programs, and they won't even bother applying to them. Whereas, the same sample of 100 premeds at PU will have a much tighter SD around that A level mean. In other words, a much larger percentage of them will be A-level students who are competitive for top med schools. Because PU selects a higher-achieving student body starting at the *high school* level when they fill their freshman class, it is not surprising that, four years later, they then have a higher-achieving graduating student body, on average, than LSU does.

So, while the very top students at LSU and PU are more or less equally accomplished, the AVERAGE students at LSU are much less accomplished than the AVERAGE students at PU are. You can easily graduate from high school with a B average, attend LSU and maintain a B average there, and be an average student at LSU. Whereas, a B student will be BELOW average at PU; there really are no C-level premed students at PU; and thus the whole curve is shifted upward. Does that make more sense?

Edit: FWIW, I'm not saying there is NO bias at all whatsover in favor of grads from certain schools. Adcoms, after all, are human like everyone else. As human beings, we all tend to look more favorably on people who are more similar to us than we do on people who are more different. That being said, biases can work both ways (e.g., maybe I am more biased TOWARD state school grads, nontrads, etc. since I am those things). And the school you attended for UG is still much less important than how you performed once you got there.

 

[Lucca]

My point though is that you are not comparing apples to apples when you compare the students attending Large State U (LSU) to the students attending Prestigious U (PU). Take a sample of 100 premed students from each school, and what do you get? At LSU, you get a wide range of students, from C students all the way up to A students. A few of the very top LSU students will have the credentials to be competitive for top medical schools, and those will be the few that you see making it. In contrast, the majority of the LSU students (and/or their advisors) will be smart enough to self-select, realizing that their grades and MCAT scores aren't competitive for top medical programs, and they won't even bother applying to them. Whereas, the same sample of 100 premeds at PU will have a much tighter SD around that A level mean. In other words, a much larger percentage of them will be A-level students who are competitive for top med schools. Because PU selects a higher-achieving student body starting at the *high school* level when they fill their freshman class, it is not surprising that, four years later, they then have a higher-achieving graduating student body, on average, than LSU does.

So, while the very top students at LSU and PU are more or less equally accomplished, the AVERAGE students at LSU are much less accomplished than the AVERAGE students at PU are. You can easily graduate from high school with a B average, attend LSU and maintain a B average there, and be an average student at LSU. Whereas, a B student will be BELOW average at PU; there really are no C-level premed students at PU; and thus the whole curve is shifted upward. Does that make more sense?

Edit: FWIW, I'm not saying there is NO bias at all whatsover in favor of grads from certain schools. Adcoms, after all, are human like everyone else. As human beings, we all tend to look more favorably on people who are more similar to us than we do on people who are more different. That being said, biases can work both ways (e.g., maybe I am more biased TOWARD state school grads, nontrads, etc. since I am those things). And the school you attended for UG is still much less important than how you performed once you got there.

I definitely understand what you mean and I agree that individual performance is probably always more important than school prestige but what I'm getting at is that there are a very limited number of spots at top MSTPs and at that level of competition the candidates are essentially homogenous in terms of qualifications with maybe a few exceptions and it seems to be the case that top MSTPs prefer students from top programs.

There are so many public institutions with their own broad spectrum of resource availability and quality; it is not the binary "top private school" "not top private school" dynamic that is often presented. There should be many, many more equally qualified applicants from public schools, even considering just the few top public schools (Mich, Berkeley, Virginia etc) than those from top private schools and yet the elite private institutions are far more represented at top programs.

The only conclusion I can draw is that there is a significant bias in top MSTP admissions in the direction of elite private school applicants, even when non elite applicants are identically qualified / desirable for whatever reason. Suggesting therefore that they perceive their pedigree to have intrinsic value. Will this matter for most applicants to most programs? No I don't think attending a public school is going to keep people from the MSTPs (especially if you excel) maybe even from top ones but I do think they are at a significant disadvantage.

 

[Bodysurfer]

As a side note, I plan on applying for MD/PhD program and I'll be interested in my success as I graduated from large state university but I work as a research technician at a prestige hospital (top 10 nationally) that's affliated with an Ivy League med school

 

[Nanorust]

I'm a bit late to the party, but while touring for interviews this semester I have been told directly, offhandedly, and in whispers by a mix of students, interviewers, and adcom members that my being from Large Southern StateU will end up translating to extra diversity points. Now, I have no idea whether diversity points always cancel out prestige points, but the consensus in my case seems to be that the elite schools (which interview and accept primarily students from elite schools) see StateU as new and different. (Read: More of the matriculating students will be from elite schools for reasons stated before, but an equivalent student from StateU actually gets a tiny boost for bringing a diverse perspective.)

It is also worth noting that while public schools might produce larger numbers of capable students, those students are not pushed toward MD/PhD (especially top tier programs) in the same way as the students that go to schools with large MSTPs. Put it this way, in the last 5 years my StateU has sent 100% of our MD/PhD applicants to MD/PhD programs (according to our pre-health advising director). It just so happens that 100% over 5 years is a total of ~5 students, mostly because many of the top MD applicants from StateU have no idea that MD/PhD is an option until its too late. If you go around looking for students from StateU, you will see a very small number of them in MSTPs, but StateU has a better record of matriculating MD/PhD students than most "top 20" institutions.

