Adcom analysis of MCAT scores

[Nietzschelover]

Seriously seeking information from anyone....and I vow that I will remain on my best behavior.

I'm curious about how adcoms analyze MCAT scores, and to what degree they actually consider the breakdowns, balance, low or comparatively lower CARS scores, and whether psych/soc at least initially will be weighted evenly. Do adcoms just look at the big number (total) and move along, or do they evaluate more closely such that one 515 (~33/34) is actually better or stronger than another 515? Will high scores on the other three cover for a low CARS? Or if the scores on the first three are equivalent with a 36/37 will some adcoms consider the overall score stronger than a 515 if the P/S is the score that brings the applicant down to a 515?

Examples are ordered CP, CARS, BB, P/S

Example #1 -- 131/127/129/128

Example #2 -- 127/126/131/131

Example #3 -- 129/129/128/129

Example #4 -- 130/124/129/132

Example #5 -- 130/129/130/126

Example #6 -- 131/130/129/125

---

Members don't see this ad.

 

[genericpremedstudent]

I thought you haven't even started undergrad yet?

 

[nwts]

I'm not an adcom, but I know of at least one school that states that they consider individual section scores below 125 to be noncompetitive, so that would make example #4 problematic for an applicant.

 

[Doudline]

Even subsections > uneven subsections (ceteris paribus).
The rest falls on the shoulders of adcoms' individuality and eccentricities, and won't help you whatsoever.

 

[Goro]

I don't think that schools will be weighed down with minutiae as to what your five hypothetical candidates have. ALL of the scores are competitive based upon %'s.

ALL schools have minimums below which an applicant will not be considered. For example, PCOM requires at least an 8 in each subsection of the old test...the schools are all trying to get a sense of what the new normal is, and so the gestalt is that they'll simply use %'s.

When I'm reviewing apps, I look at the total number, and then at the Bio section. VR/CARS will get the most slack from everyone. For most many MD schools you can see in MSAR that the bottom 10th %ile is a full point lower than either PS or Bio.

EDIT: I didn't see the 124 score in sample 4, and so that person would be in trouble.

 

[sovereign0]

Balanced scores will always trump a similar but unbalanced score. As for the weight of each section, this is something I am curious about too.

 

[sovereign0]

I don't think that schools will be weighed down with minutiae as to what your five hypothetical candidates have. ALL of the scores are competitive based upon %'s.

ALL schools have minimums below which an applicant will not be considered. For example, PCOM requires at least an 8 in each subsection of the old test...the schools are all trying to get a sense of what the new normal is, and so the gestalt is that they'll simply use %'s.

When I'm reviewing apps, I look at the total number, and then at the Bio section. VR/CARS will get the most slack from everyone. For most many MD schools you can see in MSAR that the bottom 10th %ile is a full point lower than either PS or Bio.

Why do you mention Bio in particular?

 

[Goro]

Medicine IS Biology. A low score would mean that one didn't learn anything!

Why do you mention Bio in particular?

 

[Lawpy]

Why do you mention Bio in particular?

Medicine IS Biology. A low score would mean that one didn't learn anything!

And it's typically easier to do well in biology than in other subsections. An old MCAT report from AAMC shows this.

As far as MCAT2015 score breakdown is concerned, it works exactly the same way as the old MCAT breakdown. Just ensure none of the subsection score is below a 9-equivalent (it's better to get a 10+-equivalent on all subsections), and you're good to go.

Using an efle Percentile Transform, you should aim for around a 127 in each subsection. This means Examples 1 and 3 are solid, 2 has CARS forgiveness, and 5 and 6 have some P/S leniency (being it a new section and all). 4 however is in trouble and may have to retake.

 

[sovereign0]

Medicine IS Biology. A low score would mean that one didn't learn anything!

Right, but is a high score in BS preferable to a high score in PS? Perhaps this is a pointless question because it is never so cut and dry, but would you argue that a 10 P / 10 V / 12 B is a stronger candidate than a 12 P / 10 V / 10 B?

