The numbers game: Selectivity beyond median scores

[Bovary]

When prospective applicants ask for advice on where to apply, they're usually directed to compare their stats (MCAT & GPA) with the median scores that are accepted to, or matriculate at, any given med school. This is generally good advice. However, two schools with the same median stats can't be equally selective, because at some schools there are more applicants competing for fewer spots.

For example, using public data, and letting the number of matriculants serve as a mediocre relative approximation of the number of accepted applicants:

Rush and Loma Linda both have median MCATs of 31. At Rush, 7701 applicants competed last year for a total of 128 seats (1.66% matriculated). At Loma Linda, fewer applicants (5217) competed for more seats (168) (3.22% matriculated).

Looking at LizzyM profiles alone, Rush and Loma Linda would appear equally selective; however, from these numbers one could argue that an average applicant has double the chance of getting into Loma Linda compared to Rush. (This ignores that some schools are clearly a better "fit" for any given applicant than others, but I'm trying to just crunch numbers right now).

There are a handful of schools that are attractive to many because they have relatively low median stats but are also located in large cities. Notably, George Washington has a median MCAT of only 30, but it gets more applicants than literally any other school. Last year, 14509 applicants vied for 177 spots (1.22%).
Is George Washington therefore perhaps a worse choice for someone trying to play the numbers game than its neighbor Georgetown where, despite a higher average MCAT (32), only 11733 applicants compete for 196 seats (1.67%).

Is this a productive line of reasoning when assessing relative competitiveness and selectivity? Are confounds, notably the variable self-selectivity of schools, too influential to draw anything from the results?

---

Members don't see this ad.

 

[Aerus]

The thing is that SDN DOES take into consideration the number of applicants compared to available seats. That's why people are regularly discouraged from applying to schools like Brown, Georgetown, WashU (out of state), etc. unless they have a compelling reason to apply.

Loma Linda is a religious school. Not many people would be a good fit for this school. So it isn't that Rush is that much more competitive; Loma Linda just doesn't get many applicants in general lol.

 

[seeinghowitgoes]

I agree with Aerus, but I also think matriculation rate, when looked at as a percentage to two decimal places, also paints a misleading/ignorable picture. Acceptance rates, albeit perhaps harder to get ahold of , would be slightly better (essentially I doubt the premise that matriculation rate serves as an acceptable indicator for acceptance rates), but still probably not all that important when comparing schools that are so similar.

For example, George Washington may only have 177 spots, but if it takes 300 people to fill those spots versus Georgetown only taking 200 to fill it's 196 spots (clearly unrealistic), then the acceptance rates go to 2.07% vs. 1.70%. Not that this difference is all that much anyway, so somebody applying to Georgetown vs. GW really shouldn't consider themselves as having a "better" chance at once versus the other (in essence, even if it's "statistically significant" I'd highly doubt it's "clinically significant").

In the end, we're often-times comparing schools that are on the same level (similar stats, prestige, etc) and probably getting a lot of the same applicants, so I would say yes to your second question (at least when comparing schools that are so similar) and that these factors are really embodied in most advice given on SDN anyway.

 

[RogueUnicorn]

OP I think the number crunching is too heavily influenced by confounding factors (yield, limited pool of desired applicants, etc) to be significantly worthwhile. As noted above acceptance rate/yield can pretty dramatically swing your numbers.

The thing is that SDN DOES take into consideration the number of applicants compared to available seats. That's why people are regularly discouraged from applying to schools like Brown, Georgetown, WashU (out of state), etc. unless they have a compelling reason to apply.

Loma Linda is a religious school. Not many people would be a good fit for this school. So it isn't that Rush is that much more competitive; Loma Linda just doesn't get many applicants in general lol.

Christ. GTown and WashU are private schools without a strong in-state bias of any sort. Especially since GTown doesn't even have a state. If you're going to speak for SDN you should get your facts straight

 

[Goro]

Your reasoning assumes that all of the applicants are similar. They are not. Based upon the stats at my own school, at least 50% of applicants have no business applying to medical school.

Interestingly, at the institutions I've been at, the same mentality applies to faculty job seekers. For example, for a recent anatomy position, about half of the applicants had no experience in anatomy at all (like Drosophila geneticists or pharmacologists. I guess they think "oh, I could learn to teach that").

