For those who have Step I scores back, NBME 150 questions correlation?

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

Hallm_7

Senior Member
10+ Year Member
15+ Year Member
20+ Year Member
Joined
Feb 18, 2002
Messages
685
Reaction score
3
For those of you who have already taken Step I and received your scores, what kind of correlation are you seeing with your score and the % correct on the released 150? Just curious how accurately the Medfriends.com calculator will predict Step I scores using the percent correct on the 150.

Thanks!

Members don't see this ad.
 
  • Like
Reactions: 1 user
Well it seems like everyone is acing those 150question items. So if I had to guess, I would say it's a horrible measure of how you're gonna do and it drastically overestimates your score.
 
  • Like
Reactions: 1 user
I have yet to take the exam but I wouldn't say it is a TOTAL overestimate. Yes, my score on the 150 released questions was higher than my score predicted by NBME, QBANK, MCAT, etc. but only by about 10 points. I think those questions represent the easiest 30% of the questions so there should not be too much difference. Just my humble opinion....
 
  • Like
Reactions: 1 user
Well it seems like everyone is acing those 150question items. So if I had to guess, I would say it's a horrible measure of how you're gonna do and it drastically overestimates your score.

The database at Medfriends counts for this though. It's based on regression analysis equating the practice exam to a confidence interval on the real deal. A 70% on the 150 released will translate into a lower score than a Qbank 70%. For example, 70% on the 150 translates into a 214 while 70% on Qbank translates into a 232.

However, regression models are only as good as the data they contain, so I was just curious if people here have found the Medfriends calculator accurate as it pertains to the released 150. The website actually says the 150 released analysis is the most accurate.
 
Members don't see this ad :)
why is no one actually answering the question. it's a simple one: did your score correlate or not. what's up with all this other talk . . .
 
why is no one actually answering the question. it's a simple one: did your score correlate or not. what's up with all this other talk . . .

perhaps the people who already have their scores (or took it last year) do not visit this forum any more!
 
I haven't taken the exam yet but the general consensus seems to be that the 150 free questions are much easier than the actual exam.
 
I took the usmle today....the free 150 were comparable to the easier 30-40% of the real thing IMO

Wow...that's not bad!

I scored surprisingly high on the free 150 and would absolutely love it if 40-50% of my actual exam contained similar questions :p
 
Wow...that's not bad!

I scored surprisingly high on the free 150 and would absolutely love it if 40-50% of my actual exam contained similar questions :p
everyones test will vary....but I would assume everyone has a fair share of easy-moderate questions thrown in b/w the real thinkers and wtf q's....and if not....the thing is scaled right?....ie I would assume that if someone got a test w/ the majority of q's having like a 20% correct avg accross the country that the nbme scales that sorta thign to even things out....
 
everyones test will vary....but I would assume everyone has a fair share of easy-moderate questions thrown in b/w the real thinkers and wtf q's....and if not....the thing is scaled right?....ie I would assume that if someone got a test w/ the majority of q's having like a 20% correct avg accross the country that the nbme scales that sorta thign to even things out....

go with your first assumption. There are 350 questions and the questions are randomly pulled one by one from the usmle's test bank(of god knows how many questions). So everyone's test is going to be of "about the same difficulty".

Nobody gets a "hard test" or an "easy test". Now an individual may get more questions in one area than another, and that may result in a higher or lower score due to that individual being good at the area tested.....but the general difficulty is going to be about the same
 
the test isn't really scaled like the way other tests are, like the MCAT, right? they just have data on each question so they know 90% get these right, 10% get these right, etc, then they mix and match from a database of (i've heard 17,000) questions to generate your test. that's why they put experimental q's on there, to correlate your performance on them to the total score, etc.
otherwise, how could they release scores when thousands haven't taken the test yet?
i'm sure they have checks in the algorythm (sp?) that generates the test so that you don't get multiple questions on the same thing or a test that's 70% GI or something. so you might get lucky and know a few of the "hard" questions, but that will even out over the course of the entire test
 
