MCAT and GPA prediction of success in med school

[MedPR]

---

Members do not see this ad.

Doesn't matter how I sound, what matters is the evidence that shows MCAT is a terrible predictor of success. But thank you for the kind words anyway 🙂.

You really struggle to understand basic english don't you. Nobody is saying MCAT is a good predictor. Everyone is saying MCAT is a better predictor than GPA.

 

[Midifelder10]

Lot of people can get 3.7+, not so many can get 37. GPA is the most worthless piece of information that an ADCOM has access to.

 

[mehc012]

IF variance in GPA could somehow be accounted for based on difficulty of courses taken (it's impossible to do this) then I'm sure the numbers would show it to be better than the MCAT. Sadly, GPA's vary depending on way too many factors to account for. For instance, a 3.5 from Stanford is probably different from a 3.5 at Mississippi State University.

The MCAT is the only thing that all students have in common for comparison and that is why it is used.

That...you...you just reiterated exactly the point everyone is trying to make to you 😕? IF gpa took those factors into account, it would be a different story. But it just so happens that it doesn't. GPA is a non-standardized measure which strives to be objective. MCAT is a standardized measure which comes far closer to being so.

 

[AdrianVeidt]

That...you...you just reiterated exactly the point everyone is trying to make to you 😕? IF gpa took those factors into account, it would be a different story. But it just so happens that it doesn't. GPA is a non-standardized measure which strives to be objective. MCAT is a standardized measure which comes far closer to being so.

I don't think you fully understand what I was trying to get across in that post or maybe I didn't do the best job of explaining. I am saying statistically, GPA will not be a better standard because of it's huge variance from school to school. But that doesn't mean that on the individual level, someone's GPA cannot predict their success.

I can equate it to a simple example. Person A has a 3.9 GPA and scores a 25 on the MCAT. Person B has a 3.2 GPA and scores a 35 on the MCAT. Both Persons went to the same school and took the same courses from the same teachers in the same semesters. I believe it would be foolish to think Person B has a better chance of success in medical school than Person A.

You really struggle to understand basic english don't you. Nobody is saying MCAT is a good predictor. Everyone is saying MCAT is a better predictor than GPA.

This is where you don't seem to understand what I am saying and it's clearly creating anger inside of you. Refer to the quote directly above and get a squishy ball, they are really effective at releasing anger.

 

[MedPR]

I don't think you fully understand what I was trying to get across in that post or maybe I didn't do the best job of explaining. I am saying statistically, GPA will not be a better standard because of it's huge variance from school to school. But that doesn't mean that on the individual level, someone's GPA cannot predict their success.

I can equate it to a simple example. Person A has a 3.9 GPA and scores a 25 on the MCAT. Person B has a 3.2 GPA and scores a 35 on the MCAT. Both Persons went to the same school and took the same courses from the same teachers in the same semesters. I believe it would be foolish to think Person B has a better chance of success in medical school than Person A.



This is where you don't seem to understand what I am saying and it's clearly creating anger inside of you. Refer to the quote directly above and get a squishy ball, they are really effective at releasing anger.

Which quote directly above? The one about how a 3.9/25 has a better chance at success than a 3.2/35?

If you really believe that then you're too far gone to see the truth.

 

[Spinach Dip]

This is where you don't seem to understand what I am saying and it's clearly creating anger inside of you. Refer to the quote directly above and get a squishy ball, they are really effective at releasing anger.

Obvious troll is obvious.

 

[mcloaf]

Obvious troll is obvious.

I haven't really seen him post anything in pre-allo that isn't just devil's advocate trolling.

 

[AdrianVeidt]

Which quote directly above? The one about how a 3.9/25 has a better chance at success than a 3.2/35?

If you really believe that then you're too far gone to see the truth.

Yes, that one. Have you ever considered that you are on the other end of too far gone to even consider it?

Honestly, do YOU believe that a 3 hour test can predict success more than 4 years of prior grades? No statistics involved, just a pure opinion.

I haven't really seen him post anything in pre-allo that isn't just devil's advocate trolling.