I admit that n=1 in this study, but I have seen no reason to thing that my institution has been anything but helpful in the process. I have also never felt like I had to work significantly harder than any interviewee from any institution to make it into any of my interviews to date.

 

[carpediem22]

I'm a bit late to the party, but while touring for interviews this semester I have been told directly, offhandedly, and in whispers by a mix of students, interviewers, and adcom members that my being from Large Southern StateU will end up translating to extra diversity points. Now, I have no idea whether diversity points always cancel out prestige points, but the consensus in my case seems to be that the elite schools (which interview and accept primarily students from elite schools) see StateU as new and different. (Read: More of the matriculating students will be from elite schools for reasons stated before, but an equivalent student from StateU actually gets a tiny boost for bringing a diverse perspective.)

It is also worth noting that while public schools might produce larger numbers of capable students, those students are not pushed toward MD/PhD (especially top tier programs) in the same way as the students that go to schools with large MSTPs. Put it this way, in the last 5 years my StateU has sent 100% of our MD/PhD applicants to MD/PhD programs (according to our pre-health advising director). It just so happens that 100% over 5 years is a total of ~5 students, mostly because many of the top MD applicants from StateU have no idea that MD/PhD is an option until its too late. If you go around looking for students from StateU, you will see a very small number of them in MSTPs, but StateU has a better record of matriculating MD/PhD students than most "top 20" institutions.

I admit that n=1 in this study, but I have seen no reason to thing that my institution has been anything but helpful in the process. I have also never felt like I had to work significantly harder than any interviewee from any institution to make it into any of my interviews to date.

Eh, I don't know about "better record" -- lots of top 20 schools have close to 100% matriculation too, for larger numbers of students. Here is 2012 data that I found for Yale College for example: http://ocs.yale.edu/sites/default/files/med_school_stats.pdf. Tl;dr = 20 out of 21 applicants matriculated into MD/PhDs, with the 1 who didn't choosing MD over MD/PhD.

Not sayin that the diversity thing isn't true, just that the track record stat isn't a factor.

 

[Fencer]

Here is data for the 2015 national MD/PhD class of 623 matriculants. Their undergraduate institutions were:

Primary Undergraduate College ................ 623 ..... TOTAL ... PERCENTAGE (of the class up to this school)
1 Harvard University ................................... 25
2 University of California-Los Angeles ....... 20
3 Cornell University .................................... 19
3 Yale University ......................................... 19
5 Massachusetts Institute of Technology ... 18
6 University of California-Berkeley ............. 17
7 Washington University in St. Louis ......... 16
8 Columbia University ................................. 13
8 Johns Hopkins University ....................... 13
8 University of Florida ................................ 13 ....... 173 .............. 28% of the class in the Top 10 UG Colleges
11 Stanford University ................................. 12
12 University of Michigan-Ann Arbor ........ 11
12 University of Pennsylvania .................... 11
14 University of Chicago .............................. 10
14 University of Washington ....................... 10 ........ 227 ............. 36% of the class in the Top 15 UG Colleges
16 Duke University ...................................... 9
16 Princeton University ............................... 9
18 Northwestern University-Evanston ........ 8
18 University of Pittsburgh ......................... 8 ......... 261 ............. 42% of the class in the Top 19 UG Colleges
20 University of California-San Diego ............ 7
20 University of Illinois at Urbana-Champaign.. 7
20 University of Maryland-College Park ............ 7
20 University of Wisconsin-Madison ................ 7 .......... 289 ............ 46% of the class in the Top 23 UG Colleges

325 - 52% of the class in the Top 29 UG Colleges
360 - 58% of the class in the Top 36 UG Colleges
412 - 66% of the class in the Top 49 UG Colleges​

 

[Fencer]

Out of the top 49 UG colleges (based upon the number of MD/PhD matriculants):

• 28 are private institutions and produced 253 MD/PhD matriculants (41% of the 2015 class)
• 21 are public institutions and produced 159 MD/PhD matriculants (26% of the 2015 class)

 

[Fencer]

Top PUBLIC UG universities:

Rank - UG university - # of MD/PhD matriculants in 2015 entering class
2 University of California-Los Angeles .....................20
6 University of California-Berkeley ..........................17
8 University of Florida .............................................13
12 University of Michigan-Ann Arbor .....................11
14 University of Washington ...................................10
18 University of Pittsburgh ...................................... 8
20 University of California-San Diego ..................... 7
20 University of Illinois at Urbana-Champaign ...... 7
20 University of Maryland-College Park ................. 7
20 University of Wisconsin-Madison ..................... 7
24 University of Colorado at Boulder ..................... 6
24 University of Maryland-Baltimore County ....... 6
30 University of Georgia ........................................ 5
30 University of Iowa ............................................ 5
30 University of Miami .......................................... 5
30 University of North Carolina at Chapel Hill ..... 5
37 McGill University .............................................. 4
37 Rutgers University New Brunswick Campus .... 4
37 University of Connecticut ................................. 4
37 University of Minnesota-Twin Cities ................ 4
37 University of Texas at Austin ............................ 4

