Does anyone know of a resource for checking the average USMLE scores at different med schools?
USMLE Scores
- Thread starter sandiego1
- Start date
- Joined
- Jul 2, 2010
- Messages
- 1,872
- Reaction score
- 11
call admissions and ask?
- Joined
- Jan 10, 2010
- Messages
- 2,689
- Reaction score
- 3,662
No such resource exists. Some schools publish their scores, most don't. Some school's tell applicants their average, most don't. Schools can manipulate their statistics as well, such as only giving you the average of those that passed the test, or including retakes but not originals, etc. There was a thread on SDN about this, but somehow every school was above average by a significant margin.
- Joined
- Feb 2, 2011
- Messages
- 3,739
- Reaction score
- 631
There's a thread containing a bunch of them: http://forums.studentdoctor.net/showthread.php?t=102875&page=7
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
This list is in no way complete or verifiable, so take it with a grain of salt. Going to upenn isn't going to get you a 270 if you are a ***** any more than going to harvard ensures a top mcat score. Its the people, not the institution, that matters most. The only usmle related question you should ask is do they give you time off to study.
- Joined
- Jul 27, 2011
- Messages
- 882
- Reaction score
- 33
I've heard this stated several times on SDN. Isn't 5-8 weeks considered typical for adequately preparing for Step 1? I'd imagine you'd want to kill yourself beyond 8 weeks. How many schools are not going to give their students that length of break?
2
235788
it seems like a lot of schools hide them... which i think is total crap because its conveys valuable info about the education and students the school attracts.
- Joined
- Dec 16, 2011
- Messages
- 334
- Reaction score
- 4
Sometimes it may be in their best interest to give applicants the benefit of the doubt.
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
No more than mcat averages.
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
Many schools give four or less weeks. I don't have a list, unfortunately. Six to eight weeks is ideal, eight more because you can take a week off after you take it
- Joined
- May 27, 2011
- Messages
- 1,710
- Reaction score
- 35
I'm not aware of any undergraduate schools that advertise their students' MCAT scores.
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
I wasn't referring to undergraduate schools, I was referring to the entrance MCAT average of a given medical school.
2
235009
i think we've disagreed on this in another thread but here it goes again....
i think the institution's policies and resources play a significant role in your success on step 1. here are a few questions you should ask at an interview including the one that was already mentioned above:
1. how much time is allotted as exclusive step 1 study time (i would say 6-8 weeks is what you should be looking for)
2. how often are shelf exams used during the preclinical years? (the more often the better because it's good practice for step 1, familiarizes you with the style of board questions, and helps you differentiate between high and low yield info for when you study for step 1)
3. what kind of resources are offered by the school to help with step 1 prep? (look for intangibles like tutoring from upperclassmen or step 1 review sessions...try to see through free qbanks and kaplan courses because they're probably not really free and your tuition money is being taken from something else to pay for them)
- Joined
- May 14, 2010
- Messages
- 1,189
- Reaction score
- 359
MCAT tells you about the students a given school attracts but not about the education the school provides.
USMLE provides either or both, depending on your view. Some say step 1 scores are only a function of individual effort. In that case, average scores tells you primarily about the students a school attracts. Others say step 1 scores are affected by the quality of the school. Some schools use step 1 style questions on exams, others teach more closely to step 1 content. At the intersection of these two views are ways in which the school environment (time off to study, culture of achievement, mental health resources) permits individual effort to reach its potential.
2
235788
yeah really
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
We may have, I don't recall if I have ever argued about this on SDN before, but I argue an awful lot so it is possible.
Well first I want to emphasize that I merely said that the most important thing is the students. In my post I even emphasized that time is important. Give a motivated and intelligent student time to study, and they will achieve an excellent score. As far as the other things you mentioned, I feel like the utility of these things is debatable and, as a pre-med, probably not worth trying to delve into because you really aren't situated to may judgments on the utility of a lot of that stuff. As far as your point in 2. I would argue that shelves are of mixed utility. True, you do get some level of insight into the NMBE, but as I have taken all of the clinical shelves I can tell you that much of the stuff you are arguing for isn't as easy peasy as you are saying. Taking the shelf for pre-clinicals could give you some insight into the heavily tested stuff, assuming that they infact do emphasize the same things (i'll take your word for it, as my school doesn't use pre-clinical shelves). However, I can barely remember what was on my OBGYN shelf three months ago, let alone my medicine shelf a year ago, let alone my pre-clinical exams from 2 years ago. Because of that, unless you are meticulously compiling the questions after you take the exam, I doubt it has that much effect on your usage of first aid and other study resources. There are better ways, in my opinion, to determine what to study for the step 1 besides your foggy memories of old shelf exams. I would also argue that there are some advantages to having home grown exams as well. Correct me if it is different in the pre-clinical shelves, but you don't get to see what you got wrong. So while you have some sense of that after the exam and after you receive your shelf score, you really don't know if what you thought was right was right and what you thought you got wrong was wrong (unless you score a 0 or a 100 I suppose). In contrast, while the home grown exams (which do their best to reflect boards questions at most schools anyways) may not be from the same company that makes the step 1, at least you have the opportunity to find out what you got wrong on the exam and to get a better sense of exactly where your weaknesses were, what misconceptions you had were, etc. I'm not saying this is better than taking the shelf exams, merely that I am not convinced that it is worse. Maybe there is data out there that shows that it is, I don't know. As for 3, my school has those tutoring and review options that you mention and I have not found them particularly useful relative to my own personal studying. A nice perk, but once again not something that I would recommend that students use to choose a medical school.How exactly do you gauge that, though? Do you divide the step 1 score by the MCAT and look at the ratio of every school to see which school's "experience" seems to increase the average student's score? How many points on the step 1 correlates in your mind to a significant difference versus just random noise? Remember the standard deviation is something like 20+ I believe. What about schools that have fluctuations in their score? Does that mean that one year there was a culture of achievement and then another year that culture was lost and then another year they hired good mental health resources, and then the next year the fired them? Do you average the last 3 years?