I just don't agree with the majority. You call it devil's advocate, I call it a differing opinion. Groupthink sucks

 

[Spinach Dip]

Honestly, do YOU believe that a 3 hour test can predict success more than 4 years of prior grades? No statistics involved, just a pure opinion.

Yes.

/thread

 

[HughMyron]

Honestly, do YOU believe that a 3 hour test can predict success more than 4 years of prior grades? No statistics involved, just a pure opinion.

You can retake the MCAT twice if you're not satisfied with your score, btw.

 

[AdrianVeidt]

Yes.

/thread

I was asking MedPR, but still: If you really believe that then you're too far gone to see what I refer to as the truth.

You can retake the MCAT twice if you're not satisfied with your score, btw.

As long as you've got 270 bucks you can retake as many times as you like.

 

[yellowfish]

You can retake the MCAT twice if you're not satisfied with your score, btw.

 

[yellowfish]

Yes, that one. Have you ever considered that you are on the other end of too far gone to even consider it?

Honestly, do YOU believe that a 3 hour test can predict success more than 4 years of prior grades? No statistics involved, just a pure opinion.

Repeating the same question everyone has already addressed multiple times using sensationalist phrasing doesn't change anything.

 

[MedPR]

Yes, that one. Have you ever considered that you are on the other end of too far gone to even consider it?

Honestly, do YOU believe that a 3 hour test can predict success more than 4 years of prior grades? No statistics involved, just a pure opinion.

I've said it before, but I'll say it here again. Yes, the MCAT is a 3 hour (4 hour isn't it?) test. However, to be successful you can't just show up and be a good test taker. You need to have some combination of doing well in pre-reqs (bio, chem, etc) MCAT prep, and have critical thinking skills. To get a high GPA you just need to study a couple of nights before tests for 4 years, or cheat.

So yes, I predict that a 3.2/35 will be more successful than a 3.9/25. The 3.9/25 won't even get in, so I'm right automatically.

IMO you are discounting the difficulty of doing well on the MCAT. I believe about 80k people take the MCAT each year and a 35 is in the 95%ish. If you get a 35 that means only ~4000 test takers did as well or better than you. If you get a 25, about 40,000 test takers did as well or better than you.

There's simply no reason why such a high GPA would have such a low MCAT unless the high GPA isn't legit.

 

[Spinach Dip]

I was asking MedPR, but still: If you really believe that then you're too far gone to see what I refer to as the truth.

Trolls are funny.

Let me attempt to summarize what you just said: "If you disagree with me you're hopeless."

Is that about right?

 

[HughMyron]

I was asking MedPR, but still: If you really believe that then you're too far gone to see what I refer to as the truth.

All of us are "too far gone"? Really?

Look, the reason we doubt GPA as a (good) measure of success is because there are too many lurking variables there. For example:

Grade Inflation
Grade Deflation
Major Choice
Professor Choice
Family Situation
Employment Situation
Extracurricular Commitments
Athletic Commitments
Med School Interviews (I'm tanking all my classes cause I've lost weeks to interview travel)

And the list goes on and on.

For the MCAT, it's a test you prepare for on your own time, for as long as you want. If you do bad on it, you take it again. If you do bad again, you take it a third time. It's all up to you.

 

[bollywoodlover]

I think a major reason The MCAT exists is because there simply aren't enough med spots to go around so they need a way to narrow down the pool. I have no doubt in my mind that many with a 24+ balanced score will succeed in med school and do as well as those with a 30 and sometimes do better. however, there aren't enough spots to take everyone who can succeed. I know someone who got a 26 on the MCAT one week and got a a 31 2 weeks later. luck and mentality can often increase and decrease your score drastically. also a lot of people take the MCAT two or three times before they get a 30ish but get like in the mid/high 20s the first time. Who says the person who got accepted with a mid/high 20 would not have done better if they retook the test? Most likely they would have. In my mind they have just as good of chance to succeed as the person who finally scored in the 30s after the second or third attempt because they achieved the same first attempt score.

 

[AdrianVeidt]

Trolls are funny.

Let me summarize what you just said: "If you disagree with me you're hopeless."