I took the usmle today....the free 150 were comparable to the easier 30-40% of the real thing IMO

I agree with this. And to answer the OP's question, my predicted score (from medfriends) on the 150 qs correlated almost perfectly with my score on the actual exam. But it also correlated with my last NBME, which I think contained questions that were more difficult and more like the real thing. So, both predicted my score but only the NBME qs reflected what the exam was like in terms of easy and harder questions. I may be a bit of an anomaly because I scored very well on the NBME, 150 qs, and actual exam, but if your predicted scores from the 150qs and NBME are similar (as they were for me and one of my friends) I would expect a similar score (within 5-10 points) on the real thing. Perhaps there is someone who aced the 150qs but not their last NBME who can shed light on which one better predicted their score?
 
Members don't see this ad :)
I agree with this. And to answer the OP's question, my predicted score (from medfriends) on the 150 qs correlated almost perfectly with my score on the actual exam. But it also correlated with my last NBME, which I think contained questions that were more difficult and more like the real thing. So, both predicted my score but only the NBME qs reflected what the exam was like in terms of easy and harder questions. I may be a bit of an anomaly because I scored very well on the NBME, 150 qs, and actual exam, but if your predicted scores from the 150qs and NBME are similar (as they were for me and one of my friends) I would expect a similar score (within 5-10 points) on the real thing. Perhaps there is someone who aced the 150qs but not their last NBME who can shed light on which one better predicted their score?

It looks like I will be able to shed some light on this in about 4 weeks, since my last NBME was 228, but I took the 150 today and medfriends predicted a 247 :eek:
 
that score calculator is insane. i mean, its methods may be valid, but i'm certainly not relying on it. i'm supposedly on track for a 244 . . . HA! i wish! i don't even feel worthy to type that number

i just took the 150 released items yesterday after only studying biochem and behavioral sciences and i was at 75% overall . . . but . . . i'm a total slacker. i mean, i am in NO WAY even ready for this test (still have 7 weeks on cramming and questions) and i'm supposedly doing alright? i don't buy it.
 
I don't know what to make of the score calculator....here's what I'm getting:
Free 150: 87%=253
NBME 2: according to nbme, 740 (260), but according to medfriends, it's a 260
NBME 3: according to nbme, 770 (263), but according to medfriends, it's a 274

If I use the nbme scores, I guess my free 150/nbme paid tests roughly correlate w/in the standard of error, but it's weird that the score estimator predicts like 10 points higher than nbme. I guess maybe the anonymous score submitters do better on their actual test than on nbme or something? I am not sure what to believe. I would love to do that well on the test but want to be realisitic...
 
I don't know what to make of the score calculator....here's what I'm getting:
Free 150: 87%=253
NBME 2: according to nbme, 740 (260), but according to medfriends, it's a 260
NBME 3: according to nbme, 770 (263), but according to medfriends, it's a 274

If I use the nbme scores, I guess my free 150/nbme paid tests roughly correlate w/in the standard of error, but it's weird that the score estimator predicts like 10 points higher than nbme. I guess maybe the anonymous score submitters do better on their actual test than on nbme or something? I am not sure what to believe. I would love to do that well on the test but want to be realisitic...

I believe you're gonna be ok......and even if you do slip all the way down to 253 instead of the 274 I'm sure FP in bumble*%$ will still have a spot for you
 
Haha, OK, I realize that post sounded a little ridiculous in reading it again....seriously, I'm not that kid. However, I definitely have a history of doing well on practice tests, then blowing it on the real thing.

And for the record, since I want to do FP or IM in a rural area, I would love to get one of those spots, so I am glad to hear you think they will take me. :)
 
Haha, OK, I realize that post sounded a little ridiculous in reading it again....seriously, I'm not that kid. However, I definitely have a history of doing well on practice tests, then blowing it on the real thing.

