New MCAT causing schools to be behind this year?

[barack_yomama]

Posted this on Reddit, curious to hear responses from people here as well. I'm wondering if admissions departments might be running behind this year, in part due to having to deal with the new MCAT. Does anyone know if there's any truth to this?

---

Members don't see this ad.

 

[Lawpy]

I don't know why there would be an MCAT-related delay when adcoms can readily use percentile charts for efficient comparisons and evaluations.

 

[Ismet]

No. The new MCAT scoring is not as confusing as you think. Percentiles.

 

[chocolatecheese]

No. The new MCAT scoring is not as confusing as you think. Percentiles.

Percentiles aren't as easy as you think. They're all ranges. An 11 VR is labelled as 95th percentile, but that doesn't mean it's definitively better than a 128 and a 129, since an 11 completely encompasses those ranges. How do adcoms take that into account?

 

[XxThaDoggxX]

Are schools really that far behind compared to last year? Someone aware me...Currently just riding the tide.

 

[takeurmeds02]

I doubt it. Some of my interviewers discussed my new MCAT score with me as if it has always been around.

 

[piii]

Percentiles aren't as easy as you think. They're all ranges. An 11 VR is labelled as 95th percentile, but that doesn't mean it's definitively better than a 128 and a 129, since an 11 completely encompasses those ranges. How do adcoms take that into account?

What? 95 is greater than both 93 and 87 so..

 

[Ismet]

Percentiles aren't as easy as you think. They're all ranges. An 11 VR is labelled as 95th percentile, but that doesn't mean it's definitively better than a 128 and a 129, since an 11 completely encompasses those ranges. How do adcoms take that into account?

Huh? I don't think you understand the scoring system. Each subsection score has a different percentile. On CARS, a 128 is 87th percentile and a 129 is 93rd. And also, we typically look at the overall score unless there's some crazy unbalance in the subsections, and we don't scrutinize between little points and definitely not between individual people like you suggest. The only time the percentiles meld together is at the extreme upper end, where it really doesn't matter if you're 98th or 99th or 100th because it's clear you knocked it out of the park.

Again, percentiles are very easy to use.

 

[Pantaloooons]

What? 95 is greater than both 93 and 87 so..

I think potential complications arise when comparing old vs new MCAT scores. Multiple composite scores on the new MCAT can correspond to the same reported percentile, although they probably differ by some fraction of a percentile. I'd be surprised if adcoms made any distinction between a 508 vs 509, etc., at least for this first cycle of incorporating new MCAT scores.

 

[Ismet]

I think potential complications arise when comparing old vs new MCAT scores. Multiple composite scores on the new MCAT can correspond to the same reported percentile, although they probably differ by some fraction of a percentile. I'd be surprised if adcoms made any distinction between a 508 vs 509, etc., at least for this first cycle of incorporating new MCAT scores.

A 508 is 77th and 509 is 80th. No, not a huge difference, but not the same and not a fraction different. The only time things start melting together is 519+ (98th %ile) and at that point, it doesn't really matter.

 

[DrHart]

From the chit chat I've seen it seems to me that schools just have more applicants this year which is why it may take more time. As others have said, with a conversion table it really doesn't take much longer.

 

[Lawpy]

A 508 is 77th and 509 is 80th. No, not a huge difference, but not the same and not a fraction different. The only time things start melting together is 519+ (98th %ile) and at that point, it doesn't really matter.

Analogously, as @gyngyn and others pointed out, anything above a 37+ doesn't matter, since the differences are statistically insignificant. Since a 519 corresponds to a 37, the percentile tables work well

And besides, subsection analysis is very straightforward. Anything below the median/4 (8 in the old, 125 in the new) is bad and should be retaken. Anything above the 75th percentile/4 (10 in the old, 127 in the new) is strong. The 4 is the number of subsections in the new test (so /3 for the old test). This method works well for the psych/soc section which doesn't have a counterpart in the old test.

 

[Pantaloooons]

A 508 is 77th and 509 is 80th. No, not a huge difference, but not the same and not a fraction different. The only time things start melting together is 519+ (98th %ile) and at that point, it doesn't really matter.