Is that about right?

Actually I was using MedPR's exact words against him but you had to reply first and ruin my fun. That's what he was doing.

I've said it before, but I'll say it here again. Yes, the MCAT is a 3 hour (4 hour isn't it?) test. However, to be successful you can't just show up and be a good test taker. You need to have some combination of doing well in pre-reqs (bio, chem, etc) MCAT prep, and have critical thinking skills. To get a high GPA you just need to study a couple of nights before tests for 4 years, or cheat.

IMO you are discounting the difficulty of doing well on the MCAT. I believe about 80k people take the MCAT each year and a 35 is in the 95%ish. If you get a 35 that means only ~4000 test takers did as well or better than you. If you get a 25, about 40,000 test takers did as well or better than you.

3 hours 20 minutes if you don't take any of the breaks.

You make some very good points. I'll meet you close to the middle and agree that effort is a good predictor of success.

 

[MedPR]

Actually I was using MedPR's exact words against him but you had to reply first and ruin my fun. That's what he was doing.

3 hours 20 minutes if you don't take any of the breaks.

You make some very good points. I'll meet you close to the middle and agree that effort is a good predictor of success.

Fair enough 👍

 

[mcloaf]

[I'll] agree that effort is a good predictor of success.

This definitely seems like an insightful conclusion worth three pages of pointless back and forth.

 

[AdrianVeidt]

This definitely seems like an insightful conclusion worth three pages of pointless back and forth.

You should see how long it takes my girlfriend to convince me the trash is worth taking out.

 

[UrbanRiver]

Thanks for the good discussion, guys. IMHO, MCAT is a better indicator for Step 1, not "success," which I see as a subjective term defined very differently by different people. In addition to strong critical reasoning skills, success requires good social skills, leadership, and endurance. I would think these latter attributes would matter a lot during clinicals, residency, practice, and good patient care. MCAT, I completely agree, is an irreplaceable part of the application process. That's why I'm willing to give a 3-hour test at least an equal share (if not more) than gpa, personal essay, extracurriculars COMBINED. Like you all said, we need some standard to compare applicants in critical reasoning skills. I know I may sound like a masochist, but I actually liked studying for the MCAT. I hated it, but I loved it haha (nerd i know, but aren't we all?). I liked being able to take this test and see where I stand among geniuses across the U.S. If athletes have tournaments and compare their PR's with one another's, why can't we? If anything, the MCAT was good review and a great challenge, and my effort and results gave me further confirmation that I wanted to go into medical school. glad to be saying this in hindsight of med school applications.. shoot. wishing you all success!

 

[PreMedOrDead]

There is a .7 correlation between MCAT and Step I, yet people are challenging its validity in predicting medical school performance? The Step I is absolutely the most important part of a residency application, and a .7 correlation in a humanities study is excellent. What is there to argue? The number of questions, length of test, those things don't matter at all. It could be a 1 question test given in 60 seconds, and if it had a .7 correlation to medical school success, I'd say use it over anything else. The arguments here make it clear people don't understand statistics, but the nice thing about statistics is that it doesn't care how you feel about it. It's by far the best predictor we have for medical school success, so saying it's not doing its job is baseless. Now challenging the study itself would be more apt.

By no means am I saying it can't be improved, but the arguing is ridiculous.

 

[SOMBound13]

Lot of people can get 3.7+, not so many can get 37. GPA is the most worthless piece of information that an ADCOM has access to.

But, does getting a 37 MCAT mean that you'll perform better in med school and/or perform better as a physician?

 

[mimelim]

But, does getting a 37 MCAT mean that you'll perform better in med school and/or perform better as a physician?

Sample size is small, but I've never heard of someone with a 37 MCAT struggling in medical school. People with a GPA of 3.7+ struggling are a dime a dozen. Obviously a lot fewer MCAT 37+ than GPA 3.7+, but still... on the high end, as someone else posted, MCAT scores do correlate with Step 1 scores, which while not a mark of being a good physician does help significantly with landing competitive residencies.

 

[PreMedOrDead]

But, does getting a 37 MCAT mean that you'll perform better in med school and/or perform better as a physician?

