This Can't Be True...

[filmdoc26]

---

Members don't see this ad.

Anyone from UPenn out there? I heard that this year's average on Step I was in the mid 240s. I know that in terms of preparing for the boards, UPenn's a great school and all, but mid 240s? Is this for real?

 

[MD'05]

Mid 240's - I can believe that - UPenn admits students with great credentials and they probably focus on TEACHING and preparing their students for Step I, unlike other schools that will go unnamed that I know play with their numbers to make it look like their students have a higher average score on Step I. What a bunch of bull**** medical school is!!!

 

[Bonds756]

244 was the avg this year at Penn. I guess that's not too bad, huh?

 

[filmdoc26]

That's incredible. Do other schools perform like this? Well, I'll likely bring that average down when I get there! Thanks for confirming this Bonds

 

[doepug]

No way.

The average board scores at Penn (and other similar institutions) are about 15 points lower than that. Schools do share this data (internally), and it's reviewed by a committee at my school where I know a number of the members. Most "top tier" med schools have mean scores between 220-230, with a few exceptions.

Try the statistics. If Penn's average were that high, it would mean that the overwhelming majority of test takers with scores >240 are Penn students.

Don't believe everything you hear.

 

[filmdoc26]

The average board scores at Penn (and other similar institutions) are about 15 points lower than that. Schools do share this data (internally), and it's reviewed by a committee at my school where I know a number of the members. Most "top tier" med schools have mean scores between 220-230, with a few exceptions.

hehe...

Don't believe everything you hear.

interesting.

 

[MD'05]

How many weeks does Penn give its students to study for the USMLE? I don't see this info on their website. I am curious about this since it seems like some schools play with numbers by giving more study time.

 

[Brewster]

Don't forget that people get into places like Penn and other top ranked schools by being bright and good at taking tests. And, it is bright people who are good at taking tests who are going to do well on the boards, regardless of what school they go to.

Your school's curriculum may have some impact on your board score, but in the end the boards are an individual accomplishment.

Not to take anything away from Penn or the students, as they have both clearly done a remarkable job.

 

[MacGyver]

doepug is right... that number cant be right.

Lets look at national statistics.

The mean score on the USMLE Step I is about 200, with a standard deviation of 20 points.

To get a score of 240 requires 2 standard deviations above the national mean, or in other words you score in the top 5% of all test takers.

Penn's class size is 150, so if their mean score was 240, that means approximately 75 people scored 240 or greater.

There are approximately 16,000 med students which take USMLE step I every year, with the top 3% getting 240 or above, that translates to approx 480 persons.

There are over 300 schools which participate in the USMLE step I whose scores are included in the statistical profile. If that statistic were true, it would mean that Penn accounted for more than 15% of all test scores greater than 240, with the other 299 schools combined accouting for the other 85%. Penn's a good school, but its not THAT much better than the other schools.

Bottom line: The 240 statistic is not correct. Please stop posting BS numbers and provide a link with real data, instead of made up make-believe numbers.

 

[MacGyver]

Originally posted by Bonds756
244 was the avg this year at Penn. I guess that's not too bad, huh?

So were just supposed to believe an anonymous internet poster?

link please.

 

[md_student10021]

Originally posted by MacGyver
doepug is right... that number cant be right.

Lets look at national statistics.

The mean score on the USMLE Step I is about 200, with a standard deviation of 20 points.

To get a score of 240 requires 2 standard deviations above the national mean, or in other words you score in the top 5% of all test takers.

Penn's class size is 150, so if their mean score was 240, that means approximately 75 people scored 240 or greater.

There are approximately 16,000 med students which take USMLE step I every year, with the top 3% getting 240 or above, that translates to approx 480 persons.

There are over 300 schools which participate in the USMLE step I whose scores are included in the statistical profile. If that statistic were true, it would mean that Penn accounted for more than 15% of all test scores greater than 240, with the other 299 schools combined accouting for the other 85%. Penn's a good school, but its not THAT much better than the other schools.