The 508 vs 509 example wasn't the best, but you're right that at the extremes, the differences become much less dramatic, percentile-wise. It makes me wonder why the AAMC stuck with the 15-point spread instead of increasing it and potentially spreading out the percentiles even farther. Although, I guess the 15-point spread's worked for a while and so there's no real need to change something that's not too broken to begin with.

 

[Ismet]

The 508 vs 509 example wasn't the best, but you're right that at the extremes, the differences become much less dramatic, percentile-wise. It makes me wonder why the AAMC stuck with the 15-point spread instead of increasing it and potentially spreading out the percentiles even farther. Although, I guess the 15-point spread's worked for a while and so there's no real need to change something that's not too broken to begin with.

Because it doesn't matter that much. I don't care if you're 88th or 93rd. I care even less if you're 93rd vs 91st.

I think a common misconception among pre-meds is that they think applicants are directly compared to each other, and they're not.

 

[Lawpy]

I think a common misconception among pre-meds is that they think applicants are directly compared to each other, and they're not.

Much of this misconception arises directly from secondary and interviewer questions asking:

"What makes you unique from the rest of the applicant pool?"

As well in the following rejection/waitlist scenarios:

"We have too many qualified applicants for a much smaller number of seats."

 

[Ismet]

Much of this misconception arises directly from secondary and interviewer questions asking:

"What makes you unique from the rest of the applicant pool?"

As well in the following rejection/waitlist scenarios:

"We have too many qualified applicants for a much smaller number of seats."

I hate the phrasing of that question and I've never used it in that way. There's just never a point where you think "Well Applicant A is really impressive, but Applicant B has this X factor so B is better than A." No, they're both strong in their own rights. They're evaluated on their own merits based on what we expect and value, not in comparison to each other.

The second statement is entirely true and valid, and I don't think it's possible to fully understand until you see the other side and realize the sheer number of excellent applicants there are. The vast majority of people we interview would fit and do well at the school. Only a very small fraction do poorly or don't impress in the interview stage. The fact remains that we have less than 150 seats and several hundred qualified applicants.

 

[chocolatecheese]

Huh? I don't think you understand the scoring system. Each subsection score has a different percentile. On CARS, a 128 is 87th percentile and a 129 is 93rd. And also, we typically look at the overall score unless there's some crazy unbalance in the subsections, and we don't scrutinize between little points and definitely not between individual people like you suggest. The only time the percentiles meld together is at the extreme upper end, where it really doesn't matter if you're 98th or 99th or 100th because it's clear you knocked it out of the park.

Again, percentiles are very easy to use.

Maybe I wasn't clear before. In CARS a 128 corresponds with an 81-87th percentile while a 129 encompasses an 87-93rd percentile. But for an 11 verbal reasoning the range is an 84-95th percentile. All of the scores have a range, since a certain percentage of people can achieve a given score (and the MCAT is scored on an ordinal scale). So given the larger overlap in these ranges, can we say that an 11 VR is definitively better than a 129 or even a 128? How do these schools compare? Most people (including you) seem to think that using the top of the range (i.e. the percentage of people who achieve the same or lower score) is the best, but is this necessarily the case?

Someone who has scored a 129 has technically scored better than more people than someone with an 11VR, but a someone with an 11 VR has scored greater than or equal to more people than someone with a 129 has.

 

[Ismet]

Maybe I wasn't clear before. In CARS a 128 corresponds with an 81-87th percentile while a 129 encompasses an 87-93rd percentile. But for an 11 verbal reasoning the range is an 84-95th percentile. All of the scores have a range, since a certain percentage of people can achieve a given score (and the MCAT is scored on an ordinal scale). So given the larger overlap in these ranges, can we say that an 11 VR is definitively better than a 129 or even a 128? How do these schools compare? Most people (including you) seem to think that using the top of the range (i.e. the percentage of people who achieve the same or lower score) is the best, but is this necessarily the case?

Someone who has scored a 129 has technically scored better than more people than someone with an 11VR, but a someone with an 11 VR has scored greater than or equal to more people than someone with a 129 has.

Again, people are not compared directly to each other. You're putting a lot of emphasis on a distinction that isn't highly emphasized in the great scheme of things. The percentiles are reported alongside the score so it takes 1 second to see if the score is around the target/average for the school.