Validity of LM scores

[It’s Over!!!]

Which would you rather have: an LM of 79 or an LM of 80?

The answer should be easy right? Maybe you think , "Obviously the 80", or, "Probably not that huge a difference."

Nope! Sometimes both of these thoughts are very wrong.

The answer is, "It depends", which calls into question the value of LM at all as a metric to estimate relative competitiveness.

Take, for example, two people with very high MCAT scores. A 4.0 GPA with a 99th percentile MCAT of 522 would have a LM of 79 and an expected admit rate of 84%, while a 3.5 GPA and a perfect 528 MCAT (100th percentile) would be a LM of 80 with an expected admit rate of 64 percent.

So in this case:

LM 79--predicted admit rate of 84%
LM 80--predicted admit rate of 64%

So the absolute number seems to be of almost no value, at least for the highest MCAT score. It seems for this specific scenario the MCAT is weighted too high and GPA is weighted too low. I am sure this has to do with the problems of converting new MCAT scores to old ones, but clearly the LM formula needs some revision.

---

Members don't see this ad.

 

[lacrossegirl420]

Also the LM for some of the MCAT scores is inflated (eg a 525 -> 40, 523/524 -> 39, 522 -> 38, etc., right now 525 is put as a 43)

 

[Goro]

Oh for crying out loud! They're meant to be guidelines, not part of a table in Lancet.

 

[It’s Over!!!]

Oh for crying out loud! They're meant to be guidelines, not part of a table in Lancet.

A fair point. I was merely trying to point out something that seems likely fixable.

But I am sorry for having bothered you. I hope you are ok.

 

[Stoasty]

I think even LizzyM has said she's stopped using the score since the new MCAT so I wouldn't put all that much stock into it honestly.

 

[Plut0]

Take, for example, two people with very high MCAT scores. A 4.0 GPA with a 99th percentile MCAT of 522 would have a LM of 79 and an expected admit rate of 84%, while a 3.5 GPA and a perfect 528 MCAT (100th percentile) would be a LM of 80 with an expected admit rate of 64 percent.

So in this case:

LM 79--predicted admit rate of 84%
LM 80--predicted admit rate of 64%

No, it isn't.

Assuming your source is MCAT and GPA Grid for Applicants and Acceptees to U.S. MD-Granting Medical Schools, 64% is the acceptance rate for all applicants with GPA within 3.40-3.59 and MCAT>517. The acceptance rate for GPA within 3.40-3.59 and MCAT=528 is unknown but presumably much higher than 64%.

 

[Mr.Smile12]

I want to point out history. The LM was a guideline developed before the AMCAS GPA/MCAT/acceptance table. These are still considered guidelines to show that metrics are not completely deterministic of getting into medical school. Use all the data carefully in measuring your own chances.

LM is still a nice tool to discuss your GPA/MCAT with more privacy, but don't believe admissions teams use LM... or even know what LM means.

 

[LizzyM]

The LizzyM score was developed (by me) as a tool for applicants who wanted to compare their stats to the average for the successful applicants to a given medical school. It seemed to me at the time that there were too many people "applying broadly" and getting nowhere because their idea of broadly was the Ivies and the UCs with Hopkins, Duke and Stanford thrown in for good measure.

The idea was to have your own LizzyM score (GPA*10) +old MCAT and compare it to the average (GPA*10)+MCAT of the target school. The math was dead simple. With the new MCAT, the formula had to change and, frankly, I don't even remember it anymore. I seem to remember that it involved a bit more arithmetic.

Anyway, I do think that is is useful for those folks who want to talk about metrics without being too specific. For anything else, I like WARS as a better tool for determining where one should target one's applications.

 

[It’s Over!!!]

No, it isn't.

Assuming your source is MCAT and GPA Grid for Applicants and Acceptees to U.S. MD-Granting Medical Schools, 64% is the acceptance rate for all applicants with GPA within 3.40-3.59 and MCAT>517. The acceptance rate for GPA within 3.40-3.59 and MCAT=528 is unknown but presumably much higher than 64%.