As far as things like step 1 style questions and teaching towards the exam, I didn't interview at a school that claimed NOT to do those things. They are buzz words, how would one group of medical students really know if their curriculum is better at "teaching to the boards" than average? The fluctuations in scores are much more due to individual variability. Maybe one year there were a few more students going into psych who decided to study less than they would if they were going into ortho, maybe another year there was a much more intense group of students, but it really doesn't say anything about YOUR future class. At least with MCAT averages medical schools tend to keep them pretty constant year after year.
I'm not denying that a school can have some effect on your step 1 score, I'm simply arguing that it is really hard to USE average step 1 scores to determine that effect versus simply being the result of the individuals within the group. Look for the resources directly rather than their result, because it is too complex to be distilled into a simply three digit number. I stand by the time off thing, that is probably the only factor at a medical school related to the USMLE step 1 that is worth actually factoring in with other major considerations like fit, location, and cost. Having digital microscopes in the lecture hall is fancy and cool, but much more ambiguous when the fact of the matter is self study is the primary way that you study for the step 1.
- Joined
- Dec 19, 2009
- Messages
- 2,078
- Reaction score
- 10
^ ^ ^ is probably the best articulated argument for this point i've seen to date 
- Joined
- Jun 1, 2011
- Messages
- 626
- Reaction score
- 868
-
Last edited:
2
235788
yeah it was pretty solid^ ^ ^ is probably the best articulated argument for this point i've seen to date
totally agree
- Joined
- Dec 20, 2004
- Messages
- 30,873
- Reaction score
- 10,063
the schools all tacitly agreed not to publish scores because it limits the ability of places to try new and potentially better forms of teaching (eg PBL, more clinical exposure in early years, etc). The concern us that if there's a big risk in terms of published board scores to trying new things, no new innovations will ever happen. So actually the places that publish are being disreputable. Additionally, there's more than one way to calculate scores -- do you leave out failures if there's a retest, etc, so a places score might not be a fair comparison to another places score. Then there's the notion that you can teach to the boards, while missing out on teaching real clinical medicine, so you can sometimes see good board scores from folks whose medical education is fairly limited. In academic circles, "teaching to the boards" is a dirty word, a negative, and puts you in the group with offshore degree mill type schools. Finally the USMLE folks don't publish their scores because they are concerned folks will use them the way you (OP) want to (and the way residencies do) -- in fact they were only meant to measure minimum proficiency, not to be used to suggest that eg a 230 is better than a 225. Which is why they ultimately will get changed sometime soon. For the most part, the places that admit better test takers and the places that give the most time to study and the places that "teach to the boards" will likely have the highest scores. This may or may not jibe with the US News ranking, but based on the couple of schools that publicize their high scores, it probably doesn't. At any rate I agree with the above poster that you should be very suspicious of numbers listed on SDN. They are often fake and often not all calculated the same way.
Last edited:
- Joined
- Dec 20, 2004
- Messages
- 30,873
- Reaction score
- 10,063
There was a thread on SDN about this, but somehow every school was above average by a significant margin.
They probably excluded failing scores if there was a retake. Or they were just made up.
- Joined
- Jul 18, 2004
- Messages
- 3,293
- Reaction score
- 224
One day you will realize that it's more of an independent effort than that of what the school is doing.
- Joined
- Jan 10, 2010
- Messages
- 2,689
- Reaction score
- 3,662
My thoughts on school factors that likely have a big impact on step1 scores.
(1) More time off for board study = better board scores (within reason, I'm sure there are diminishing returns to increased time)
(2) Mandatory lecture: less time for board study = worse scores
(3) High avg MCAT for a school: There is an MCAT to Step1 correlation, though its far from perfect. High MCAT likely indicates that you are a good test-taker and/or have good study skills. This says nothing about the school except that they weight the MCAT highly.
Clinical time before boards?? This is hard to say. Seeing lots of patients can help you remember diseases, management, and common presentations, but step1 tends to be more about minutiae and pathophys. I know that I would lose a lot of the things I learned in first 2 years, but I'm sure that it would be helpful for some. Probably a tossup that depends on the kind of learner you are.
