Calculating Chance of Getting into Med School from LizzyM Score

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
RapplixGmed

RapplixGmed

Looking for the Ether
15+ Year Member
Joined
Nov 27, 2007
Messages
710
Reaction score
17
Hey everybody! While trying to decide where to apply, I had a little too much fun with the LizzyM scores and came up with a way to guestimate your chances of being accepted at any specific school given your stats and residency status. I've had a lot of fun using it to create a list of schools that are very well regarded but I also believe I have a decent chance of getting into. I thought that I'd share my method with you guys on SDN to see what you think and to let others use it too:

The file is here! (Changed hosts because SDN killed the old file) =(

Its kind of technical but not too difficult to understand. Basically, I just took the difference between a school's and my LizzyM score, divided by the standard deviation of LizzyM scores and used that to create a multiplier for the average acceptance rate for the school. This new number is your probability of being accepted into the school with your numbers given that you are otherwise the average matriculating student at that school.

Before I get flamed, obviously there is much more to an application than just stats but these numbers are the only "objective" things we have to go on right now. I thought it was pretty cool and thought others could benefit as well. Time to sleep!
Members don't see this ad.
 
Members don't see this ad :)
You love statistics, dont you?? lol reading that makes my head hurt.
 
sounds interesting. i wasn't able to open the attachment though.
 
Hmmm when I click on the attachment it just tells me "Invalid attachment specified". Anyone else getting that?
 
Pretty amazing stuff there. If I were just now applying then I think I would use that.

But then again, even if you only have a 9% chance of getting in someplace (like Duke for you), won't you still apply just to see if you can go there? People can tell where they are and are not competitive without all the algorithms--it is just won't give them a numerical value for it.

Moral of the story, OP was <50% chance to get in to all schools and based off his L' value his stats are likely ~4.0/40 with a decrease in one area leading to increase in another. APPLY BROADLY NO MATTER WHO YOU ARE
 
SnoPearl said:
Hmmm when I click on the attachment it just tells me "Invalid attachment specified". Anyone else getting that?
Click to expand...

Same here. It worked when I looked at it 20 minutes ago though. :confused:
 
When you are calculating the LizzyM of the schools, wouldnt it make more sense to use the LizzyM of Applicants rather than matriculants? Bc obviously a person with stats close or above the schools average matriculants will have a much greater shot then someone with stats = the average applicants. Am I missing something here?
 
Members don't see this ad :)
justme456 said:
what is LizzyM??
Click to expand...

She is apparently one of the administrator from a Medical School where she invented a score that distinguishes who is qualify to receive the secondary and/or an interview vs. who isn't via using simple Statistical Calculations. The process makes it fairly easy for the school if there are a large batch of applicants trying to matriculate 50-100 spots. I could be wrong though! :/
 
Yup going back and looking at it, I think instead of using the average accepted students stats to calculate the schools LizzyM it would be a much better approximation to use the average applicants stats to calculate the schools lizzyM.

Since schools dont publish that, this either wont work, or you could probably get away with using the national applicants stats of 28 and 3.4something. Atleast it would work for all the schools except the top ones.
 
Yeah it doesnt work anymore, maybe the OP took it down?
 
I swear... if I wasn't a neurotic, procrastinating premed coming in, I sure will be going out!

I want to see the damn attachment! :)
 
Gah you guys are right, the attachment is down. I thought maybe SDN would be a reliable place to host it. =(. I will find another place to host it.
Here it is!
 
Wow. This is pretty awesome. Thanks.
Question: On the mdpotential.com site, is the % Acceptance equal to % Acceptance given that you have been offered an interview? Or just % Acceptance if you apply? (Since that would make a big difference.)
Thanks!
 
Yep it doesn't work.
 
Wow there are a lot of people clicking on that file! The hotlinks place had bandwidth issues. Thanks for the person who mentioned 2shared.com!
 
I was able to download your most recent link just fine. Very pretty formatting. And your mdapps is crazazy!

However, I'm too lazy to look up all the numbers needed to use your formula. ;)
 
bozz said:
www.mdpotential.com

basically uses mdapps data to calculate %chances
Click to expand...

Not exactly the best calculator.... I just used it and it said my chances of getting into BU were 0%, and i got in on the first round of acceptances. dont rely on that one for anything.
 
doesn't work.