Performance? Yes. There are plenty of studies showing this.
Practicing Physician? Too subjective to determine. That's a loaded question.

To determine one's ability to be a physician is far too complex for strict analysis for admissions decisions. Interviews and activities/LoRs/PSs are key components of determining this, but are subjective. MCAT is the best objective measure we have, period.

 

[danny55]

In my opinion the MCAT is a great weed-out test. The science sections focus on stuff that you should have learned your first 2 years of undergrad so if you have a 3.9+ GPA and got lower than 10 on the science sections my question is why? If you understood the material good enough to get an A+ in the course why did you all of a sudden forget it during the MCAT? I can understand if someone has a bad day, but then just re-write it again. MCAT tests how well you actually recall the material you should have been learning during undergrad and how you can APPLY that knowledge in different situations. For people who can't seem to do well on the MCAT but have stellar GPAs it indicates that something is wrong with you understanding of science or you are just a bad test-taker all the time. (Sorry if this sounds harsh, it is just my opinion).

 

[SixStringPsych]

In my opinion the MCAT is a great weed-out test. The science sections focus on stuff that you should have learned your first 2 years of undergrad so if you have a 3.9+ GPA and got lower than 10 on the science sections my question is why? If you understood the material good enough to get an A+ in the course why did you all of a sudden forget it during the MCAT? I can understand if someone has a bad day, but then just re-write it again. MCAT tests how well you actually recall the material you should have been learning during undergrad and how you can APPLY that knowledge in different situations. For people who can't seem to do well on the MCAT but have stellar GPAs it indicates that something is wrong with you understanding of science or you are just a bad test-taker all the time. (Sorry if this sounds harsh, it is just my opinion).

+1

The MCAT evaluates a certain level of intellectual ability and dexterity that college courses cannot, period. Why do you think you can't use a calculator for PS? Because there is no plug and chug. It's all logic based on the principles of Physics/Chem.

 

[def1]

Performance? Yes. There are plenty of studies showing this.
Practicing Physician? Too subjective to determine. That's a loaded question.

To determine one's ability to be a physician is far too complex for strict analysis for admissions decisions. Interviews and activities/LoRs/PSs are key components of determining this, but are subjective. MCAT is the best objective measure we have, period.

While I agree the MCAT is a decent indicator of med school success, there seems to be a point where the amount of success plateaus and this is before getting into really high MCAT scores.

According to this, https://www.aamc.org/students/download/267622/data/mcatstudentselectionguide.pdf

A student with a 27-29 MCAT has pretty much the exact same chance of graduating med school as someone with a 39-45 MCAT.

The same is true for the statistics involving withdrawal or dismissal from med school with a difference of about 1-2%. 1st time pass rates of Step 1 are roughly the same as well.

Even scores lower than 27 fair pretty well with medical school success.

From a rather brief look at the data, it seems after a MCAT score of 27 the amount of additional success from a higher score becomes rather insignificant. Yet, top schools go after top scores. I think this has less to do with "predicted success" and more to do with a school wanting to show off high scores to seem more elite.

To be fair, the data does not go into specifics about scores and just shows passing rates, and its probably fair to assume scoring a 45 on the MCAT means you are likely to score better on Step 1 than someone who scored a 27 on the MCAT, but I don't think it will be that statistically significant.

 

[nabilesmail]

Just curious, Though I agree about the MCAT being the better correlator to step 1, doesn't a strong GPA show good work ethic -->> being able to study longer hours and more ahead of time--> med school is volume not conceptual, therefore GPA might be better?

For example, I had a decent GPA applying (3.74), and all my B's have been due to things like not turning in HW, or forgetting to show up for a test, studying the night before, forgetting to turn in a lab report etc. I'm someone that starts LEARNING 2-3 days before a test, s O-chem would test 3-4 chapters and I would start reading the book 3 days before (never attended lectures), read the chapters by sunday, and study sunday until Monday's test.