That was not my source. SDN is. And I guess that is just my point. It seems very fixable. Here is the link. LizzyM Score Calculator - What are my medical school chances?

I merely was trying to point out that at least at the high end, there appears to be an error. I wasn't trying to ruffle any feathers.

 

[GoSpursGo]

That was not my source. SDN is. And I guess that is just my point. It seems very fixable. Here is the link. LizzyM Score Calculator - What are my medical school chances?

I merely was trying to point out that at least at the high end, there appears to be an error. I wasn't trying to ruffle any feathers.

Except it is not easily fixable, and if you think it's easy then I challenge you to propose a simple formula that accounts for all possible outliers. It is very common that "catch all" formulas like LM or WARS will break down for examples which are at the extreme end of the spectrum. In this case that means high MCAT/low GPA or vice/versa, but you can find similar statistical oddities like this in other fields where people try to boil down an extremely complicated process into a small number of digestible variables.

This question has been asked ad nauseum since at least 2007: CFO: Would you be interested in a spreadsheet of LizzyM scores? . Yes, everyone knows that LM and WARS and whatever else isn't "valid." It is intended as a very rough tool to give applicants a broad idea of their competitiveness for medical school. Assuming that MCAT and GPA are roughly matched, it may be useful for narrowing down a list of schools. For applicants with no frame of reference, this can be a very useful starting point. Of course if you have any suggestions for how to improve it, by all means do so

 

[It’s Over!!!]

Except it is not easily fixable, and if you think it's easy then I challenge you to propose a simple formula that accounts for all possible outliers. It is very common that "catch all" formulas like LM or WARS will break down for examples which are at the extreme end of the spectrum. In this case that means high MCAT/low GPA or vice/versa, but you can find similar statistical oddities like this in other fields where people try to boil down an extremely complicated process into a small number of digestible variables.

This question has been asked ad nauseum since at least 2007: CFO: Would you be interested in a spreadsheet of LizzyM scores? . Yes, everyone knows that LM and WARS and whatever else isn't "valid." It is intended as a very rough tool to give applicants a broad idea of their competitiveness for medical school. Assuming that MCAT and GPA are roughly matched, it may be useful for narrowing down a list of schools. For applicants with no frame of reference, this can be a very useful starting point. Of course if you have any suggestions for how to improve it, by all means do so

Except it is not easily fixable, and if you think it's easy then I challenge you to propose a simple formula that accounts for all possible outliers. It is very common that "catch all" formulas like LM or WARS will break down for examples which are at the extreme end of the spectrum. In this case that means high MCAT/low GPA or vice/versa, but you can find similar statistical oddities like this in other fields where people try to boil down an extremely complicated process into a small number of digestible variables.

This question has been asked ad nauseum since at least 2007: CFO: Would you be interested in a spreadsheet of LizzyM scores? . Yes, everyone knows that LM and WARS and whatever else isn't "valid." It is intended as a very rough tool to give applicants a broad idea of their competitiveness for medical school. Assuming that MCAT and GPA are roughly matched, it may be useful for narrowing down a list of schools. For applicants with no frame of reference, this can be a very useful starting point. Of course if you have any suggestions for how to improve it, by all means do so

Please, just forget I asked. I had no idea this was such a sensitive subject.

Have a nice weekend everyone.

 

[zakimran23]

Is there anything that can be used in present day in place of LizzyM? I've seen WARS, but even that seems outdated.

 

[Goro]

MSAR

 

[Angus Avagadro]

In addition, there are intangibles LM scores don't address when looking at medical school acceptance rates. Most notably from your posts, the lower GP, 3.5 in your example, and a 528 MCAT has a 64% acceptance rate. My Spidy Senses begin to tingle when I see a low GPA and a very high MCAT. I will give that app a little more scrutiny. Often it represents a bright but lazy student who does just enough to get by. Maybe they have personality issues? Not my idea of the doctor I want taking care of me and my family. LM scores can't address these issues.

