"National Median MCAT"

[DoctorWhoo]

---

Members don't see this ad.

"*National Median refers to data derived by including MCAT scores from all accepted applications and includes duplicate scores from individuals with multiple acceptances. The national median for each individual test section remains identical when calculated using duplicate scores or a single record from an accepted applicant, except in the case of the national median for the Total MCAT Score, which is 31 when derived using unduplicated scores."


I stumbled upon this interesting side note in MSAR... Any thoughts as to why they prefer the national median that is derived from multiple acceptances(ie double counting high scores)? The obvious answer is that it's more daunting to premeds, but I'd like to think that the AAMC is a bit more kindhearted than that haha.

I had a hard time believing that the national median mcat for accepted applicants was above the 90th percentile... But ~80th seems much more reasonable.

 

[LizzyM]

Keep in mind that in a given year almost 60% of applicants do not get in anywhere and the lower the MCAT score the greater the likelihood of not being admitted. Consider, too, that some folks with poor MCAT scores never even apply to med school. Thus the national median for accepted students will be at a percentile that is much higher than might otherwise be expected.

On the other hand, half of all accepted applicants had a MCAT of 31 or less.

 

[NickNaylor]

Why? For some people the MCAT is the first shock to reality that they may not be able to be a physician - either because they don't want to work hard enough or don't have an interest in the material. It's one thing to rationalize to yourself not doing well in courses or whatever your excuse might be up to that point. It's another altogether to take the MCAT and get smacked with a percentile score that probably wasn't what you were expecting. The mean MCAT for 2013 applicants was 27-28. I'm sure the data is out there, but my guess is that the mean for test-takers generally is at least a few points below that.

 

[Heday]

Why? For some people the MCAT is the first shock to reality that they may not be able to be a physician - either because they don't want to work hard enough or don't have an interest in the material. It's one thing to rationalize to yourself not doing well in courses or whatever your excuse might be up to that point. It's another altogether to take the MCAT and get smacked with a percentile score that probably wasn't what you were expecting. The mean MCAT for 2013 applicants was 27-28. I'm sure the data is out there, but my guess is that the mean for test-takers generally is at least a few points below that.

Mean was 25.3 for test-takers. Mean MCAT for applicants was 28.4. Mean MCAT for matriculants was 31.3.

Sources:
https://www.aamc.org/students/download/361080/data/combined13.pdf.pdf
https://www.aamc.org/download/321494/data/2013factstable17.pdf

Now, OP's question as to why they use duplicated scores for acceptances. The reason I would imagine is that the average MCAT for an acceptance should be appropriately represented. While both a 28 and a 38 may get an acceptance, let's say the 28 gets 2 and the 38 gets 10, the 38 should have a heavier weight because it resulted in more acceptances. I'm not sure if that makes sense, but I think it's very similar to using a weighted average (eg GPA) rather than a simple average because 1 hour credit courses would have much more bearing than a 3 hour course (which would "inflate" someone's GPA).

 

[NickNaylor]

Mean was 25.3 for test-takers. Mean MCAT for applicants was 28.4. Mean MCAT for matriculants was 31.3.

Sources:
https://www.aamc.org/students/download/361080/data/combined13.pdf.pdf
https://www.aamc.org/download/321494/data/2013factstable17.pdf

Now, OP's question as to why they use duplicated scores for acceptances. The reason I would imagine is that the average MCAT for an acceptance should be appropriately represented. While both a 28 and a 38 may get an acceptance, let's say the 28 gets 2 and the 38 gets 10, the 38 should have a heavier weight because it resulted in more acceptances. I'm not sure if that makes sense, but I think it's very similar to using a weighted average (eg GPA) rather than a simple average because 1 hour credit courses would have much more bearing than a 3 hour course (which would "inflate" someone's GPA).

If that was the intent then the AAMC should describe it as a weighted mean rather than a strict mean to prevent confusion (as in this case). I agree with the reasoning in that it makes the information more useful in a way, but perhaps it would be better to report both.

 

[V5RED]

Mean was 25.3 for test-takers. Mean MCAT for applicants was 28.4. Mean MCAT for matriculants was 31.3.

Sources:
https://www.aamc.org/students/download/361080/data/combined13.pdf.pdf
https://www.aamc.org/download/321494/data/2013factstable17.pdf

Now, OP's question as to why they use duplicated scores for acceptances. The reason I would imagine is that the average MCAT for an acceptance should be appropriately represented. While both a 28 and a 38 may get an acceptance, let's say the 28 gets 2 and the 38 gets 10, the 38 should have a heavier weight because it resulted in more acceptances. I'm not sure if that makes sense, but I think it's very similar to using a weighted average (eg GPA) rather than a simple average because 1 hour credit courses would have much more bearing than a 3 hour course (which would "inflate" someone's GPA).

That method might be interesting for some people, but for applicants, it is not very helpful. The weighted average will tell you where schools give their acceptances. The unweighted(unduplicated) average tells you who ends up matriculating.

The first tells you what medical schools hope to get in an applicant. The second tells applicants how strong of an MCAT an applicant needs to get in somewhere.

 

[ascott189]

It would be interesting if each school would publish a distribution of their MCAT scores for each class.

 

[lmn]

It would be interesting if each school would publish a distribution of their MCAT scores for each class.

Why would they want to do that? Schools want to boast the highest number they (reasonably) can in terms of gpa's and mcat's to increase their reputation, quality of applicants, etc. If they can say their average was a 35, that is much better than saying 25% of the class had a score of 30 or lower.

 

[Goro]

See MSAR Online.
The spread goes something like this (10th-90th %ile). The higher the tier of school, the higher the median.
Wash U: 34-41
Harvard: 32 (!)-41
Pitt and similar: 31-39
U Iowa: 29-37
Rosy Franklin: 27-36
It would be interesting if each school would publish a distribution of their MCAT scores for each class.

To add to my learned colleague's comment, keep in mind that the school medians are lower in state schools in the midwest (say, U KS) and the south (U AL, LSU), which seems to pull the numbers down, and limit the competitiveness of applicants with scores <30, since state schools favor the home team. The rough rule of thumb I use is the one should have a score of 30 or more. 28, 29 aren't lethal, but do limit your chances. 27 is circling the drain for MD schools. Interestingly, DO schools have seen their MCAT medians go up in recent years as well.

Keep in mind that in a given year almost 60% of applicants do not get in anywhere and the lower the MCAT score the greater the likelihood of not being admitted. Consider, too, that some folks with poor MCAT scores never even apply to med school. Thus the national median for accepted students will be at a percentile that is much higher than might otherwise be expected.
On the other hand, half of all accepted applicants had a MCAT of 31 or less.