 

[Fencer]

Top PRIVATE UG universities

Rank - UG university - # of MD/PhD matriculants in 2015 entering class
1 Harvard University .................................................. 25
3 Cornell University ................................................... 19
3 Yale University ........................................................ 19
5 Massachusetts Institute of Technology .................. 18
7 Washington University in St. Louis ........................ 16
8 Columbia University in the City of New York ........ 13
8 Johns Hopkins University ..................................... 13
11 Stanford University ............................................... 12
12 University of Pennsylvania ................................... 11
14 University of Chicago ........................................... 10
16 Duke University .................................................... 9
16 Princeton University ............................................ 9
18 Northwestern University-Evanston ..................... 8
24 Dartmouth College ............................................... 6
24 Davidson College .................................................. 6
24 New York University ............................................ 6
24 University of Rochester ....................................... 6
30 Emory University ................................................. 5
30 Temple University ................................................ 5
30 University of Notre Dame .................................... 5
37 Bowdoin College ................................................... 4
37 Brown University ................................................. 4
37 Georgetown University ......................................... 4
37 Haverford College ................................................. 4
37 Pomona College .................................................... 4
37 Swarthmore College .............................................. 4
37 Tufts University ..................................................... 4
37 University of Southern California ......................... 4

 

[Lucca]

So I went ahead and did some quick recalculations by matriculants from each "tier" of undergrad institution rather than pure volume.

Here is what I got:

Percentage of MD/PhD seats filled by undergraduates from the top 10: 22%
...from the top 20: 34%
...from the top 30: 43%

Of the top 30 schools, 4 of them are public and they are [ school (USNWR rank) ]: UC Berkeley (20), UCLA (23), U Michigan - Ann Arbor (29), and UNC Chapel Hill (30).

I also used @efle 's data from a thread in pre-allo to determine what percentage of the pre-med applicants from a particular institution eventually became MD/PhD matriculants. The fractions were predictably very small.

Take the few schools that supply the most MD/PhD matriculants for example:

School Name (# of MD/PhD matriculants) - Percentage of supplied applicants who become MD / PhD matriculants

1. Harvard (25) - 7.7%
2. UCLA (20) - 2.1%
3. Cornell (19) - 3.7%
3. Yale (19) - 8.2%
5. MIT (18) - 17.8% (nearly 1/5th of MITs applicants go on to become MD/PhD matriculants!! Notably, MIT only had 101 MD or MD / PhD applicants in 2014)
6. UC Berkeley (17) - 2.2%


So one fifth of MD/PhD seats are filled by top 10 applicants and nearly half are filled by applicants from just 30 schools, most of them private. Also, MIT is an extraordinary producer of potential physician scientists.

What is missing is data of class composition by program ranking. My hypothesis is that schools at the tippity top are almost exclusively populated by applicants from top 10 institutions (maybe top 20) and as you decrease in ranking class diversity actually begins to increase, suggesting that the bias for pedigreed undergraduates is most present at, somewhat predictably, elite dual degree programs themselves.

UPDATE: As an exemplar of my above hypothesis I tabulated Harvard's MSTP class data from the past 7 years and got the following results:

N = 163
% From Top 10 Undergrad: 72%
% From Top 20 Undergrad: 81%
% From Top 30 Undergrad: 86%

A few schools within the Top 10 are very well represented:
% HMS MSTP students from Harvard: 24.5%
% ...from MIT: 11%
% ...from Yale: 9.8%
% ...from Stanford: 7.4%
% ...from Princeton: 4.3%

57% of HMS MSTP students come from just 5 undergrads, typically known among helicopter parents and high school students as HYPSM (or in this particular case, HMYSP) and fully a quarter were Harvard undergrads as well.

URM Info:
% URM: 6.8%
% From Top 10 and URM: 1.2% (precisely 2 people lol)
% From Top 20 and URM: 1..8% (3 people)
% From Top 30 and URM: 1.8% (3 people)
% Of URM coming from Top 10, 20, 30 as a fraction of all URM students: 18%, 27%, 27%

Furthermore, of the HMS MSTP students who did not graduate from a Top 30 undergraduate, 35% were URM. Two thirds of the URM students at HMS' MSTP graduated from non-top 30 undergraduate institutions.



NOTE: Universities without USNWR ranking such as Williams, Amherst, Swarthmore, Oxford / Cambridge etc were assigned the next best available ranking. There were very few of these schools included in the data (Williams is the largest with 4 matriculants across 7 years).

 

[neurotroph]

What is missing is data of class composition by program ranking. My hypothesis is that schools at the tippity top are almost exclusively populated by applicants from top 10 institutions (maybe top 20) and as you decrease in ranking class diversity actually begins to increase, suggesting that the bias for pedigreed undergraduates is most present at, somewhat predictably, elite dual degree programs themselves.