 

[Pusheen]

interesting thread.
Even though low CARS gets slack, do high CARS scorers (130 and up) still get some brownie points? I believe I've seen here that high verbal correlates to something good. USMLE scores maybe? I may be totally wrong here, so feel free to correct.

 

[pageantry]

interesting thread.
Even though low CARS gets slack, do high CARS scorers (130 and up) still get some brownie points? I believe I've seen here that high verbal correlates to something good. USMLE scores maybe? I may be totally wrong here, so feel free to correct.

Ppl say it and then other people will be like, naw.
And then someone will break out the PDFs and we'll have a real party.

 

[Cyberdyne 101]

interesting thread.
Even though low CARS gets slack, do high CARS scorers (130 and up) still get some brownie points? I believe I've seen here that verbal is one of the strongest predictors of med school success. (I may be totally wrong here, so feel free to correct).

I've heard that for the old MCAT, Columbia would multiply the verbal score by 2 and use a 60 point scale to evaluate candidates. I could be totally off though...

@WedgeDawg, can you confirm this?

 

[Neuroplasticity]

When I'm reviewing apps, I look at the total number, and then at the Bio section.

I would assume that Goro's stance is not unique:
"Statistically significant predictors for step 1 and step 2 included science grade point average (SGPA), the biologic science (BS) section of the Medical College Admissions Test (MCAT), college selectivity, race, and age of the applicant."
http://link.springer.com/article/10.1007/s10459-007-9087-x

 

[ac62994]

I was just thinking about this and was about to post about it! Definitely interesting.

Add-On Question: Do AdComs/Schools see a general breakdown of how well/badly you did on each of the specific subjects in each section of the MCAT, e.g. Biochemistry, Orgo, etc.?

 

[Goro]

I consider both to be equal, and yes, I took Orgo and Physics.

A 32 is a 32.

Also remember that the rest of the package is also looked at!

Right, but is a high score in BS preferable to a high score in PS? Perhaps this is a pointless question because it is never so cut and dry, but would you argue that a 10 P / 10 V / 12 B is a stronger candidate than a 12 P / 10 V / 10 B?

 

[Goro]

No, just the subsections. We do see how our own students do in the disciplines like Anatomy on COMLEX though.

I was just thinking about this and was about to post about it! Definitely interesting.

Add-On Question: Do AdComs/Schools see a general breakdown of how well/badly you did on each of the specific subjects in each section of the MCAT, e.g. Biochemistry, Orgo, etc.?

 

[Lawpy]

interesting thread.
Even though low CARS gets slack, do high CARS scorers (130 and up) still get some brownie points? I believe I've seen here that high verbal correlates to something good. USMLE scores maybe? I may be totally wrong here, so feel free to correct.

Definitely not for the old MCAT because it's essentially a crapshoot to score higher than a 12. The MCAT2015 CARS supposedly balanced things out in the end, but it's too early to know. I'd still wager probably not since it appears anywhere above a 129 is still equivalent. Having a better ability to rapidly decipher abstract humanities passages may be pretty cool, but really not so praiseworthy in the end.

And of course the MCAT subsection scores have some correlation with Step 1, but people are exaggerating the correlation. As long as each subsection is a 10+/127+, you are more than capable of handling a medical school curriculum.

I would assume that Goro's stance is not unique:
"Statistically significant predictors for step 1 and step 2 included science grade point average (SGPA), the biologic science (BS) section of the Medical College Admissions Test (MCAT), college selectivity, race, and age of the applicant."
http://link.springer.com/article/10.1007/s10459-007-9087-x

I don't understand these studies. As long as someone is admitted to medical school, the factors needed to get there become irrelevant. So Step 1 success is attributed entirely to effective studying in preclinical years. Even then, Step 1 and MCAT test two different things and are held in entirely two different environments.