 

[lacrossegirl420]

I just think it’s funny how a 3.6 has become a mark of “laziness” for med schools, when in stuff like business it’s considered something to be proud of and put on a resume lmao

 

[Angus Avagadro]

I just think it’s funny how a 3.6 has become a mark of “laziness” for med schools, when in stuff like business it’s considered something to be proud of and put on a resume lmao

Wow. Sorry I hit a nerve, I never said a 3.5 was a sign of laziness. If you read what I said, I said a 3.5, OP's number, and a 99th percentile MCAT, 528, often represents a bright but lazy student who does just enough to get by. You have to ask why a 99th%tile mcat is not aligned with a 99th%tile GPA? Often life issues, medical issues, or mental health issues can explain this. This is why I said these apps get some extra scrutiny from me. Certainly not all are lazy, but many are.

 

[Mr.Smile12]

I also want to point out that any GPA/MCAT formula that may have been used up to 10 years ago may likely have been modified to include additional non-cognitive characteristics. A couple of presentations I have listened to have discussed having an extra scoring factor that shows mission fit or adjusts to challenges overcome. As mentioned, even setting priorities for applicant review can be geographic, notably in-state and in-region sorts by zip code as one possible example. Mission fit is a big deal, as scored by your essays and interviews, which cannot be captured here.

 

[LunaOri]

Wow. Sorry I hit a nerve, I never said a 3.5 was a sign of laziness. If you read what I said, I said a 3.5, OP's number, and a 99th percentile MCAT, 528, often represents a bright but lazy student who does just enough to get by. You have to ask why a 99th%tile mcat is not aligned with a 99th%tile GPA? Often life issues, medical issues, or mental health issues can explain this. This is why I said these apps get some extra scrutiny from me. Certainly not all are lazy, but many are.

To me, it often correlates with someone who was not premed as an undergrad, then changes course and studies like heck for the MCAT (and usually does very well in postbac courses).

 

[Hzreio]

Wow. Sorry I hit a nerve, I never said a 3.5 was a sign of laziness. If you read what I said, I said a 3.5, OP's number, and a 99th percentile MCAT, 528, often represents a bright but lazy student who does just enough to get by. You have to ask why a 99th%tile mcat is not aligned with a 99th%tile GPA? Often life issues, medical issues, or mental health issues can explain this. This is why I said these apps get some extra scrutiny from me. Certainly not all are lazy, but many are.

Funny how 3.5 with a positive is interpreted as worse than 3.5 without the positive...

 

[Angus Avagadro]

Funny how 3.5 with a positive is interpreted as worse than 3.5 without the positive...

Actually, it's not funny. I would love to teach the student @LunaOri describes. They make the faculty's job much easier and represent our school well when they move on to residency. I have had several of the bright and lazy variety and they are not easy to teach, always scrambling at the end, and don't represent our school as well in residency. This is why their app gets a closer inspection from me. PD's aren't looking for residents committed to doing the bare minimum. Their co residents don't care for lazy classmates either.

 

[eigen]

Actually, it's not funny. I would love to teach the student @LunaOri describes. They make the faculty's job much easier and represent our school well when they move on to residency. I have had several of the bright and lazy variety and they are not easy to teach, always scrambling at the end, and don't represent our school as well in residency. This is why their app gets a closer inspection from me. PD's aren't looking for residents committed to doing the bare minimum. Their co residents don't care for lazy classmates either.

Sounds like a similar reason to why solid B/B+ students are often the ones who shine in my research lab.

 

[LizzyM]

Remember, LizzyM score was developed to help applicants identify schools where they were most likely to be inivited to interview (or unlikely to interview). I was trying to keep the LM 65's from applying to places where the combination of GPA(10)+MCAT was 78. Of course, more goes into interview invite decisions than metrics but it is a start and narrowing down one's list.