This data is easy to compile from the current students pages that most MD/PhD programs have. Aside from Harvard's program, I haven't found your hypothesis to be true. The almost exclusively top 10 part at least. I'm sure there are fewer top 10 grads at lower ranked programs for reasons other than pure bias that have already been discussed.

 

[Lucca]

This data is easy to compile from the current students pages that most MD/PhD programs have. Aside from Harvard's program, I haven't found your hypothesis to be true. The almost exclusively top 10 part at least. I'm sure there are fewer top 10 grads at lower ranked programs for reasons other than pure bias that have already been discussed.

Im in the process of compiling it for a few more schools across different rankings but its time consuming so it might take a while. Harvard's program might be a hard outlier but its the only one ive done!

 

[Doctor&Geek]

Good work, team.

 

[Lucca]

Here is the data for Johns Hopkins from 2008-2015. I wrote some scripts so this is a lot easier now if the website is uniformly formatted.

N = 90
% From Top 10 Undergrad: 49%
% From Top 20 Undergrad: 59%
% From Top 30 Undergrad: 67%

The diversity of school representation is greater than at HMS. In spite of the program being 50% the size of Harvard's, JHU's MSTP sample represented over 50 undergraduate institutions, while 44 were represented at Harvard; obviously, a far smaller proportion were from the top 30 schools at JHU than at HMS.

The most represented university is JHU whose students comprise 12% of the sample. The next most represented universities are Harvard, Duke, and Cornell - each at 5.6% of the sample.


URM Info:
N = 11
% URM: 12.2%
% From Top 10 and URM: 5.6% (n=5)
% From Top 20 and URM: 6.7% (n=6)
% From Top 30 and URM: 6.7% (n=6)
% Of URM coming from Top 10, 20, 30 as a fraction of all URM students: 45%, 54%, 54%

Once again, in spite of having nearly half the sample size of HMS's MSTP from the same years, JHU's program has just as many URMs (11) and a higher portion come from Top 30 institutions when compared to URMs ar Harvard.

If other schools follow JHU's trend, I might have been very wrong from the start (and I am happy to be wrong in this case). Kind of ridiculous how different JHU's composition looks to Harvard's it actually reflects pretty badly on HMS in my eyes.

 

[Lucca]

Stanford

N=86
% Top 10: 42%
% Top 20: 58%
% Top 30: 66%

Most represented undergrads:
Harvard - 9.3%
Stanford - 8.1%
MIT - 7.0%
Yale, UCLA- 5.8%

41 total undergrads are represented.

 

[efle]

So I went ahead and did some quick recalculations by matriculants from each "tier" of undergrad institution rather than pure volume.

Here is what I got:

Percentage of MD/PhD seats filled by undergraduates from the top 10: 22%
...from the top 20: 34%
...from the top 30: 43%

Of the top 30 schools, 4 of them are public and they are [ school (USNWR rank) ]: UC Berkeley (20), UCLA (23), U Michigan - Ann Arbor (29), and UNC Chapel Hill (30).

I also used @efle 's data from a thread in pre-allo to determine what percentage of the pre-med applicants from a particular institution eventually became MD/PhD matriculants. The fractions were predictably very small.

Take the few schools that supply the most MD/PhD matriculants for example:

School Name (# of MD/PhD matriculants) - Percentage of supplied applicants who become MD / PhD matriculants

1. Harvard (25) - 7.7%
2. UCLA (20) - 2.1%
3. Cornell (19) - 3.7%
3. Yale (19) - 8.2%
5. MIT (18) - 17.8% (nearly 1/5th of MITs applicants go on to become MD/PhD matriculants!! Notably, MIT only had 101 MD or MD / PhD applicants in 2014)
6. UC Berkeley (17) - 2.2%


So one fifth of MD/PhD seats are filled by top 10 applicants and nearly half are filled by applicants from just 30 schools, most of them private. Also, MIT is an extraordinary producer of potential physician scientists.

What is missing is data of class composition by program ranking. My hypothesis is that schools at the tippity top are almost exclusively populated by applicants from top 10 institutions (maybe top 20) and as you decrease in ranking class diversity actually begins to increase, suggesting that the bias for pedigreed undergraduates is most present at, somewhat predictably, elite dual degree programs themselves.

UPDATE: As an exemplar of my above hypothesis I tabulated Harvard's MSTP class data from the past 7 years and got the following results:

N = 163
% From Top 10 Undergrad: 72%
% From Top 20 Undergrad: 81%
% From Top 30 Undergrad: 86%

A few schools within the Top 10 are very well represented:
% HMS MSTP students from Harvard: 24.5%
% ...from MIT: 11%
% ...from Yale: 9.8%
% ...from Stanford: 7.4%
% ...from Princeton: 4.3%

57% of HMS MSTP students come from just 5 undergrads, typically known among helicopter parents and high school students as HYPSM (or in this particular case, HMYSP) and fully a quarter were Harvard undergrads as well.

URM Info:
% URM: 6.8%
% From Top 10 and URM: 1.2% (precisely 2 people lol)
% From Top 20 and URM: 1..8% (3 people)
% From Top 30 and URM: 1.8% (3 people)
% Of URM coming from Top 10, 20, 30 as a fraction of all URM students: 18%, 27%, 27%

Furthermore, of the HMS MSTP students who did not graduate from a Top 30 undergraduate, 35% were URM. Two thirds of the URM students at HMS' MSTP graduated from non-top 30 undergraduate institutions.