Bottom line: The 240 statistic is not correct. Please stop posting BS numbers and provide a link with real data, instead of made up make-believe numbers.

Except that the national mean is about 215, not 200. A 240 would be a liitle more than 1 sd above the mean; around 90th %ile. -->1600 students, 75 from penn. Not as exceptional as you make it out to be. Still, 240 would be a bit high. I suspect the school's average is more like 230-235.

 

[Jalby]

Originally posted by MacGyver

The mean score on the USMLE Step I is about 200, with a standard deviation of 20 points.

To get a score of 240 requires 2 standard deviations above the national mean, or in other words you score in the top 5% of all test takers.

You forgot about the "other" 2.5%. I heard the average was about 215, but I agree that 244 does sound extremely high.

 

[doepug]

Ok, here's the math. I hate being the BS police.

mean score = 215
standard deviation = 15
Source is my USMLE Step I score report, which I received 3 months ago.

Therefore, a mean of 244 is almost equal to 2 SD above the mean.
2 SD includes 95% of the distribution.
Therefore, the top end of the curve is 2.5% of the distribution.
2.5% of 16,000 test takers is 400 people with awesome scores.

Should we honestly believe that 150/400 (3/8) of great scores belong to Penn students? If the other 250 great scores get divided evenly around the country, it would mean that only 2 students at each American medical school would have a score that is equal to the average score at Penn. It's BS, plain and simple.

Yes, people from Penn are smart. Yes, they're great test takers. But the rest of us at "similar" schools are great test takers too. The simple truth is that Penn's board scores are in the 220-230 range, in line with Harvard, Hopkins, Duke, Wash U, Cornell, UCSF, Stanford, etc. If you're still a skeptic, ask your dean.

 

[MD2b06]

Originally posted by doepug
Ok, here's the math. I hate being the BS police.

mean score = 215
standard deviation = 15
Source is my USMLE Step I score report, which I received 3 months ago.

Therefore, a mean of 244 is almost equal to 2 SD above the mean.
2 SD includes 95% of the distribution.
Therefore, the top end of the curve is 2.5% of the distribution.
2.5% of 16,000 test takers is 400 people with awesome scores.

Should we honestly believe that 150/400 (3/8) of great scores belong to Penn students? If the other 250 great scores get divided evenly around the country, it would mean that only 2 students at each American medical school would have a score that is equal to the average score at Penn. It's BS, plain and simple.

Yes, people from Penn are smart. Yes, they're great test takers. But the rest of us at "similar" schools are great test takers too. The simple truth is that Penn's board scores are in the 220-230 range, in line with Harvard, Hopkins, Duke, Wash U, Cornell, UCSF, Stanford, etc. If you're still a skeptic, ask your dean.

I agree with everything you've said, but just want to point out a couple of things. Isn't the number of test takers higher than 16000 since you have to include all the Carribean grads + FMG's, and the DO's that take the exam in addition to the COMLEX? Also, why does the Step 1 score report arrive 3 months ago if you take it after the summer of your 2nd year? It doesn't take that long to receive your scores does it?

 

[doepug]

Originally posted by MD2b06
I agree with everything you've said, but just want to point out a couple of things. Isn't the number of test takers higher than 16000 since you have to include all the Carribean grads + FMG's, and the DO's that take the exam in addition to the COMLEX? Also, why does the Step 1 score report arrive 3 months ago if you take it after the summer of your 2nd year? It doesn't take that long to receive your scores does it?

Ok, so the number is liable to be slightly higher than 16000, but 2.5% of a couple hundred more people isn't going to add up to anything. Even if you figure that there are 1,000 non-US allopathic students taking the exam, that only comes to 25 people who can't change the conclusion. If anything, they make it even less likely that Penn students have an average that high, since FMGs/osteopaths have a lower mean score, as reported by the NBME.

My school doesn't require the USMLE, so I took it on 12/31/02.