It's common sense really that students who do well academically in college are likely to do just as well in medical school. So, it's unnecessary to carry out Step 1 vs MCAT regression analyses.

 

[Pusheen]

Definitely not for the old MCAT because it's essentially a crapshoot to score higher than a 12. The MCAT2015 CARS supposedly balanced things out in the end, but it's too early to know. I'd still wager probably not since it appears anywhere above a 129 is still equivalent. Having a better ability to rapidly decipher abstract humanities passages may be pretty cool, but really not so praiseworthy in the end.

I disagree with your analysis of what verbal is and how it works, but this isn't the place for that.

It seems that the adcoms care about all this correlation stuff, they're really the only ones who have any reason to.

 

[Lawpy]

I disagree with your analysis of what verbal is and how it works, but this isn't the place for that.

It seems that the adcoms care about all this correlation stuff, they're really the only ones who have any reason to.

You can disagree if you want, but it's actually a common sentiment shared by even the top scorers. The correlation studies are largely irrelevant as long as the applicant does well (and not have any subsection land below the 10th percentile). From there, the process becomes self-selecting, as students with higher scores will choose the better schools etc.

All these correlation studies just restate what everyone knows: people who do well academically in college are likely to do well in medical school.

 

[Nietzschelover]

Thanks for all the responses.

Here's a more focused question, since there already has been a lot of discussion in general about the risk of a relatively low CARS.

Would straight 130s on the first three and a soft psych/soc score potentially be viewed as higher than the total score at first glance (130, 130, 130, 125). In other words, if someone is at a 36/37 equivalent on CP, CARS and BB are they potentially more competitive for schools with a median a couple of points higher than their total with all four that lands them at 33/34?

 

[Nietzschelover]

I thought you haven't even started undergrad yet?

You were generous. Actually a 53 year old 8th grader.

 

[Neuroplasticity]

And of course the MCAT subsection scores have some correlation with Step 1, but people are exaggerating the correlation. As long as each subsection is a 10+/127+, you are more than capable of handling a medical school curriculum.

I don't understand these studies. As long as someone is admitted to medical school, the factors needed to get there become irrelevant. So Step 1 success is attributed entirely to effective studying in preclinical years. Even then, Step 1 and MCAT test two different things and are held in entirely two different environments.

It's common sense really that students who do well academically in college are likely to do just as well in medical school. So, it's unnecessary to carry out Step 1 vs MCAT regression analyses.

I think you are committing a straw man fallacy, I was not arguing against "if your MCAT is above X you are able to pass medical school," I agree with this. And I was not arguing that the correlation study I referenced was saying "doing well in college correlates to doing well in medical school." This thread's topic is what section of the MCAT do adcoms care most about if any, so I was pointing out that the bio MCAT section has the highest predictive validity of all the sections.

 

[Nietzschelover]

So @Goro, sounds like you won't analyze too much beyond total score UNLESS something really stands out as a major anomaly with the rest of the sub-scores. So, in general, a 515 is a 515 is a 515? There just seems to be a bit of a different arithmetic dynamic with four sections versus three, as hitting three sections out of the park out of four might trump two out of three.

 

[cardzkp]

Perhaps the wrong place to bring it up, but how about comparing old MCAT to new MCAT? For example, a student who scores a 512 is around a 32 on the old exam according to percentiles, but say that student received a 126 on his P/S section. If his other scores were 130/128/128, his percentiles could technically equate to around a 33 if you don't include the MCAT2015 addition of P/S. So, then how do you compare a student with a 12/10/11 and a 130/128/128/126? I could be totally wrong though, just curious!

 

[excited_alumni]

Just an applicant here... But in my opinion people should stop worrying so much about how small point differences and how it might potentially affect your application. Study hard and do your best once you have your score that's it you can't change it. Focus your efforts on improving the rest of your app rather than pondering what an extra point on one section rather than another could've done for you.

 

[Goro]

Moderators, please sticky!!