NOTE: Universities without USNWR ranking such as Williams, Amherst, Swarthmore, Oxford / Cambridge etc were assigned the next best available ranking. There were very few of these schools included in the data (Williams is the largest with 4 matriculants across 7 years).

I was not aware that the # of applicants per undergrad reported by AAMC included both MD and MD/PhD apps. Good work as always

 

[Lucca]

I was not aware that the # of applicants per undergrad reported by AAMC included both MD and MD/PhD apps. Good work as always

It is unclear to me if they do however I am guessing that they do because you have to apply MD on AMCAS and check off MD/PhD so I imagine all of this data is compiled all at once. I could be wrong.

 

[efle]

It is unclear to me if they do however I am guessing that they do because you have to apply MD on AMCAS and check off MD/PhD so I imagine all of this data is compiled all at once. I could be wrong.

And even then I'd guess the majority apply to a couple pure MD programs so it's a safe assumption

 

[Fencer]

Please note that the MD/PhD data about primary undergraduate colleges pertains to Matriculants into MD/PhD programs, and specific data from each MD/PhD program is publicly available on each of the MD/PhD program websites. I only want to indicate that the 623 MD/PhD students came from 207 different primary undergraduate institutions including many very small colleges (go for the little non-elite college). Data on the UG colleges of MD/PhD applicants is not publicly available. Information about the number of applicants for each MD/PhD program is available in the AAMC table 33: https://www.aamc.org/download/321544/data/factstable33.pdf

 

[youououa]

I would caution against using website photos (apologies if this isn't what you did!) to identify URM students. While this isn't likely the case at JHU and HMS, different schools can select populations that they believe are under-represented - gyngyn has mentioned this in the pre-allo forum on occasion.

Once again, in spite of having nearly half the sample size of HMS's MSTP from the same years, JHU's program has just as many URMs (11) and a higher portion come from Top 30 institutions when compared to URMs ar Harvard.

 

[Lucca]

I would caution against using website photos (apologies if this isn't what you did!) to identify URM students. While this isn't likely the case at JHU and HMS, different schools can select populations that they believe are under-represented - gyngyn has mentioned this in the pre-allo forum on occasion.

While that may be true for the MD side, the T32 training grant from the NIH / MSTP funding utilizes the following definition:

"Blacks or African Americans, Hispanics or Latinos, American Indians or Alaska Natives, Native Hawaiians and other Pacific Islanders and B) individuals with disabilities, defined as those with a physical or mental impairment that substantially limits one or more major life activities."

 

[youououa]

Thanks - I didn't realize there was a strict definition for MSTP students. Still, the broader point is that there's inherent bias in using appearance to decide whether a student falls into those categories.

While that may be true for the MD side, the T32 training grant from the NIH / MSTP funding utilizes the following definition:

"Blacks or African Americans, Hispanics or Latinos, American Indians or Alaska Natives, Native Hawaiians and other Pacific Islanders and B) individuals with disabilities, defined as those with a physical or mental impairment that substantially limits one or more major life activities."

 

[Lucca]

Thanks - I didn't realize there was a strict definition for MSTP students. Still, the broader point is that there's inherent bias in using appearance to decide whether a student falls into those categories.

Yah there are inaccuracies but the number of URM applicants is so small for MD PhD to begin with it is kind of specious to draw any conclusions from URM data at all. I included it since I was just curious.

 

[Fencer]

Here is the 2014 MD/PhD data on self-reported race and ethnicity:

M.D.-Ph.D. Matriculants, 2014
American Indian or Alaska Native ........................ 1
Asian .................................................................. 137
Black or African American ..................................26
Hispanic, Latino, or of Spanish Origin ................33
Native Hawaiian or Other Pacific Islander ............ 1
White ..................................................................308
Other ..................................................................... 11
Multiple Race/Ethnicity ....................................... 37
Unknown Race/Ethnicity .................................... 58
Non-U.S. Citizen and Non-Permanent Resident .. 14
Total .....................................................................626

Using NIH Training grant definition of URM .......... 61
Training grant eligible (US Citizens & Residents) .... 612
URMs ....................................................................... 10%

Source: https://www.aamc.org/download/321546/data/factstable34.pdf

 

[Nanorust]

This thread is why I love physician scientists.

 

[Elixir6]

The admission's committee at my program uses a point system for selecting candidates. If you come from an elite school, you get an extra 10 pts added to your score. On the flip side, I think it might just be a self-selecting bias for MD/PhD candidates. If you go to a school with a lot more resources, then there is a greater chance you'll do research or work with mentors who are MD/PhDs which will then influence your decision to apply to MD/PhD programs. Also, I feel like there is a certain amount of hubris among students at elite schools (I went to a "top 20" private research university for undergrad), so when you see that you can get two degrees at the same time, who wouldn't want to gun for that?