And for the record, since I want to do FP or IM in a rural area, I would love to get one of those spots, so I am glad to hear you think they will take me. :)

For the more record... you are DEFINITELY "that kid." :laugh:
 
How does everyone on this site get >260 on the NBMEs its crazy...i'm going to go cry myself to sleep now
 
I thought the real thing was much harder than the free 150. I have yet to receive my score...

free 150 was easier... much easier... the questions on the free 150 were prob representative of the easiest 30-40% from real thing.
 
free 150 was easier... much easier... the questions on the free 150 were prob representative of the easiest 30-40% from real thing.

I suppose that's one way of looking at it. The problem is, there are enough people who do these questions intending them to be a reflection of the WHOLE exam, not just the easiest questions.
 
For the more record... you are DEFINITELY "that kid." :laugh:

OK, that's fair, I guess. Sorry, really wasn't trying to be a jerk, but can see how it came off...anyway, my point is just that for me, the scores were also a little variable depending on which source you use to predict, plus they disagree w/ what the nbme score conversion is. So yea...like everyone else, not sure how accurate the score predictor is (or the nbme's themselves, for that matter). Something to try to remember to include in a "step 1 experiences" post...if I survive til then. :) Good luck, all.
 
my free score was 78% which equals 235....my real score was in the 230's!

wow, really? did you take it as a diagnostic or leading up to the real deal? i mean, again, i scored 75% and i've barely started studying . . . .and i'm shooting for 230 but have serious apprehension about setting my sights that high. seriously, the most adamant advice i get from other med students is to STAY AWAY FROM SDN 'cause all the 240+'ers are not realistic . . . but i just can't, i'm like a moth drawn to bright blue light. so anyway, i don't know what to think about all this score prediction hoopla . . . i heed the advice on here because, well, it all sounds pretty solid and has proved useful for many people i know ahead of me. but somehow i still have reservations that i'll be able to live up to SDN standards . . . sigh . . . 7 more weeks of studying and i'll get back on here and might have stronger feelings one way or the other. anyway, i don't know where this little rant came from.

f*&%ing NBME and their stupid exam . . . i want my life back!
 
wow, really? did you take it as a diagnostic or leading up to the real deal? i mean, again, i scored 75% and i've barely started studying . . . .and i'm shooting for 230 but have serious apprehension about setting my sights that high. seriously, the most adamant advice i get from other med students is to STAY AWAY FROM SDN 'cause all the 240+'ers are not realistic . . . but i just can't, i'm like a moth drawn to bright blue light. so anyway, i don't know what to think about all this score prediction hoopla . . . i heed the advice on here because, well, it all sounds pretty solid and has proved useful for many people i know ahead of me. but somehow i still have reservations that i'll be able to live up to SDN standards . . . sigh . . . 7 more weeks of studying and i'll get back on here and might have stronger feelings one way or the other. anyway, i don't know where this little rant came from.

f*&%ing NBME and their stupid exam . . . i want my life back!

haha..ull be fine. Ya i did that like 3 days b4 my exam and thats what i got.

I also did NBME 3 3 days b4 my exam and i got a 214...SO my point is just keep working and you will get the score you want. All i wanted was above a 230.
 
According to this (http://www.usmle.net/step-1/messages2006b/209674.html) a 221 is equal to an 89. I don't know anything about this correlation, so this could be entirely inaccurate. I just googled and this came up.

its not very accurate....

i believe it varies by the difficulty of the test. I got a 97 as my two digit. and according to the scale, i should be much lower than that.
 
I've heard that the score you get on the NBME exam correlates well with what you'll get on the real thing so I took it #3 'cause I heard that one was suppose to be the hardest. However, for the real thing I thought it was considerably more difficult than the NMBE exam. It was actually similar to Kaplan questions. I haven't gotten back my scores yet but I'd like to know if anyone else felt this way.
 