In Medical school, however, I feel that the amount of volume of material will be entirely too high to do this method. So wouldn't the 4.0 student that studies daily be in a better place to succeed in med school than me (assuming I don't step anything up?) 4.0 30 > 3.7 34

 

[PreMedOrDead]

To be fair, the data does not go into specifics about scores and just shows passing rates, and its probably fair to assume scoring a 45 on the MCAT means you are likely to score better on Step 1 than someone who scored a 27 on the MCAT, but I don't think it will be that statistically significant.

There's a R^2 = 0.7 correlation between MCAT and Step I. That's very much statistically significant. But you're also right in that people scoring a 28 or higher will do just fine passing the boards and getting through medical school. The reason that schools select for higher scores is because this is a competitive process in which schools pick 'the best of the best.' Why pick a 28 when you could pick a 36? The 36 is going to increase your school's average matriculant stats, likely have an easier time with material, score higher on the Step, and place into a more desired residency. Of course, this isn't always the case, but the odds are better. It's pretty obvious, then, why schools select toward higher numbers.

Just curious, Though I agree about the MCAT being the better correlator to step 1, doesn't a strong GPA show good work ethic -->> being able to study longer hours and more ahead of time--> med school is volume not conceptual, therefore GPA might be better?

For example, I had a decent GPA applying (3.74), and all my B's have been due to things like not turning in HW, or forgetting to show up for a test, studying the night before, forgetting to turn in a lab report etc. I'm someone that starts LEARNING 2-3 days before a test, s O-chem would test 3-4 chapters and I would start reading the book 3 days before (never attended lectures), read the chapters by sunday, and study sunday until Monday's test.

In Medical school, however, I feel that the amount of volume of material will be entirely too high to do this method. So wouldn't the 4.0 student that studies daily be in a better place to succeed in med school than me (assuming I don't step anything up?) 4.0 30 > 3.7 34

Higher GPA can be oversimplified to better work ethic, but that's not always the case. Certainly, there is a flood of information in medical school, but a higher MCAT is likely to indicate a student will have a much easier time understanding and internalizing that material. The 4.0/30 vs. 3.6/34 (changed so their LizzyMs were the same) comparison is likely going to come down to a factor other than their marks, as these applicants will both be successful in medical school, though they will likely excel in different ways. Though, if you think about it, the potential of the better MCAT student may be higher, assuming their GPA was from a lack of work ethic due to something other than simply being lazy (for example, having to take some irrelevant general ed requirements they hated and got Cs in vs played video games all day and night instead of studying).

 

[def1]

In my opinion the MCAT is a great weed-out test. The science sections focus on stuff that you should have learned your first 2 years of undergrad so if you have a 3.9+ GPA and got lower than 10 on the science sections my question is why? If you understood the material good enough to get an A+ in the course why did you all of a sudden forget it during the MCAT? I can understand if someone has a bad day, but then just re-write it again. MCAT tests how well you actually recall the material you should have been learning during undergrad and how you can APPLY that knowledge in different situations. For people who can't seem to do well on the MCAT but have stellar GPAs it indicates that something is wrong with you understanding of science or you are just a bad test-taker all the time. (Sorry if this sounds harsh, it is just my opinion).

I don't think this is exactly true. There is a lot of advanced bio/physiology stuff on the MCAT that you might not learn if you take only the required pre-reqs.

There are also non-trads who are out of school for long periods of time before taking the MCAT.

I think there are too many variables to simply say someone who doesn't do excellent on the MCAT lacks thorough understanding of science.

 

[MedPR]

Just curious, Though I agree about the MCAT being the better correlator to step 1, doesn't a strong GPA show good work ethic -->> being able to study longer hours and more ahead of time--> med school is volume not conceptual, therefore GPA might be better?

For example, I had a decent GPA applying (3.74), and all my B's have been due to things like not turning in HW, or forgetting to show up for a test, studying the night before, forgetting to turn in a lab report etc. I'm someone that starts LEARNING 2-3 days before a test, s O-chem would test 3-4 chapters and I would start reading the book 3 days before (never attended lectures), read the chapters by sunday, and study sunday until Monday's test.