No offense, but what a crock of horse**** that policy reflects. I guess I can't knock you too hard because I assume most programs do this to one degree or another, but I guess your program had the decency to not hide the fact that pedigree can trump merit and actually baked it right into their algorithm.

We should value merit over pedigree where possible, understanding that sometimes they are intertwined since often pedigree comes from merit. However, I would argue the opposite and I favor someone who comes from a non-elite place who has a similar application to someone from an elite place on the grounds that the non-elite background person likely had to work harder to build her credentials.

 

[Fencer]

The admission committee of my program adds up to a 0.5 point to disadvantaged students and/or underrepresented students to an academic score that is the beginning of the conversation as to whether we should consider that particular student or not for interviews [interviews are most critical for admission or not]. The thing here is that those applicants who are truly disadvantaged might had to be working 2 or 3 part-time jobs along with trying to do research and coursework. Their GPAs would have likely suffered, and the actual question is whether the applicant did the most with the opportunities that they had available. Once in a lab, they might need to pursue a post-bac to truly succeed (shedding those part-time jobs) earning their stripes, which would have been more easily earned if the applicant came from an Elite private or public Ivy school.

In my view, the real sign of an outstanding MSTP is that they take some of these students and produce terrific clinician scientists who otherwise would have been lost by our society. An elite MD/PhD program who only accepts elite students is not adding much "quality" training to them. This might be a minority view, not popular or reflective of my peers, but that is my view.

 

[Maebea]

I would not argue with Fencer on this, though I think many would (particularly those programs that are heavily populated with graduates from Top 10 schools.

 

[Maebea]

Here's data on where the students come from.
Legend: Total number of reported MD-PhD matriculants from 2000 to 2014
Percentage of matriculants from USNWR Rankings of Top 10 Universities (U 10), 11-20 (U 20), and 21-30 (U 30); Top 10 LA Colleges (C 10), 11-20 (C 20), and 21-30 (C30); and all Other schools.