EDIT: Got it to work.
 
bozz said:
www.mdpotential.com

basically uses mdapps data to calculate %chances
Click to expand...

zeb414 said:
Not exactly the best calculator.... I just used it and it said my chances of getting into BU were 0%, and i got in on the first round of acceptances. dont rely on that one for anything.
Click to expand...
That's because MDapplicants (the site this calculator draws from) is not at all reliable as an accurate source of admissions data. If I wanted to, I could go in to MDapps right now, make a fake profile, completely skew the stats, and therefore change the results of the calcluator for everyone.
 
Can someone attach it through SDN?
 
yay it works now thanks :thumbup:
 
Jacka1776 said:
She is apparently one of the administrator from a Medical School where she invented a score that distinguishes who is qualify to receive the secondary and/or an interview vs. who isn't via using simple Statistical Calculations. The process makes it fairly easy for the school if there are a large batch of applicants trying to matriculate 50-100 spots. I could be wrong though! :/
Click to expand...

Actually, I came up with this quick and dirty method for applicants to determine whether they have a good shot at being offered an interview - or not. My school does compare an applicant's stats to the average stats for our current matriculants and grades applicants as average, below average or above average. Your best shot at getting an interview comes if you are at or above average in comparison to the current M1 class.

We don't use the formula in making adcom decisions.
 
LizzyM said:
Actually, I came up with this quick and dirty method for applicants to determine whether they have a good shot at being offered an interview - or not. My school does compare an applicant's stats to the average stats for our current matriculants and grades applicants as average, below average or above average. Your best shot at getting an interview comes if you are at or above average in comparison to the current M1 class.

We don't use the formula in making adcom decisions.
Click to expand...

Also LizzyM would you agree that your score tends to be more accurate with "balanced" scores like 3.5 30 instead of 4.0 and 26..I don't know if that works out exactly but thats close.
 
novastorm said:
Also LizzyM would you agree that your score tends to be more accurate with "balanced" scores like 3.5 30 instead of 4.0 and 26..I don't know if that works out exactly but thats close.
Click to expand...

Does a 4.0/26 actually exist or is it like a Dr. Seuss character? Frankly, something that odd would have to be 1) an inflated gpa or 2) someone who had a very bad day and should not have been taking the MCAT. Obviously, regardless of the LizzyM score, someone with a 26 should retake the MCAT. Even then a 26 in light of the 4.0 looks like a severe case of hubris and the adcom may steer clear.
 
LizzyM said:
Does a 4.0/26 actually exist or is it like a Dr. Seuss character? Frankly, something that odd would have to be 1) an inflated gpa or 2) someone who had a very bad day and should not have been taking the MCAT. Obviously, regardless of the LizzyM score, someone with a 26 should retake the MCAT. Even then a 26 in light of the 4.0 looks like a severe case of hubris and the adcom may steer clear.
Click to expand...

I know a 3.9/27 no grade inflation, they are just allergic to standardized tests.
 
novastorm said:
I know a 3.9/27 no grade inflation, they are just allergic to standardized tests.
Click to expand...

Now that is a bit of a problem because the Step 1 and 2 and the board exams are similar tests. Some adcom members will worry that someone who can't do exceedingly well on the the MCAT won't pass the boards on the first try either.

If the 27 is 9/9/9, then the applicant might have a good shot at schools where the median or mean MCAT is in that range despite having a gpa that is obviously above the median gpa at (almost?) every school in the nation.

Still, a LizzyM score of 66 is going to rule out most of the upper tier schools regardless of how it is arrived at. The question then becomes is the 27 an application killer at a school with a Lizzy score of 66 or less?.
 
I know a 3.9/27 no grade inflation, they are just allergic to standardized tests
Click to expand...

That sucks, a lot of their life is going to be determined by standardized tests.
 
amwatts0322 said:
You love statistics, dont you?? lol reading that makes my head hurt.
Click to expand...

I love math and highly respect those who do mathematics professionally, they are gods among men. Those with tenure are gods among gods. In another life (and if medicine wasn't my primary interest), I would be a pianist and mathematician, those are two things that I have a interest in but just haven't had enough time to develop any talent in. In my present form, I am neither intelligent nor creative enough to do much in math. David Hilbert once had a student who dropped out of the graduate program at Gottingen University. Hilbert's explanation was that the student wasn't creative enough to do mathematics so he became a poet.

j-weezy said:
OP: nice use of LaTex
Click to expand...