In Medical school, however, I feel that the amount of volume of material will be entirely too high to do this method. So wouldn't the 4.0 student that studies daily be in a better place to succeed in med school than me (assuming I don't step anything up?) 4.0 30 > 3.7 34

Go away with your humblebrags.

GPA means very little because there are too many variables. You (people in general) could have a high GPA because you went to an easy school, or because you had an easy major, or because you took only the easy professors, or because you cheated.

None of those things apply to the MCAT.

 

[eDITN]

Just curious, Though I agree about the MCAT being the better correlator to step 1, doesn't a strong GPA show good work ethic -->> being able to study longer hours and more ahead of time--> med school is volume not conceptual, therefore GPA might be better?

For example, I had a decent GPA applying (3.74), and all my B's have been due to things like not turning in HW, or forgetting to show up for a test, studying the night before, forgetting to turn in a lab report etc. I'm someone that starts LEARNING 2-3 days before a test, s O-chem would test 3-4 chapters and I would start reading the book 3 days before (never attended lectures), read the chapters by sunday, and study sunday until Monday's test.

In Medical school, however, I feel that the amount of volume of material will be entirely too high to do this method. So wouldn't the 4.0 student that studies daily be in a better place to succeed in med school than me (assuming I don't step anything up?) 4.0 30 > 3.7 34

Maybe but not every single 4.0 is created equal. If every premed in the entire country took the same exact courses from the same exact professors at the same exact time and then we took all the 4.0s out there, then it might correlate more closely to step 1 scores. But, that is obviously not the case for a plethora of reasons that people have discussed in this thread. Who knows how hard one person who got a 4.0 had to work in comparison to another person at a different school with a different set of circumstances.

The MCAT is the universalizer (albeit with flaws). Premeds come from hundreds of different colleges throughout the world that all have differing levels of difficulty. Heck, even at the same school you can take the same course and have a differing level of difficulty. There are too many factors to know exactly how that 3.9 GPA shapes up against another applicant with a 3.9.

Is it flawed? Sure. The MCAT can't test every single concept you learned in your prereqs and so there is a level of luck to the test. But, it has a 0.7 correlation which makes it about as good of an indicator as we have.

 

[PreMedOrDead]

Go away with your humblebrags.

GPA means very little because there are too many variables. You (people in general) could have a high GPA because you went to an easy school, or because you had an easy major, or because you took only the easy professors, or because you cheated.

None of those things apply to the MCAT.

I went to an easy school, got an easy major, took the easy professors, and cheated off my friend named Saul. Suck on my 4.0, brah.

 

[MedPR]

I went to an easy school, got an easy major, took the easy professors, and cheated off my friend named Saul. Suck on my 4.0, brah.

I did nothing. Got into med school with <3.00. Suck my URM, brah.

 

[SixStringPsych]

I don't think this is exactly true. There is a lot of advanced bio/physiology stuff on the MCAT that you might not learn if you take only the required pre-reqs.

There are also non-trads who are out of school for long periods of time before taking the MCAT.

I think there are too many variables to simply say someone who doesn't do excellent on the MCAT lacks thorough understanding of science.

I agree about the non-trads, that's true.

But there is nothing on the MCAT that can't be learned in a Bio 101/102 sequence in college. They can't give advantage to bio majors otherwise it would be unfair to Humanities majors. Being a bio major might help you more easily and quickly get through the material, but that's about it.

 

[SixStringPsych]

I did nothing. Got into med school with <3.00. Suck my URM, brah.

 

[def1]

There's a R^2 = 0.7 correlation between MCAT and Step I. That's very much statistically significant. But you're also right in that people scoring a 28 or higher will do just fine passing the boards and getting through medical school. The reason that schools select for higher scores is because this is a competitive process in which schools pick 'the best of the best.' Why pick a 28 when you could pick a 36? The 36 is going to increase your school's average matriculant stats, likely have an easier time with material, score higher on the Step, and place into a more desired residency. Of course, this isn't always the case, but the odds are better. It's pretty obvious, then, why schools select toward higher numbers.

Where are you getting that r^2 from? I did a quick search and found a meta-analysis that shows r to be .6, which gives an r^2 of 0.36. Thats if I remember my statistics correctly.

