Approximate questions right (out of 308 questions) on the REAL THING to get a certain score?

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

kehlsh

Medic Commando
10+ Year Member
Joined
Dec 28, 2009
Messages
1,137
Reaction score
37
Hey guys,

I know majority of you have already finished step 1 and now are moving onto clerkships.
Some of you are very smart so I am asking this out for the curiousity's sake.
I have taken 5/6 of the NBMEs online so I know approximately how scoring works there.
However, I was curious as to know if curve in real thing is any more lenient or is just as steep.

Any of the people who have great intuition on things... what would you say is the approximate % you have to get right to get a 230/240/250/260 bench mark?

I know this will be just speculative but I would still like to know.

Thanks
 
This is difficult for people to answer b/c no one gets a detailed report of how many they got wrong. From NBMEs, it looked like 80% correct was 240, 90% correct was 250. So you can probably go with that for the real thing, which is of course going to be curved, just as the NBMEs are curved I'm sure.

The short answer is no one knows b/c no one knows the curve, which is different for every exam.
 
how many questions correct on the NBMEs correspond to a passing score (196)? I failed my first one and now freaking. Not sure how I got such a low score, if 70% correct is the mean I feel like I should have at least been close to passing with the percentage of questions I got right.
 
The curve/scale varies on each exam, but when I compared multiple scores from one NBME (taking scores from friends), multiplying your raw score out of 200 by 1.4 was generally close to what the online NBME website gave me as the score (but that was when the scores were in the 240 range. Not sure what the multiplier might be in the 190 range)
 
Probably <2 wrong per block for >270, <3 wrong per block for for >260. Just based on NBME data. There are of course rumors about experimental questions that get "thrown out" and if those rumors are true then the NBME relationship isn't going to be accurate. A friend of a friend knows someone who writes for the USMLE and claims that any given exam is up to 15% new/experimental questions. For all we know the exam is actually only 6 blocks and everyone has a full block of experimental questions.
 
Probably <2 wrong per block for >270, <3 wrong per block for for >260. Just based on NBME data. There are of course rumors about experimental questions that get "thrown out" and if those rumors are true then the NBME relationship isn't going to be accurate. A friend of a friend knows someone who writes for the USMLE and claims that any given exam is up to 15% new/experimental questions. For all we know the exam is actually only 6 blocks and everyone has a full block of experimental questions.

that is an interesting point... if there are that many experimental questions I'm assuming those are for data gathering purposes and thus are related to content outside of UFAP.
Maybe that is why people feel bad about the exam then those questions that they screwed up on got thrown away and they end up doing close to the NBMEs.
 
The experimental questions are not all difficult in nature. Some are easy questions, I believe, according to a resource via Google (search for "experimental Step 1 questions"). Sorry, I can't find the link at the moment. It did seem somewhat legit though--I realize how this looks--"ok, a random internet citation!"

Curious, has anyone who has received their scores recently done better (>5-7pts) than their nbme practice exams? I have seen very few thus far. Thanks all.
 
Totally depends, some high scores say their test was easy and marked only a few, so their curve was probably close to NBMEs. My test was crazy hard, pretty sure I missed a lot more and still got curved close to m NBME scores.
 
The experimental questions are not all difficult in nature. Some are easy questions, I believe, according to a resource via Google (search for "experimental Step 1 questions"). Sorry, I can't find the link at the moment. It did seem somewhat legit though--I realize how this looks--"ok, a random internet citation!"

Curious, has anyone who has received their scores recently done better (>5-7pts) than their nbme practice exams? I have seen very few thus far. Thanks all.

Got my scores back this week and did 25 points better than NBME 17, which I took about 8 days before my actual exam. Interesting.
 
NBME 17 seems to have garnered a reputation as a tough exam. I still haven't seen more than a handful that improved vs their NBME averages.

Is the consensus also that the free 132 were easier relative to the NBME practice exams? Anyone feel significantly different? Thanks again.
 
NBME 17 seems to have garnered a reputation as a tough exam. I still haven't seen more than a handful that improved vs their NBME averages.

Is the consensus also that the free 132 were easier relative to the NBME practice exams? Anyone feel significantly different? Thanks again.

Free 138 was way harder for me.
 
The experimental questions are not all difficult in nature. Some are easy questions, I believe, according to a resource via Google (search for "experimental Step 1 questions"). Sorry, I can't find the link at the moment. It did seem somewhat legit though--I realize how this looks--"ok, a random internet citation!"

Curious, has anyone who has received their scores recently done better (>5-7pts) than their nbme practice exams? I have seen very few thus far. Thanks all.
I got a 245 on NBME 17 (my last full length, timed NBME; also my highest NBME score) and got a 252 on the real thing. I was also taking all the NBME tests in 50q blocks throughout my study period (but not as one continuous 200q exam) and my scores ranged in the 240s. I never hit 250 on any NBME and still broke it, and hope that gives encouragement to anyone out there who hasn't been able to break it on a practice test yet.
 
Top