Cheers,
doepug

 

[MD2b06]

Originally posted by doepug
Ok, so the number is liable to be slightly higher than 16000, but 2.5% of a couple hundred more people isn't going to add up to anything. Even if you figure that there are 1,000 non-US allopathic students taking the exam, that only comes to 25 people who can't change the conclusion. If anything, they make it even less likely that Penn students have an average that high, since FMGs/osteopaths have a lower mean score, as reported by the NBME.

My school doesn't require the USMLE, so I took it on 12/31/02.

Cheers,
doepug

Actually, according to this link, the total number of USMLE Step 1 test takers is closer to 32,000.

http://www.usmle.org/news/2001perf.htm

However, I still think a 244 avg for UPenn is BS. Not that it's not a great school, but there are too many other great schools out there as well.

BTW, how can Hopkins not require the USMLE? Did you mean you're not required to take it before 3rd year?

 

[doepug]

It doesn't matter how many non-US or non-allopathic med students took Step I. Confine your population to US allopathic med students, and the sample size is large enough for a normal distribution. I just refuse to believe that the "average" Penn student performs better than 98.4% (1 - 250/16000) of the rest of us in American allopathic schools.

Originally posted by MD2b06

BTW, how can Hopkins not require the USMLE? Did you mean you're not required to take it before 3rd year?

Hopkins is sort of anti-standardized testing. The reason I was given was that among a handful of variables, the USMLE score had the worst correlation with medical student/resident success. Besides, plenty of Hopkins grads end up in research, public health, health policy, or another field. So, we're not required to take any step of the USMLE. For that matter, we don't have a prescribed third year curriculum either... for example, I started third year with two electives, and I'm putting off OB until after the match.

Hopkins didn't even require the MCAT until the school joined AMCAS about 4 years ago.

-doepug

 

[filmdoc26]

Okay...this link is getting a bit more heated than I had anticipated. However, it's still pretty interesting. To be sure, I'm still doubtful, but I'm not convinced that the average at UPenn was NOT 244 this past year. At least not yet. I guess we'll find out at the Penn Preview weekend in two weeks. As for now, let's keep playing with numbers. This is amusing. I've never taken a stats class or anything, but let's do this:

Should we honestly believe that 150/400 (3/8) of great scores belong to Penn students?

Regarding the above quote, shouldn't the number be 75/400 because only half of the entire class scored in that top 2.5%?

BUT,
if we rightly assume that the pool is indeed closer to 32000, then the top 2.5% is going to be 800 rather than 400. The 50% of Penn's class that supposedly scored in this 244 + range should come to around 75/800, which is about 9.3%.

SO...(and I admit, I may be wrong because math and stats are definitely NOT my forte), this means that 9.3% of the top 2.5% of all USMLE Step I testakers are Penn students. Certainly this is a very large proportion favoring Penn students, but is this SO unreasonable that it's "BS"? The other 90.7% (or the other 725 out of 800) could come from a number of very well qualified students from other excellent institutions (yes...including Hopkins)

I don't know...to me it seems reasonable, but perhaps still a little "crazy". Anything wrong with my math up there? Hmm...maybe doepug has some thoughts?

 

[kcrd]

Do you know why no one from Penn has posted to confirm or deny the 244 mean? Because they are all studying for the step 1!! 😀

 

[MacGyver]

You show me a link where Penn has a confirmed 244 mean USMLE score and I'll believe it.

The fact is there is no such evidence. If their scores were really that high, then surely they would brag about it on their website or release it along with match statistics or whatever.

 

[doepug]

Sorry if I made it seem like this was getting heated. I'm not offended or upset; I just take a special interest in debunking certain posts.


Regarding the above quote, shouldn't the number be 75/400 because only half of the entire class scored in that top 2.5%?


So... we have to assume that Penn's scores are evenly distributed around a mean of 244. Since we don't know their standard deviation, we can't really compare the distribution of Penn's scores to the national distribution of scores. What we can do is compare the "average" Penn student (with a score of 244) to the rest of us.