I don't think thats a strong correlation, but its at least a moderate correlation. Definitely not something I think one needs to concern themselves with.

 

[MedPR]

I don't think this is exactly true. There is a lot of advanced bio/physiology stuff on the MCAT that you might not learn if you take only the required pre-reqs.

There are also non-trads who are out of school for long periods of time before taking the MCAT.

I think there are too many variables to simply say someone who doesn't do excellent on the MCAT lacks thorough understanding of science.

Idk what you learned in bio1/2, but everything on the MCAT was covered in mine.

Even then, there is no "advanced" bio/physio on the MCAT. It's all very basic even by undergrad bio/physio standards.

 

[def1]

I agree about the non-trads, that's true.

But there is nothing on the MCAT that can't be learned in a Bio 101/102 sequence in college. They can't give advantage to bio majors otherwise it would be unfair to Humanities majors. Being a bio major might help you more easily and quickly get through the material, but that's about it.

I think this depends greatly on individual schools. My Bio 101/102 classes were a joke and mostly dealt with organisms and plants. They did not talk about the physiology of the human organs, and they certainly didn't get into some of the more advanced cell biology questions. Obviously if you have a decent base in intro bio classes you can learn everything through MCAT review books, but someone who took advanced physiology, cell bio, genetics, virology/bacteriology classes will definitely be slightly in a better position

edit* I should mention I thought the Bio part was the easiest (besides verbal) section. But I'm just saying seeing how my intro bio classes were so abysmal I can see how someone only taking the pre reqs could struggle with it.

 

[IslandStyle808]

There's a R^2 = 0.7 correlation between MCAT and Step I. That's very much statistically significant. But you're also right in that people scoring a 28 or higher will do just fine passing the boards and getting through medical school. The reason that schools select for higher scores is because this is a competitive process in which schools pick 'the best of the best.' Why pick a 28 when you could pick a 36? The 36 is going to increase your school's average matriculant stats, likely have an easier time with material, score higher on the Step, and place into a more desired residency. Of course, this isn't always the case, but the odds are better. It's pretty obvious, then, why schools select toward higher numbers.

There is a difference between correlation and it being statistically significant. A high correlation means that one variable affects the other very strongly and vica versa for a low correlation. A statistically significant result means that your confidence interval does not have much variance or your p value is below a certain value.

For example, you could have r=0.4 and have a CI 95% between 0.35 and 0.45.
1) Does the result have a high correlation? No (closer to zero)
2) Is the result statistically significant? Yes (very small variance)

EDIT: That is actually r= 0.7 not r^2= 0.7

There is a .7 correlation between MCAT and Step I, yet people are challenging its validity in predicting medical school performance? The Step I is absolutely the most important part of a residency application, and a .7 correlation in a humanities study is excellent. What is there to argue? The number of questions, length of test, those things don't matter at all. It could be a 1 question test given in 60 seconds, and if it had a .7 correlation to medical school success, I'd say use it over anything else. The arguments here make it clear people don't understand statistics, but the nice thing about statistics is that it doesn't care how you feel about it. It's by far the best predictor we have for medical school success, so saying it's not doing its job is baseless. Now challenging the study itself would be more apt.

By no means am I saying it can't be improved, but the arguing is ridiculous.

This is what set off my trigger.

Where are you getting that r^2 from? I did a quick search and found a meta-analysis that shows r to be .6, which gives an r^2 of 0.36. Thats if I remember my statistics correctly.

I don't think thats a strong correlation, but its at least a moderate correlation. Definitely not something I think one needs to concern themselves with.

👍

 

[PreMedOrDead]

EDIT: That is actually r= 0.7 not r^2= 0.7

Ugh, and that's why I usually read articles myself. Another user cited R^2, not R. That makes the .4 for GPA even more of a joke, too. But your trigger was well-deserved, that's my bad. Either way, the MCAT still stands as the best predictor of performance.

 

[CaribbeanBlue]

Why do people even care about how correlated MCAT or GPA is to success in med school?

I feel like it has more to do with the individual.