(1) More time off for board study = better board scores (within reason, I'm sure there are diminishing returns to increased time)
(2) Mandatory lecture: less time for board study = worse scores
(3) High avg MCAT for a school: There is an MCAT to Step1 correlation, though its far from perfect. High MCAT likely indicates that you are a good test-taker and/or have good study skills. This says nothing about the school except that they weight the MCAT highly.
Clinical time before boards?? This is hard to say. Seeing lots of patients can help you remember diseases, management, and common presentations, but step1 tends to be more about minutiae and pathophys. I know that I would lose a lot of the things I learned in first 2 years, but I'm sure that it would be helpful for some. Probably a tossup that depends on the kind of learner you are.
2
235788
I can see that. But i'm sure second year education does help with the preparation.
But then again, my med school pre-reqs didn't do wonders preparing me for the MCAT
- Joined
- Dec 20, 2004
- Messages
- 30,873
- Reaction score
- 10,063
it helps -- you are certainly at a disadvantage if you did poorly second year and have to start from scratch. That's your foundation, on which you really want to build your knowledge. But that being said, you are going to really learn what is going to get you through that test by reading first aid and other prep materials, and doing qbank type problems 11 hours a day every day for the 6-8 weeks you spend preparing for the test.
- Joined
- Aug 11, 2011
- Messages
- 1,216
- Reaction score
- 16
Your step1 score is a function of your own studying/work ethic and intelligence - you can get a 260 from a below average school and a 200 from an ivy league. Go to a med school where you're happy and that you can afford (that's a relative term, obviously).
- Joined
- Dec 19, 2009
- Messages
- 2,078
- Reaction score
- 10
exactly. your pre-med experience should tell you something, then. There are far, far too many confounders with this kind of analysis to make it worthwhile. At the same time, everyone wants them. Did you choose your undergrad based on average MCAT score?
mmmcdowe
Duke of minimal vowels
Staff member
Administrator
Volunteer Staff
Lifetime Donor
15+ Year Member
- Joined
- Sep 13, 2008
- Messages
- 9,912
- Reaction score
- 1,946
To provide some support to that, reported scores at my school have range from low 200 to low 280 in the last four years.
- Joined
- May 14, 2010
- Messages
- 1,189
- Reaction score
- 359
If there were USMLE data available by school publicly, it would be possible to perform a simple regression analysis using matriculant MCAT average as the independent variable and step 1 average as the dependent variable.
The regression would provide a confidence interval. This is what you call "experience." Schools that fall outside the confidence interval are under- or over-performing. This confidence interval defines the "noise."
The standard deviation of step 1 scores in 2011 nationally across individual test-takers is 22. The standard deviation of step 1 scores at one school, UVA, which provides its information straight from the NBME, is 20. That's pretty close (pretty variable). We can see, however, from previous threads of reported step 1 averages that the inter-school variability seems to be less than the intra-school variability.
For the purposes of calculating the population standard deviation, it would be consistent to assume that each school is a random sample of students.
If you assume that, purely for the purpose of looking at the variation, then it follows that the standard deviation of the averages (standard error of the mean) would approximately be equal to the sample standard deviation divided by the square root of the sample size (class size at a given school).
The standard deviation of school-level averages would then only be on the order of 1-3 points.
This seems low. You'll say, well there's a clear hierarchy of schools with some schools performing much better. I'll say that the standard deviation across schools will still not be as big as say, 15.
Another line of evidence for relatively low noise levels is on the school level, step 1 averages are relatively stable across time, except with a slight upward trend over time, unless there are dramatic changes in curriculum or something. Averaging across several years is a valid way to reduce the effects of noise. At UVA, the standard deviation of mean step 1 scores over the past four years is only 1.5. The standard deviation of means over the last 11 years is only 4.5.
All this to say that in my mind, a difference of 10 points above predicted step 1 based on matriculant MCAT average would be significant. Mayo comes to mind.
- Joined
- Dec 19, 2009
- Messages
- 2,078
- Reaction score
- 10
your study design fails at the level of your assumptions. it is a mistake to assume that student bodies are a random sample. people pick specific places to attend med school precisely because they believe they are a better fit for them. and the school accepts them because of a reciprocal sense at the adcom's end. using your suggested approach, based on this kind of mechanistic method, to choose a school, is crazy.
a simple example: i don't know if Mayo is overperforming compared to other schools or not. no one really does, unless they've seen the NBME numbers for Mayo as UVA shows them for UVA. even if we knew that Mayo was overperforming, it would require going to pretty much every class, because that's the kind of place Mayo is. if you wind up being someone for whom that doesn't work, you are not going to be a person for whom the supposed institutional advantage at Mayo is likely to work, because you yourself are an outlier within that culture.
This is a classic example of the value of stats breaking down at the level of the individual's choice. if your analysis showed that one school was beating the tar out of everyone else, and that school was all PBL, and you attended a class which made you hate the idea of PBL, would you still go?