Program................................ Total...............U 10..........U 20 ......... U 30 ............C 10............C 20................ C 30 .....................Other
AECOM ...................................66 ................11% ............9% .............12% .............2% .............2% ...................2%....................... 64%
Arizona .....................................5 .................0% ............0%.............. 0% .............0% .............0% ...................0% .....................100%
Arkansas ..................................12 ................0%............ 0%............... 0% ............0%............. 0% ...................0%..................... 100%
Baylor ......................................49 ...............22% ..........16%.............. 6% .............6% .............0% ...................0% .......................49%
Boston U .................................33 ................15% ...........9% ...............9% ............9% ..............3%................... 0% ......................55%
Buffalo ....................................18 ................11% ............0% ..............17% ............0% .............0% ...................0% ......................72%
Case Western ..........................65 ................17% ..........20% .............11%............ 0% ..............0% ...................2% .......................51%
Chicago ..................................40 ................45% ..........13% .............18% .............3%............. 0% ..................0% .......................23%
Cincinnati.............................. 36 ................28% ..........3% ................3% .............0% .............6% ...................0%....................... 61%
Colorado................................ 46 ..................9% ..........11% ...............7% ..............2%............ 0% ....................2% .......................70%
Columbia ...............................78 .................55% ........15% ...............4% ...............5% .............3% ...................1% ........................17%
Connecticut ...........................20 .................15% ........10% ...............5% ..............0% ............10% ...................0% ......................60%
Cornell ...................................75................. 36% .........17% ..............5% ..............7% ..............0% ...................1% .......................33%
Dartmouth ............................15 ..................13%......... 13% ..............13% .............7% ............13% ...................0% ......................40%
Drexel ..................................20 ..................10% ...........5% ...............5% ..............5% ............0% ....................0% ......................75%
Duke ....................................48 ..................25% ..........13% ..............15% ............0% .............2% ....................0% .....................46%
Emory ...................................54 .................35% ...........4% ..............24% ............2% .............2% .....................0% ....................33%
Florida ..................................25 ...................4% ............8% ..............8% .............0%............ 0% ....................0% ......................80%
Georgetown ............................6.................. 17% ............0% ..............17% ............0% ............0% ....................0% ......................67%
GRU-MCG ............................20 .................15% .............0% ..............5% .............0% ............5% ....................0% ............ .....75%
Harvard.................................68...................74% ............3% ...............3% .............7% ............0% ....................0% .......................13%
Hofstra ..................................16.................. 19% ............6%............... 6% .............0% ............0% ...................0%...................... 69%
Howard ...................................3................... 0% ...........0%................0% ..............0%........... 0% ....................0% ....................100%
Illinois ..................................70 ..................10% ...........10%............ 11% ...............3%............ 1% ....................0% .......................64%
Illinois-Chicago .....................31 .................19% ............13% ..............3% ..............6%........... 0% ....................0% .......................58%
Indiana................................. 33 ..................6% ............18% ..............6% .............0% ............0% ....................0% ........................70%
Iowa ......................................47 ..................2% .............6% .............19% ............0% .............4% ....................0% ........................68%
Jefferson ...............................24 ................25% ............4% ................4% ...........0% .............4% ....................0% ........................63%
Johns Hopkins..................... 55 .................38% ...........9% ................2% ............7% .............2% ...................5% ........................36%
Kansas ...................................22 ...................9% ...........5% ...............0% ............0% ............0% ...................0% .........................86%
Kentucky ...............................16 ...................6% ...........6% ...............19% ...........6% ............0% ....................0% ........................63%
Loma Linda ...........................13 ...................0% ............8% ..............0% .............0% ...........0% ....................0%....................... 92%
Louisville ...............................12 ...................8% ...........8% ..............17% .............0% ...........8% ....................0% .......................58%
Loyola-Chicago .....................10 ...................10% ........40% ..............0% ..............0% ...........0% ....................0% .......................50%
LSU-New Orleans ................27 ....................11% ...........0% .............4% ...............0% ...........0% ....................0% .......................85%
Marshall................................ 4 ....................25%.......... 0% .............0%............... 0%........... 0%.................... 0% ........................75%
Maryland .............................33 ...................21% ...........9% ..............15%............. 6% ............0% ....................0% .......................48%
Massachusetts .....................51 ....................14% ..........6%.............. 10% .............4%............ 4%.................... 2% .........................61%
Mayo ...................................31 .....................0% ...........3% ...............3% .............3% ............0%.................... 0%........................ 90%
MCW ..................................25 .....................4% ..........12% ..............12%............ 0% ............0%.................... 0%......................... 72%
Miami .................................26 ....................15% ..........12%............... 8%............ 4% ............0% ....................0%........................ 62%
Michigan .............................61 ....................26% ..........16% .............20% ...........7% .............3% ...................0% .........................28%
Michigan State ...................12 .....................0% .............0% .............25%...........0% .............0%....................0%.......................... 75%
Minnesota .........................34 .....................9% ..............9% .............6% ............9% ..............3% ...................0%......................... 65%
Mississippi ...........................7..................... 0% .............0%............. 0%............ 0%.............. 0% ...................0%....................... 100%
Missouri.............................. 7 ....................14% .............0% .............0% ............0% ..............0%.................... 0% ........................86%
Mt. Sinai ..........................54 ....................30% .............13% ............4%............ 6% ...............4% ....................0%........................ 44%
MUSC ...............................32..................... 0%.............. 3% ............9% ..............3% .............0% .....................3% ........................81%
Nebraska ..........................21.......................0%............. 0%............ 10% ............0% ...............0%................... 0%........................ 90%
Nevada............................... 1 ......................0% .............0% ..........100% ............0% ...............0% ...................0% ..........................0%
New Mexico ......................6 .....................0%.............. 0% .............17% ............0% .................0%................. 0% ........................83%
Northwestern ..................67 ....................31%............ 16% .............10% ............3% .................0% .................0% ........................39%
NYU ................................51 .....................33% ............10% .............14% ............2% ................2% ..................0% ........................39%
Ohio State ......................28 .....................18% ............14%............. 18% ............4% ...............0% ...................4% ........................43%
Oklahoma ........................5...................... 0% .............0% ...............0% .............0%............... 0% ...................0% ......................100%
Oregon Health ..............20 ......................0% .............15% ...............0% .............0% ..............5% ...................5% .........................75%
Penn ............................109 .....................53% .............11% ..............8% ..............5%...............2% ...................2% .........................19%
Penn State.................... 39 ......................21% .............10%.............. 5% .............0% ...............3%................... 0% .........................62%
Pitt ...............................60 .......................18%............ 15% ..............8%............. 5% ...............2% ...................0% .........................52%
Rochester.................... 40....................... 10%............. 18% ..............5% .............0% ..............3% ....................0% .........................65%
Rutgers-RWJ ................46...................... 15% ..............0%............... 9% .............2%.............. 2% ....................2%.......................... 70%
South Dakota ...............10........................ 20%............ 20% ............0% .............0% ..............0% ....................0% ..........................60%
South Florida .................2 .........................0% ..............0% ..............0%............. 0% ............0% .....................0%........................ 100%
St. Louis U.................... 13 .........................0%.............. 0% ..............8% ............0% ..............0% ....................0%.......................... 92%
Stanford .......................52........................ 44% .............10% ............10% ...........4% ...............0% ....................0% .........................33%
Stony Brook ................32 .........................38%............... 13% ...........3%........... 3% ...............3% .....................0% .........................41%
SUNY Downstate ..........8..........................13% ................0% ............25% ............0%........... 0% ......................13% ......................50%
SUNY Upstate ................7 .........................0%................. 57% ............0% ............0% ..............0% ....................0% .........................43%
Temple ..........................17........................ 12% .................0% .............6% ............0% ..............0%..................... 0% ........................82%
Texas ...............................7....................... 29%................ 14% ............0%.............14%............. 0% .....................0%......................... 43%
Texas A&M................... 18 ........................0% ..................17%..............6% ...........6% ...............0% ....................0% ........................72%
Texas Houston .............23 ........................13% .................22% ...........13% ...........0% ...............0% ....................0%........................ 52%
Texas Tech................... 10 ........................0%.................. 20% ...........10% ............0% ...............0%................... 0% ........................70%
Texas-San Antonio ......22 .........................5% ..................18%............ 5% ..............0% .............0% ....................0% ........................73%
Texas-Southwestern ......47 ......................19%................... 9% .............9% ..............4% .............0% ....................0%....................... 60%
TMB-Galveston .............19 ......................11% ...................5% .............5%.............. 0% ..............0% ....................0% .......................79%
Toledo........................... 10....................... 0%.................. 10% ...........10% .............0% ..............0%.................... 0%....................... 80%
Tufts ............................23....................... 22% .................22%............. 4% .............4% ..............4%..................... 0% .......................43%
Tulane ..........................3 .........................0%.................. 0% ..............0% ..............0% .............0%..................... 0%.................... 100%
U. Washington ...........51 ........................18% ................22%............ 12% ..............6% ..............0% ....................0%....................... 43%
UAB ............................37 ........................5% .................14% ..............0% ..............0%............. 3% ....................0% .......................78%
UC-Davis........ .............5..........................20% ................20%.............20%.............20%.............0%....................0%.......................20%
UC-Irvine..............28...............................25%.................11%................4%..............4%...............0%.....................0%......................57%
UC-Los Angeles.61 38% 15% 21% 3% 2% 0% 21%
UC-San Diego 42 38% 10% 14% 5% 0% 0% 33%
UC-San Francisco 52 48% 17% 8% 4% 4% 0% 19%
UNC 47 30% 6% 23% 0% 2% 0% 38%
Uniformed Services 9 0% 0% 0% 0% 11% 0% 89%
USC 17 29% 6% 24% 6% 0% 0% 35%
Utah 17 18% 12% 6% 6% 0% 0% 59%
Vanderbilt 58 14% 26% 14% 5% 5% 0% 36%
VCU 43 9% 9% 12% 2% 0% 0% 67%
Virginia 32 25% 13% 19% 3% 0% 3% 38%
Wake Forest 12 33% 8% 25% 0% 8% 0% 25%
Washington U 123 20% 28% 9% 0% 2% 2% 39%
Wayne State 25 4% 4% 32% 0% 0% 0% 60%
Wisconsin 44 14% 14% 5% 2% 9% 0% 57%
Wright State 1 0% 0% 100% 0% 0% 0% 0%
Yale 68 40% 16% 9% 3% 0% 1% 31%
Total 3143