People that work hard and are good at managing their time usually succeed with whatever is thrown at them as long as it INTERESTS them and it is something they really want to do.

If you lucked your way through the MCAT (Yes it happens every day it is administered at every site) or lucked your way through college and you aren't that smart, you probably may not do very well in medical school.

If you studied your a** off to pull off a 3.5+ GPA or 30+ MCAT, you will most likely succeed in medical school.

If you had to modify your study habits per class, succeeded, and are more than willing to make sacrifices to receive the grades, you will probably do well in medical school.

From what I have heard from medical school], it is mostly a bunch of memorization and the "reasoning" of the USMLE/COMLEX is most like the BS of the MCAT. The reason people don't really agree with the VR being a good correlation is because some people can be very good at critically reading science articles, but not so good at stock market/history articles.

 

[IslandStyle808]

Ugh, and that's why I usually read articles myself. Another user cited R^2, not R. That makes the .4 for GPA even more of a joke, too. But your trigger was well-deserved, that's my bad. Either way, the MCAT still stands as the best predictor of performance.

No worries. There are times it is good to post about something even if it is wrong. It helps one to understand if one's knowledge is correct or not. I have been shut down various times on SDN but thanks to those people I know my stuff better.

Edit: r=0.49 is not bad of a correlation between GPA and USMLE step 1 but is still lower than MCAT

 

[PreMedOrDead]

No worries. There are times it is good to post about something even it is wrong. It helps one to understand if one's knowledge is correct or not. I have been shut down various times on SDN but thanks to those people I know my stuff better.

Edit: r=0.49 is not bad of a correlation between GPA and USMLE step 1 but is still lower than MCAT

I'm like any other premed, I don't like spreading misinformation. 😉

But it depends on the study it seems. Most show the correlation for the BS section to be between .6 and .75. GPA is usually around .3 to .4. Someone would have to check but I think PS was about ,5 and VR was lower than .3.



And people care because this is how medical school decide how much weight to put on GPA and MCAT. If you can't tell, looking at the trend, MCAT averages at every school has been increasing dramatically in the 2000s, when the first of these studies got press, IIRC.

 

[IslandStyle808]

I'm like any other premed, I don't like spreading misinformation. 😉

But it depends on the study it seems. Most show the correlation for the BS section to be between .6 and .75. GPA is usually around .3 to .4. Someone would have to check but I think PS was about ,5 and VR was lower than .3.



And people care because this is how medical school decide how much weight to put on GPA and MCAT. If you can't tell, looking at the trend, MCAT averages at every school has been increasing dramatically in the 2000s, when the first of these studies got press, IIRC.

I agree about the misinformation and the main thing is just to research it before hand and then post. There are times when the information is wrong or the interpretation is wrong (we ourselves may not know). People will chime in with their views and this helps lead to what is true or false.

I have an article that I posted a little early that has more information than the first study.

http://internationalgme.org/Resources/Pubs/Donnon%20et%20al%20(2007)%20Acad%20Med.pdf

The r for the MCAT and USMLE correlation here is 0.6 but the studies done on the subject vary.

 

[PreMedOrDead]

Interesting article. My favorite parts are about the writing sample having literally no correlation with medical school success. Even the mean and confidence interval go negative, pretty significantly so even.
However, interesting that while the BS and PS sections both showed nearly the same range of r values for Step I, with step II and step III included they suggested med schools only look at verbal and biological sciences scores. Maybe I'm interpreting it wrong but with step I being of paramount importance and step II and III just 'there' I'm surprised at that proposition.

 

[Narmerguy]

I'm also going to add in that correlation between USMLE and MCAT score isn't necessarily the most important indicator. Sure we want to make sure our students are passing USMLE, but there's no evidence to show that high STEP scores lead to better doctors. Just because residency programs use it to select students doesn't mean medical schools need to.

Now the prestige/ranking interests may be such that medical schools want to select students that will go on to prestigious residencies and high board scores help that. But I would say that our goal shouldn't be maximizing USMLE scores among matriculants any more than our goal is to maximize MCAT scores among matriculants. Thus, a USMLE threshold measure makes the most sense.