Let me preface this post with two things: 1) I really want an honest discussion with facts involved without flaming/racism, and that 2) I believe some URM consideration is necessary because of racism and cultural issues (e.g. few race-appropriate role models) independent of financial considerations, and also because of the disparity in the proportion of URMs in the doctors and patient population.
However, I strongly believe that the current magnitude of URM consideration is too high. To address this I compare African-American to Asian-American populations because they represent the statistically lowest and statistically highest groups. For counterargument #5, all non-URM and all URM are considered.
The average MCAT for accepted Asian-Americans is 6 points higher than that for accepted African-Americans.*
Accepted African-Americans: 26.1 average, 3.6 SD
Accepted Asian-Americans: 32.2 average, 3.6 SD
The average sGPA for accepted Asian-Americans is 0.4 higher than that for accepted African-Americans.*
Accepted African-Americans: 3.28 average, 0.41 SD
Accepted Asian-Americans: 3.62 average, 0.29 SD
*Source #1: http://www.aamc.org/data/facts/applicantmatriculant/start.htm
**Source #2: MSAR
a. Counterargument 1: Morehouse, Howard, and Meharry are responsible for the relatively low average MCAT/GPA of African-Americans.
This is not true. If we factor out the three schools, the Accepted African average MCAT goes from 26.1 to 26.3 and the Accepted African average GPA decreases from 3.28 to 3.27.
Calculations: (26.1*1336-25*82-25*80-27*42) / (1136-82-80-42) = 26.3. **
b. Counterargument 2: The Carribean schools are responsible for relatively low average MCAT/GPA of African-Americans.
This is not likely. Only 10% of the Ponce, U Central del Carribean, and San Juan Bautista are black (total of 18 people at the three schools).** This does not significantly affect averages.
c. Counterargument 3: Many African-Americans are accepted with higher MCATs than Asian-Americans.
This is partially true, but it is dependent on your definition of many. Based on the standard deviations listed above, and the 95th percentile African-American accepted applicant is equivalent to the 50th percentile Asian-American applicant.
This means that if you are an average Asian-American in medical school, only 5% of fellow students who are African-American will have a higher MCAT than you.
If you are a 25th percentile Asian-American in medical school, only 16% of fellow students who are African-American will have a higher MCAT than you. If you are a 75th percentile Asian-American in medical school, only 1% of fellow students who are African-American will have a higher MCAT than you.
d. Counterargument #4: Only 36.5% of African-Americans get accepted, versus 44% of Asian-Americans.
This statistic addresses the need for URM, but is irrelevant to the extent of URM. It is explained mostly by the statistics of applicants*.
African-American applicant: average GPA=3.23, MCAT=21.6
Asian-American applicant: average GPA=3.52, MCAT=29.0 (note, higher statistics than the average African-American accepted)
e. Counterargument #5: URMs form a small enough proportion of the medical school class that it does not influence your chances of acceptance.
This is not true. The extent of influence can be addressed statistically.
The grand mean MCAT for all matriculants is 30.8, with a standard deviation of 4.1. There exists 18390 matriculants.*
matriculants:
URM: average MCAT=27.2, stdev=4.24, n=2593
non-URM: average MCAT=31.5, stdev=3.71, n=15763
applicants:
URM: average MCAT=24, stdev=4.24, n=6278
non-URM: average MCAT=28.6, stdev=3.71, n=35875
Currently, in order to have a 50% chance of being accepted as a non-URM your MCAT needs to be at the 78th percentile of the non-URM group. Since half of the accepted people are above and half are below, the calculations suggest that 7892*2, or 15784 total non-URM are accepted (very close to the actual number, 15763).
In order to have a 50% chance of being accepted without any URM consideration (assume the new non-URM mean = current grand mean), you need to be in the 72nd percentile, or 20090 total non-URM are accepted.
This means that 4,306 people (about 12%) are not accepted as a direct result of them not being URM.
However, I strongly believe that the current magnitude of URM consideration is too high. To address this I compare African-American to Asian-American populations because they represent the statistically lowest and statistically highest groups. For counterargument #5, all non-URM and all URM are considered.
The average MCAT for accepted Asian-Americans is 6 points higher than that for accepted African-Americans.*
Accepted African-Americans: 26.1 average, 3.6 SD
Accepted Asian-Americans: 32.2 average, 3.6 SD
The average sGPA for accepted Asian-Americans is 0.4 higher than that for accepted African-Americans.*
Accepted African-Americans: 3.28 average, 0.41 SD
Accepted Asian-Americans: 3.62 average, 0.29 SD
*Source #1: http://www.aamc.org/data/facts/applicantmatriculant/start.htm
**Source #2: MSAR
a. Counterargument 1: Morehouse, Howard, and Meharry are responsible for the relatively low average MCAT/GPA of African-Americans.
This is not true. If we factor out the three schools, the Accepted African average MCAT goes from 26.1 to 26.3 and the Accepted African average GPA decreases from 3.28 to 3.27.
Calculations: (26.1*1336-25*82-25*80-27*42) / (1136-82-80-42) = 26.3. **
b. Counterargument 2: The Carribean schools are responsible for relatively low average MCAT/GPA of African-Americans.
This is not likely. Only 10% of the Ponce, U Central del Carribean, and San Juan Bautista are black (total of 18 people at the three schools).** This does not significantly affect averages.
c. Counterargument 3: Many African-Americans are accepted with higher MCATs than Asian-Americans.
This is partially true, but it is dependent on your definition of many. Based on the standard deviations listed above, and the 95th percentile African-American accepted applicant is equivalent to the 50th percentile Asian-American applicant.
This means that if you are an average Asian-American in medical school, only 5% of fellow students who are African-American will have a higher MCAT than you.
If you are a 25th percentile Asian-American in medical school, only 16% of fellow students who are African-American will have a higher MCAT than you. If you are a 75th percentile Asian-American in medical school, only 1% of fellow students who are African-American will have a higher MCAT than you.
d. Counterargument #4: Only 36.5% of African-Americans get accepted, versus 44% of Asian-Americans.
This statistic addresses the need for URM, but is irrelevant to the extent of URM. It is explained mostly by the statistics of applicants*.
African-American applicant: average GPA=3.23, MCAT=21.6
Asian-American applicant: average GPA=3.52, MCAT=29.0 (note, higher statistics than the average African-American accepted)
e. Counterargument #5: URMs form a small enough proportion of the medical school class that it does not influence your chances of acceptance.
This is not true. The extent of influence can be addressed statistically.
The grand mean MCAT for all matriculants is 30.8, with a standard deviation of 4.1. There exists 18390 matriculants.*
matriculants:
URM: average MCAT=27.2, stdev=4.24, n=2593
non-URM: average MCAT=31.5, stdev=3.71, n=15763
applicants:
URM: average MCAT=24, stdev=4.24, n=6278
non-URM: average MCAT=28.6, stdev=3.71, n=35875
Currently, in order to have a 50% chance of being accepted as a non-URM your MCAT needs to be at the 78th percentile of the non-URM group. Since half of the accepted people are above and half are below, the calculations suggest that 7892*2, or 15784 total non-URM are accepted (very close to the actual number, 15763).
In order to have a 50% chance of being accepted without any URM consideration (assume the new non-URM mean = current grand mean), you need to be in the 72nd percentile, or 20090 total non-URM are accepted.
This means that 4,306 people (about 12%) are not accepted as a direct result of them not being URM.