 

[Fencer]

I suspect this is data from the national MD/PhD survey. Given that the USNWR Rankings for R1 and Liberal Universities change slightly every year, I assume that you only used one given year of rankings for the 2000-14 period.

 

[Maebea]

Given that the USNWR Rankings for R1 and Liberal Universities change slightly every year, I assume that you only used one given year of rankings for the 2000-14 period.

I used the most recent USNWR rankings, figuring that the annual changes would not significantly alter the picture.

Overall, 50% of MD PhD matriculants hail from "non-elite" schools (i.e., non-Top 30 colleges and universities). Pitt (52%) is the only Top 20 med school this is particularly meritorious in matriculating individuals from non-elite schools, while a handful of others (Duke, MSSM, UW, Northwestern, NYU, Wash U, and Hopkins) exceed 33%.

 

[downthedrain]

No offense, but what a crock of horse**** that policy reflects. I guess I can't knock you too hard because I assume most programs do this to one degree or another, but I guess your program had the decency to not hide the fact that pedigree can trump merit and actually baked it right into their algorithm.

We should value merit over pedigree where possible, understanding that sometimes they are intertwined since often pedigree comes from merit. However, I would argue the opposite and I favor someone who comes from a non-elite place who has a similar application to someone from an elite place on the grounds that the non-elite background person likely had to work harder to build her credentials.

I should've specified that this is for the entire medical program, not just the MD/PhD program. And the definition of an "elite school" that gets used is one that has a notoriously difficult undergraduate program. How they determine that, I'm not sure, but for example, my undergrad program used to curve the class avg down to a C, even if your score gave you an 80%, which really sucked.

I agree it's not the best strategy for selecting students, but I guess it's another way to figure out if a student from a really tough, competitive school with a 3.5 is equivalent to a student from a less rigorous program with 4.0, beyond just using an MCAT. In the end, it's worked well for my program, with people doing really well on boards and residency matches.

 

[Fencer]

I agree it's not the best strategy for selecting students, but I guess it's another way to figure out if a student from a really tough, competitive school with a 3.5 is equivalent to a student from a less rigorous program with 4.0, beyond just using an MCAT. In the end, it's worked well for my program, with people doing really well on boards and residency matches.

One of the tools that program directors (PDs) have is a book (published annually by AAMC) with the stats for GPA & MCAT of every UG school with more than 5 applicants in AMCAS. This annual book shows the mean academic numbers for applicants and matriculants, providing good data to PDs about grade inflation or deflation.

 

[Elixir6]

One of the tools that program directors (PDs) have is a book (published annually by AAMC) with the stats for GPA & MCAT of every UG school with more than 5 applicants in AMCAS. This annual book shows the mean academic numbers for applicants and matriculants, providing good data to PDs about grade inflation or deflation.

Now that's interesting to know