I've heard that the score you get on the NBME exam correlates well with what you'll get on the real thing so I took it #3 'cause I heard that one was suppose to be the hardest. However, for the real thing I thought it was considerably more difficult than the NMBE exam. It was actually similar to Kaplan questions. I haven't gotten back my scores yet but I'd like to know if anyone else felt this way.

Don't know about the real thing being like Kaplan, but I thought form 3 did not represent the real thing at all. The real thing was much harder.
 
come on guys the free 150 Qs they gave us are just a joke !
I've done the 1999 sample ( which was the first or second year of computer based testing ) and i did the 2007 sample questions and they were almost the SAME !
so in 10 years' time they changed about 10 out of the 150, i would never rely on these Qs as a correlation for the real score.
 
What percent correct is everyone getting on the free 150?
 
Not that accurate in my opinion. I got 96% on the 150 and 247/99 on the real thing. Too many variables.
 
Not that accurate in my opinion. I got 96% on the 150 and 247/99 on the real thing. Too many variables.

I don't understand...how is that not an accurate correlation? You rocked the 150 and rocked the exam...how much more accurate should it be?
 
I got 70% on the 150 free questions in the final days before the test.

The score estimator predicted a 214.

My actual score is a 209.

Pretty good correlation if you ask me.

By the way, I averaged 45% to 50% on Kaplan and the score estimator said I was going to fail based on my Kaplan score.

Kaplan qbank questinos were NOTHING like the actual test. Use it to help you learn the material and get familiar with the testing software format, but not as a "mock test" because it won't prepare you for the question styles on the actual exam. Also the Kaplan questions are way too picky and test minutae which I guess forces you to broaden your knowledge base but is annoyingly low yield.
 
Mine correlated very well, better than NBME & especially Kaplan (for some reason I never really broke through on Q bank until the last week or so, but I did every single question so I think it paid off even if my average was low). I was shocked by the correlation \to be honest because I thought the 150 were a lot easier than the actual exam.

83% correct, medfriends predicted a score of 245. My actual score was 240.
 
I don't know what to make of the score calculator....here's what I'm getting:
Free 150: 87%=253
NBME 2: according to nbme, 740 (260), but according to medfriends, it's a 260
NBME 3: according to nbme, 770 (263), but according to medfriends, it's a 274

If I use the nbme scores, I guess my free 150/nbme paid tests roughly correlate w/in the standard of error, but it's weird that the score estimator predicts like 10 points higher than nbme. I guess maybe the anonymous score submitters do better on their actual test than on nbme or something? I am not sure what to believe. I would love to do that well on the test but want to be realisitic...

Got my score today......258/99. Looking at my various predicted scores (253, 260, 263, and my qbank average predicted score of 256), I'd say the estimator was right on the money for me. I was concerned my nerves would get the best of me on test day, but things seemed to have worked out for the best, or at least to my extreme satisfaction.
 
i took it about 2 wks before the exam and i scored 3 pts higher than what was projected. the questions on the free 150 were easier than the real thing, but the correlation was there.
 
I already posted this correlation in another thread, but:
free 150 2 weeks before: 254
2 weeks of massive studying later:
241 on the real thing.

So yes, I thought the free Q's were easier than the real thing. They aren't a joke, but the $45 tests are going to be better predictors. You've already spent tens of thousands on your medical education-suck it up and pay for the forms.
 
definitely fits well with the free 150 from usmle, but agree with other posters that the nbme forms 1-4 may be better indicators of scores. good luck to those still studying/waiting on the mail.
 
My actual Step 1 score was 6 points off from the medfriends prediction based on the 150 free. As others have stated, the correlation is solid. I actually thought those 150 questions were a pretty good representation of my real exam.
 
I thought I would post my Free 150 percent and my actual score since I started this thread:

Free 150 two days before the test: 86% which correlated to 251.
Real deal Step I: 231/96

So for me it didn't correlate all that well, but I wasn't really expecting it to since it projected me at a 251 (I thought that was ludicrous given my performance on Qbank and the NBME Forms). I'm still incredibly pleased with my score.
 
Top