Just an applicant here... But in my opinion people should stop worrying so much about how small point differences and how it might potentially affect your application. Study hard and do your best once you have your score that's it you can't change it. Focus your efforts on improving the rest of your app rather than pondering what an extra point on one section rather than another could've done for you.

 

[WedgeDawg]

I've heard that for the old MCAT, Columbia would multiply the verbal score by 2 and use a 60 point scale to evaluate candidates. I could be totally off though...

@WedgeDawg, can you confirm this?

Sorry! I have no idea if this is true or not

 

[Cyberdyne 101]

Sorry! I have no idea if this is true or not

Lol, it's cool. It's probably just another crazy pre-med myth

 

[Nietzschelover]

Just an applicant here... But in my opinion people should stop worrying so much about how small point differences and how it might potentially affect your application. Study hard and do your best once you have your score that's it you can't change it. Focus your efforts on improving the rest of your app rather than pondering what an extra point on one section rather than another could've done for you.

Of course. Many are overly neurotic and each person has to focus on what he or she can control.

That said, isn't SDN about gaining a better understanding of the admissions process, what adcoms value, and how adcoms go about differentiating applications? And isn't there value in applicants knowing how their scores will play in terms of selecting what mix of schools to apply to? The MCAT seems to be the single largest factor considered from what I have read here, and certainly the WAMCs focus heavily on whether someone is within a couple of points up or down from medians at various tiers. Especially those who have a total score that lands on one of the boundary areas around or between a couple of tiers might benefit from hearing how adcoms assess a person's MCAT. Is it more about the total score, or does how the sub-scores fall out make a difference? Does a 130/130/130/125 count differently in the overall picture for an applicant than a 128/127/128/132? As someone else noted, the first set of scores there compared to the old MCAT would seem to look significantly better than the second example.

Which leads to another question I'm sure has been discussed many times. How much time does an adcom spend on each app? If it's only 10 minutes total for the whole thing then obviously they won't be staring at the MCAT breakdown for more than a few seconds, but maybe if there is a more extensive discussion at a cmte meeting the nuances might matter more?

 

[mr.mkitty]

In regards to "importance" of each sub section, would it be fair to say that adcoms will rank them in this order (for 2015) : BIO>CARs>P/C>Psych. Or am I off.

 

[Goro]

Yes, it has. I'll spend 30 mins, minimum on each person's app.

The MCAT breakdown, old or new, is very easy to take in. I just write down the numbers on the eval forms we get. We pay far more attention to retakes and look for (and value) improved performance. Likewise, knowledge decay and worse performance is duly noted.

And no, we don't bother with "his Bio went down but the VR went up" discussions at all. It's either "his score went up", or "his score went down."

Which leads to another question I'm sure has been discussed many times. How much time does an adcom spend on each app? If it's only 10 minutes total for the whole thing then obviously they won't be staring at the MCAT breakdown for more than a few seconds, but maybe if there is a more extensive discussion at a cmte meeting the nuances might matter more?

We haven't had those discussions yet. Hell, we haven't even started meeting yet.

In regards to "importance" of each sub section, would it be fair to say that adcoms will rank them in this order (for 2015) : BIO>CARs>P/C>Psych. Or am I off.

 

[Doug Underhill]

Getting a 7 in Bio is always unacceptable. Getting a 7 in Verbal is bad but conditionally acceptable. You don't want either.

It's not just that mean ol' adcom members are demanding high scores for the High Score God. The material in the bio section is pretty close to the sort of multiple-choice material you'll see in med school and on the USMLE. If you're unable to cough up a measly 8 (which is like 60% correct IIRC), your ability to handle this type of question is suspect.

 

[nwts]

On this subject, there is speculation in another thread about an applicant with a cumulative score of 508, but his CARS score is 121. A 508 is between a 29 and a 30 on the old scale, and a 121 in CARS is between a 4 and a 5. There is a lot of speculation that even though his cumulative score is okay, his CARS score is just too low and will sink his application.