The way scoring on USMLE STEP 1 most likely works is as follows:
All questions you get were given as experimentals several times to different pools of test takers. The statistics regarding %who got it right vs. %who got it wrong for those questions are used to compute a difficulty rating for each of the questions. On your exam you get say 280 questions. Say 40 are experimental. Therefore 240 count towards your score. Each of those questions has its own individual difficulty rating, computed by the test makers via the aforementioned technique. The difficulty ratings of all of your questions is averaged. That rating is used to generate a "curve" or standardization for your test. This is then used to finally compute your score.
Now your questions actually come from a pool of 20,000 questions. NBME essentially makes sure you get a nearly random set of 280 questions. However, to ensure that you get a certain number in every topic, all of the topics have a minimum number of questions that NEED be asked for them. For example, "pathology" probably has the highest minimum and something like biochem would have a lower minimum. Hope this makes sense
Also questions, where say under 15% got it right and there were only 5 choices, are probably thrown out. This is because it is hard to tell if people got it right because of luck or knowing the answer.
I agree with everything you said, however this is up for discussion. I think the people that really get all of those questions right end up with the super-duper stellar scores (268, 271, etc)
I respectfully disagree. In general, I strongly believe that the difference between the top scorers and those who are merely above average or even average is due to a disparity in the number of average and medium difficulty questions answered correctly. Maybe there is a slight rate of difference for the high difficulty questions, but I think that makes up a small fraction of the score differential. My school also told us this with regards to our in-house exams.
Anecdotally: I scored a 247. I counted 15 questions I missed for sure. Of those 15, I missed a few straight recall microbio questions and 9 anatomy questions. Of the 9 anatomy questions, only 1 was truly obscure. Among the other 8, half of them were directly in first aid. The only subjects I didn't get to finish on my crash 2 week review at the end were microbio and anatomy. I actually didn't look at anatomy at all during my dedicated study period and relied solely on my accumulated knowledge. BRS anatomy actually doesn't look all that bad, and I think if I had a bit more time, I could have burned through it and gotten many of the questions I missed correct.
Also, I realized a bit too late in the game that Sketchy wasn't a complete resource for me. I generally remember by connecting bugs via knowing in depth mechanisms of how, for example, they invade or colonize the body. While Sketchy had a lot of that, it just wasn't quite enough for me to deduce the right answer for different pathogenesis questions. Part of that, I think, is that I didn't have a lot of the small details sketchy had memorized. Others could deduce the answer to those questions via exclusively sketchy because they even knew the obscure factoid like things sketchy presented; I imagine this was because of their relatively more adept visual memory. My brain doesn't quite give me that luxury, despite having watched Sketchy about 5 times through slowly and twice with notes.
In general, my friends who scored in the 260s, showed a strong tendency towards reporting fewer mistakes on easy or medium questions, especially memory heavy but straightforward anatomy/microbio. Another trend among the top scorers seemed to be a huge level of familiarity with all of the study material. Of the five I know, four have graduate degrees of some sort in the biomedical sciences. The one outlier, claims to have a photographic memory. Three of them are junior AOA, but I know that all of them busted their butts during the first two years.
Finally, for what it's worth, biochem, genetics, and molecular bio were my worst subjects at the started of my dedicated study period. I became paranoid about those and studied my ass off. They ended up as the only subjects I starred in that subject breakdown subcategory. The systems subject I was most cocky about, repro+endocrine, I had the worst performance in. The subject I studied my absolute hardest for in school, despite not actually getting too much of a reward in terms of grades, neurology, I starred.
Side note: Apparently, I'm also an outlier because I loved DIT...
Anyway, I think my original point stands and my tangent has gotten me trying to address a different but equally important one: figure out how you study best!
)