BUT, if we rightly assume that the pool is indeed closer to 32000, then the top 2.5% is going to be 800 rather than 400. The 50% of Penn's class that supposedly scored in this 244 + range should come to around 75/800, which is about 9.3%.


No, the 32,000 figure doesn't matter. Let's stick to comparing apples to apples, and compare the average Penn student to the rest of the pool of 16,000 American allopathic medical students. By increasing the pool to 32K, all we do is account for all the FMGs and osteopaths who rocked the exam.


SO...(and I admit, I may be wrong because math and stats are definitely NOT my forte), this means that 9.3% of the top 2.5% of all USMLE Step I testakers are Penn students. Certainly this is a very large proportion favoring Penn students, but is this SO unreasonable that it's "BS"? The other 90.7% (or the other 725 out of 800) could come from a number of very well qualified students from other excellent institutions (yes...including Hopkins)


So I was a little bit off with my previous post. The average Penn student would still be 2 SD above the mean, putting them in the top 2.5% (n=400) of American allopathic test takers. Only half of their class would have a score greater than 244, so that would account for 75 of 400 great scores, instead of the 150 I cited earlier. Sorry for the confusion.

Distribute the 325 remaining great scores across 125 American medical schools, and it is still exceedingly unlikely that Penn's average is that high. Let's say that Penn outperformed the other 9 schools in the "top ten," having twice as many students with a great score as peer schools. That's 37.5 students x 9 schools = 337.5, which is greater than 325. I guess that means no one outside of the top ten got a score of 244 or better, right? Of course not.

In order for the 115 schools outside of the top ten to have just one student with a great score, Penn would have to have more than three times as many students with great scores than other "top ten" schools. (325 remaining scores - 115 for other schools = 210 for the "top ten" divided by 9 other "top ten" schools = 23 per class, being less than 1/3 of the 75 at Penn.)

Who honestly believes their average is that high? Show me your reasoning.

 

[md_student10021]

I think doepug has sufficiently explained why 244 is clearly not the average at Penn. You don't need complicated statistics. You just need a little common sense.

 

[filmdoc26]

Yeah...everything in the numbers had always made me incredulous about that mid 240s average. In fact, md_student, common sense seemed to be telling me that. But we know that common sense has failed mankind from the beginning of time (except in cases like when common sense tells early humans not to play with the angry grizzly bear). 😀

Anyhow, of course it's still a little "crazy". It's just that I had two Penn students confirming this number, which is why this thread is very interesting. Of course Penn has reason to lie, but I'm not convinced that they would, nor am I convinced that the average was NOT 244 this past year. Hehe...let's just say that I still stand in a state of socratic doubt about this whole thing: I know nothing.

The reason I'm not 100% convinced is because if we assume that 75/400 testakers (if we choose to use American/Canadian allo schools only) with "great scores" are from Penn, then 325 other students had "great scores", as doepug said. Again, this seems outlandish, but in my estimation not necessarily impossible.

I guess there's nothing left to do but wait til the Penn Preview weekend. Don't worry guys, I'll be sure to grill them about the score. Maybe I'll print out this whole thread and take it with me (doubtful)! Sorry MacGyver, there's no link to the current scores. So please put it to rest. All I know is that Penn's average Step I last year was 232, which is believable and is on the web for accepted students. Oh well...anymore thoughts?

 

[filmdoc26]

All righty ladies and gents! It's over. I just checked the UPenn status page that has the Step I and Step II average over the past several years. In 2001, like I had said before, the average was 232. In 2002, the average was 235. This must have been posted fairly recently.

Nice corresponding with you all (esp. doepug). Best of luck!

 

[doepug]

The 232 and 235 averages are much more believable.

Sorry about my low tolerance for BS about Step I,
doepug

 

[GAdasooka]

If you care enough to do the math, why not go ahead and research it instead? Wew... I think maybe I ought to go study for Step 1 now... (or maybe passing my next test?!)