less than 60% needed to PASS?

[Upstairsmamma]

I too have heard from numerous sources that to pass USMLE or COMLEX one must achieve ~60% of the (non-pilot) questions correct. However, based on Kaplan Q bank this “60%” is puzzling. Of the Q bank questions I’ve done, and I’ve done almost all of them, the Other Kaplan Users’ Overall Correct is 61%.

While I realize that Q-bank questions certainly differ from the real test questions, and percentages are not intended to be true indicators of actual test performance, it seems that there would at least be some correlation. Are people just answering a higher percent of questions correctly on the real thing compared to Q bank?

If the overall Kaplan AVERAGE is 61%, then (assuming Kaplan is accurate) on the real test wouldn’t ~60% earn you an AVERAGE score--215 or 500? Moreover wouldn't the score needed to merely PASS be much lower than average, say like 50% correct?

What do you guys think?

---

Members don't see this ad.

 

[lord_jeebus]

Are people just answering a higher percent of questions correctly on the real thing compared to Q bank?

Yes

 

[Law2Doc]

If the overall Kaplan AVERAGE is 61%, then (assuming Kaplan is accurate) on the real test wouldn't ~60% earn you an AVERAGE score--215 or 500? Moreover wouldn't the score needed to merely PASS be much lower than average, say like 50% correct?

You have to realize that the overall Kaplan average is meaningless. You have people taking tests to see where they are at before studying. You have people who score poorly early in their studying but will be better by the real test (but never quite bring the average back, or worse, switch to another question bank by the time they get good). You have people retaking sections in which they have already seen the answers once or twice. You have people doing Kaplan under timed, nontimed, or tutorial versions, only one of which is a fair simulation.

So there is no useful correlation from that 61% because that simply isn't what the average person scores when they are taking the test seriously, after having studied for 4-6 weeks, the first time seeing questions, under a timed situation. So no, the 61% average is not correlated to the average board score at all.

It is wishful thinking to think that if you are scoring a 50% on Kaplan, you will be at passing range. The medfriends site score estimator puts someone getting a 50% on Kaplan on course to score about 149 plus or minus 13. Sorry.

 

[Miami_med]

As an example, I started with scores around 50% and ended with scores over 70% on Kaplan. My overall average was 61%, but I was doing much better than that before I actually took Step I.

 

[Scottish Chap]

I too have heard from numerous sources that to pass USMLE or COMLEX one must achieve ~60% of the (non-pilot) questions correct. However, based on Kaplan Q bank this “60%” is puzzling. Of the Q bank questions I’ve done, and I’ve done almost all of them, the Other Kaplan Users’ Overall Correct is 61%.

While I realize that Q-bank questions certainly differ from the real test questions, and percentages are not intended to be true indicators of actual test performance, it seems that there would at least be some correlation. Are people just answering a higher percent of questions correctly on the real thing compared to Q bank?

If the overall Kaplan AVERAGE is 61%, then (assuming Kaplan is accurate) on the real test wouldn’t ~60% earn you an AVERAGE score--215 or 500? Moreover wouldn't the score needed to merely PASS be much lower than average, say like 50% correct?

What do you guys think?

I don't think an average score on Kaplan's Qbank means a predicted average score on Step 1. Keep in mind that a lot of people don't use Qbank in the way which allows it to reflect Step 1 (timed, mixed, unused mode). If you use Qbank in subject-based mode after you have studied that subject or in the untimed mode, your score is going to be higher. It's pretty much believed that a consistently average Kaplan score will allow one to comfortably pass Step 1. Of course, there are always exceptions at either extreme. Bottom line: use Qbank only as a teaching tool, and you'll do well.

 

[werd]

as a side note, ever notice that rumors say ~60% to pass step I, and that 182 is pretty much 60% of the 300 questions that are scored.... it would be kind of crazy if they just threw your raw score back at you

 

[godschosenone]

as a side note, ever notice that rumors say ~60% to pass step I, and that 182 is pretty much 60% of the 300 questions that are scored.... it would be kind of crazy if they just threw your raw score back at you

there's little doubt in my mind that that is exactly what they do (out of 300).

 

[manning18]

The NBME is A VERY secretive company. One of my frustrations having had trouble with an early Step is that I received virtually NO feedback. I had no idea how many items I answered right, how many wrong and how many were needed to pass. Some people say it's somewhere between 57-60% to pass, but others say 60-70%. The NBME is NOT a company whose policies I would try to predict, so just try to get as many as you can. The correlations between QBank, UW etc are OK, but can vary quite a bit.