 

[zakimran23]

Wow. Sorry I hit a nerve, I never said a 3.5 was a sign of laziness. If you read what I said, I said a 3.5, OP's number, and a 99th percentile MCAT, 528, often represents a bright but lazy student who does just enough to get by. You have to ask why a 99th%tile mcat is not aligned with a 99th%tile GPA? Often life issues, medical issues, or mental health issues can explain this. This is why I said these apps get some extra scrutiny from me. Certainly not all are lazy, but many are.

A 3.5 is between a B+ and A- average? Idk if I would call someone who averages a 90 in their classes "bright but lazy who does just enough to get by." Just my uneducated 2 cents.

 

[Angus Avagadro]

A 3.5 is between a B+ and A- average? Idk if I would call someone who averages a 90 in their classes "bright but lazy who does just enough to get by." Just my uneducated 2 cents.

I see what you are saying. Said another way, what is a person doing with the cutoff GPA, usually 3.5 for most good schools doing with a 99%tile MCAT? Why doesn't that person have a 99%tile GPA? Lots of reasons, but bright and lazy is one of them. There are some scary smart people who don't have to work hard.

 

[LizzyM]

In response to the 3.5/99th percentile MCAT discussion. Lazy might be a harsh word. Sometimes it is more like brilliant but not good over the long haul with minor matters.

"Homework worth 10% of points? I'll blow it off 50% of the time, no big deal. Tardy to class and therefore missed a pop quiz worth 2% of the grade.... no big deal, I'm going to ace the final, just like I aced the midterm." Such a student might be taking the approach that the little day-to-day stuff doesn't matter as long as you rise to the occasion for big, high stakes tests.

We'd rather the student/applicant who manages both the little things and the big ones, too.

 

[MyOdyssey]

I just think it’s funny how a 3.6 has become a mark of “laziness” for med schools, when in stuff like business it’s considered something to be proud of and put on a resume lmao

Not for top flight MBA programs.

 

[eigen]

A 3.5 is between a B+ and A- average? Idk if I would call someone who averages a 90 in their classes "bright but lazy who does just enough to get by." Just my uneducated 2 cents.

The issue isn't the 3.5.

It's an "average" GPA for a medical school applicant, coupled with a well above average MCAT score.

Since the MCAT is a one-time exam vs. grades that often require sustained effort, when there's a definite imbalance between the two of them, it suggests that someone might be better at a burst effort than sustained effort.

 

[kath00]

I also want to point out that any GPA/MCAT formula that may have been used up to 10 years ago may likely have been modified to include additional non-cognitive characteristics. A couple of presentations I have listened to have discussed having an extra scoring factor that shows mission fit or adjusts to challenges overcome. As mentioned, even setting priorities for applicant review can be geographic, notably in-state and in-region sorts by zip code as one possible example. Mission fit is a big deal, as scored by your essays and interviews, which cannot be captured here.

I have to be honest though, I just don't understand mission fit. I feel like EVERY school has the same mission. So for a well rounded applicant who has done some research, some community service, working with underserved, they are literally a mission fit at every single school.... or none?! It's so hard to see how they would separate out students.

 

[Goro]

Some schools are very big into service, like Gtown, Drexel, Albany and Netter.

Others prefer to train the medical leaders of tomorrow, like the Ivies or JHU

 

[eigen]

I have to be honest though, I just don't understand mission fit. I feel like EVERY school has the same mission. So for a well rounded applicant who has done some research, some community service, working with underserved, they are literally a mission fit at every single school.... or none?! It's so hard to see how they would separate out students.

Some schools have fairly generic missions, but there are a lot with much more specific focuses.

 

[LizzyM]

Some schools are very big into service, like Gtown, Drexel, Albany and Netter.

Others prefer to train the medical leaders of tomorrow, like the Ivies or JHU

Agreed. Some schools (particularly some state sponsored schools) are charged with producing primary care providers for a specific geographic area or a specific population such as rural Americans. Someone who wants to be a academic subspecialist might not be a great fit for such a school.