Thanks! LaTeX is amazing and saves bazillions of time. For many documents now, I won't even consider WYSIWYG editors. WYSIWYM all the way!
 
LizzyM said:
Does a 4.0/26 actually exist or is it like a Dr. Seuss character? Frankly, something that odd would have to be 1) an inflated gpa or 2) someone who had a very bad day and should not have been taking the MCAT. Obviously, regardless of the LizzyM score, someone with a 26 should retake the MCAT. Even then a 26 in light of the 4.0 looks like a severe case of hubris and the adcom may steer clear.
Click to expand...

What about a 3.3/40S?
:laugh:

(no seriously...)
 
That's a Lizzy Score of 73.3. That's going to get a good look at any school with a Lizzy score < 73, particularly with the great MCAT.

But why a 3.3? That's going to be the question that the adcom members reviewing your application are going to ask. I would try to determine if there was something going on that explains what would generally be considered a low gpa. I would look at the school, the major, the breakdown by BCPM/All other (AO) and by year, and at each individual class grade. I'd also read the LORs and the essay looking for some explanation (LOR may tell me about the school's grading policy or about a problem the student experienced during the academic year), the essay may also cover an exceptional circumstance that relates to the low gpa. If the grades in the pre-reqs are in the 3.6 range and there has been no downward trend,and all other factors (experiences, research, LOR, "no red flags") are good then I'd recommend an interview.
 
amwatts0322 said:
Yup going back and looking at it, I think instead of using the average accepted students stats to calculate the schools LizzyM it would be a much better approximation to use the average applicants stats to calculate the schools lizzyM.

Since schools dont publish that, this either wont work, or you could probably get away with using the national applicants stats of 28 and 3.4something. Atleast it would work for all the schools except the top ones.
Click to expand...

I think I responded to you by PM but I guess I should do it here so everybody else can see it too. I think thats a great idea but like you said, schools don't publish the stats of the average person who applies. If you get the national rate of acceptance and the national rates that you have here, you can use this method to find your probability of being accepted "anywhere." This might be useful but its hard to say what it tells you because it doesn't take into account what schools you are applying to.

I think its too bad that schools don't publish applicant data. I could try assuming that applicants are 1 to 2 standard deviations below acceptances but that would just be a guess. In any case, my calculation will underestimate your chances of acceptance because it is based on matriculant data rather than applicant data.

So from what I can see, the two major flaws are:
1) It will underestimate your chances of getting into a particular school because of using matriculant stats instead of applicant stats
2) It will overestimate your chances of getting into at least one school since the independence assumption is fundamentally false in this situation

Hahaha thats almost paradoxical isn't it?

Thank you for your insightful comments!
 
OP, Thanks for the great post it is an interesting statistical idea. I think it would be cool if we could do a validation study of how well this model actually does in prediction once the cycle is finished. My other question would be is there a way using the curves in the front of the MSAR that we could try and model to deal with the skew of the distribution. Theoretically we should be able to generate a similar looking function. Normally I do a lot of work with binary data so this would be a stretch for me, but if I have some time Ill dig out my books and give it a go. One idea that occurred to me is trying to model the probability with some sort of redraw system whereby as since independence is violated as med schools make decisions during the cycle you can use this prior probability to determine future probability. Some sort of Bayesian priors I think.
 
You guys are intense. I love it.
 
You must log in or register to reply here.
Next unread thread

Similar threads

R
Building a School List - "Targets" in relation to LizzyM Score
Replies
2
Views
697
Mr.Smile12
Mr.Smile12
D
Where do I have better chance of getting into med school VA or TX?
Replies
11
Views
431
Mr.Smile12
Mr.Smile12
Happyswimmingawg
  • Locked
Do schools actually use LizzyM scores?
Replies
10
Views
1K
Happyswimmingawg
Happyswimmingawg
D
How should I calculate my LizzyM score if I did an SMP?
Replies
5
Views
869
Goro
Goro
D
@LizzyM Holiday Rule still applicable?
Replies
2
Views
1K
Mr.Smile12
Mr.Smile12
Top