The of the remaining 50%, another half, while they may be above our stated minimums for GPA and MCAT) fall below the "floor" we set as our actual minimum. Again, I've seen this at other medical schools. So knock out another 25%. Of the remaining 25%, another half are lacking in the ECs we looks for (or have some other aspect that the Admissions dean, in the magic he uses to choose who to interview, weeds out). Now were down to what, some 12% of the original applicant pool? Of those, the Dean works his magic to choose about half of those to invite for interviews.

Again, using the faculty job seekers as a comparison, about another 25% we cull because even though they have anatomy experience (a PhD in Anatomy/Cell Bio), they lack another important qualifier. It could be no research, no teaching experience, too little experience...but it's enough to make up a short list of the remaining 25% of the applicant pool. The the Search Committee brings out the knives and we fight each other over who gets the interview!

So, my guess is that at Loma Linda, only some 600-1000 people are really competitive for those 168 seats. I'll bet that their median stats are close to what LL ends up with in their matriculates.

So don't let the number of applicants scare you away. You're just as qualified, if not more so, than most of them!

When prospective applicants ask for advice on where to apply, they're usually directed to compare their stats (MCAT & GPA) with the median scores that are accepted to, or matriculate at, any given med school. This is generally good advice. However, two schools with the same median stats can't be equally selective, because at some schools there are more applicants competing for fewer spots.

For example, using public data, and letting the number of matriculants serve as a mediocre relative approximation of the number of accepted applicants:

Rush and Loma Linda both have median MCATs of 31. At Rush, 7701 applicants competed last year for a total of 128 seats (1.66% matriculated). At Loma Linda, fewer applicants (5217) competed for more seats (168) (3.22% matriculated).

Looking at LizzyM profiles alone, Rush and Loma Linda would appear equally selective; however, from these numbers one could argue that an average applicant has double the chance of getting into Loma Linda compared to Rush. (This ignores that some schools are clearly a better "fit" for any given applicant than others, but I'm trying to just crunch numbers right now).

There are a handful of schools that are attractive to many because they have relatively low median stats but are also located in large cities. Notably, George Washington has a median MCAT of only 30, but it gets more applicants than literally any other school. Last year, 14509 applicants vied for 177 spots (1.22%).
Is George Washington therefore perhaps a worse choice for someone trying to play the numbers game than its neighbor Georgetown where, despite a higher average MCAT (32), only 11733 applicants compete for 196 seats (1.67%).

Is this a productive line of reasoning when assessing relative competitiveness and selectivity? Are confounds, notably the variable self-selectivity of schools, too influential to draw anything from the results?

 

[Bovary]

These are good responses, and I'm mostly convinced that it's not worthwhile to pay a lot of attention to how many applicants contend for each seat.
However:

Loma Linda is a religious school. Not many people would be a good fit for this school. So it isn't that Rush is that much more competitive; Loma Linda just doesn't get many applicants in general lol.

But discounting "fit", it seems like a contradiction to say Loma Linda doesn't get many applicants, but also to say that Rush isn't more competitive. All things being equal (which they aren't in real life, as has been discussed and I agree with), Rush would be more selective.

OP I think the number crunching is too heavily influenced by confounding factors (yield, limited pool of desired applicants, etc) to be significantly worthwhile. As noted above acceptance rate/yield can pretty dramatically swing your numbers.

I'd imagine that Georgetown and George Washington have similar yields, which is why they can be compared. I wouldn't compare George Washington with Mayo.

Christ. GTown and WashU are private schools without a strong in-state bias of any sort. Especially since GTown doesn't even have a state. If you're going to speak for SDN you should get your facts straight

Maybe that person confused WashU with UWashington, which loves in-state (or rather, in-region) applicants? And Brown takes a sizable portion of their class from their undergraduate pool. I don't know why Georgetown would have been mentioned though.

 

[gumdrops]

Sheesh, just seeing the title of this thread makes me want to tell the OP to take a chill pill.

 

[Aerus]

Christ. GTown and WashU are private schools without a strong in-state bias of any sort. Especially since GTown doesn't even have a state. If you're going to speak for SDN you should get your facts straight

Georgetown gets a huge number of applicants per year and many people on SDN do normally discourage applying there as a result of that. I never mentioned they had an in-state bias.

The latter was my mistake. I meant University of Washington, not George Washington in DC or Washington in St. Louis.