UW Questions with <20% user correct.. a mathematical improbability? You decide.

[Re3iRtH]

---

Members don't see this ad.

As I was reviewing the answers to my latest block of questions I was
thinking... There are tons of questions in UW with less than 20% of users
having gotten it correct.

Now, most of these questions have only 5 answer choices. So theoretically,
even if not a SINGLE user knew the concept, and from just sheer
guessing, the percentage gotten it correct should be over 20%. Now,
there are tons of smart people doing the qbank, so one would think the
percentage should be even higher than 20%.

Which leads me to think, are these questions legit/high-yield for the
Step? I mean, when I read these questions, they don't seem poorly worded
or ambiguous like some Kaplan qbank questions I've seen. So why is it that
even with 5 answers choices I am seeing 14% and 17% all over the place
(see reasoning above)?

 

[pseudoknot]

Questions have distractors, and people are not choosing answers randomly. The hardest UW questions probably have a wrong answer that looks right unless you understand it.

 

[joeylux]

Disclaimer: huge brag post, so skip it if you want.

I had a UWORLD question yesterday that 9% answered correctly (pretty sure that's the lowest I've seen) and I knew the answer. It wasn't even that hard was the weirdest part - it involved p450 induction and a sulfonamide cephalosporin. Sometimes I wonder how such a low % is possible but then I remember that many of my errors come on questions that 80% of UWORLD seems to know so I really have no idea.

 

[Re3iRtH]

Disclaimer: huge brag post, so skip it if you want.

I had a UWORLD question yesterday that 9% answered correctly (pretty sure that's the lowest I've seen) and I knew the answer. It wasn't even that hard was the weirdest part - it involved p450 induction and a sulfonamide cephalosporin. Sometimes I wonder how such a low % is possible but then I remember that many of my errors come on questions that 80% of UWORLD seems to know so I really have no idea.

Which leads me to believe that these distracters pseudo is referring to,
are they really distracters? I think I remember that p450 fact about
cephalosporins... I mean, 9%!! Hmmz. Even if the smart people saw the
distracters, maybe they picked the wrong answer from it. What about
most people who had no idea, why are they picking the wrong answer
that consistently?

 

[harrisonIM]

i use the % correct to gauge the questions quality. if reading the explanation it seems like the question was testing minutiae and less than 20% got it right i'll just not annotate the facts into first aid. there are already more facts in first aid then i can actually remember. no need to add more facts from UW that aren't high yield.

 

[nope80]

Some of the questions are really poorly worded - I tend to notice it more when the questions have low correct percentages..maybe I'm wrong though.

 

[DrGregoryHouse]

a sulfonamide cephalosporin

Umm what?

 

[sebsvenmdc]

As I was reviewing the answers to my latest block of questions I was
thinking... There are tons of questions in UW with less than 20% of users
having gotten it correct.

Now, most of these questions have only 5 answer choices. So theoretically,
even if not a SINGLE user knew the concept, and from just sheer
guessing, the percentage gotten it correct should be over 20%. Now,
there are tons of smart people doing the qbank, so one would think the
percentage should be even higher than 20%.

Which leads me to think, are these questions legit/high-yield for the
Step? I mean, when I read these questions, they don't seem poorly worded
or ambiguous like some Kaplan qbank questions I've seen. So why is it that
even with 5 answers choices I am seeing 14% and 17% all over the place
(see reasoning above)?

As someone pointed out, this logic only applies to randomly picking answers. You would think that a knowledge base would be superior to random picking, but it is not if there are incorrect choices that are very tempting to the majority of students. Thus, it's actually possible for the percent correct to be in the single digits in terms of percent.

 

[Ihateverbal]

I noticed a bunch of pedigree Genetics questions + calculating probabilities in Genetics got a ton of 20% correct...

People must suck at Genetics or something?

 

[Captopril]

Disclaimer: huge brag post, so skip it if you want.

I had a UWORLD question yesterday that 9% answered correctly (pretty sure that's the lowest I've seen) and I knew the answer. It wasn't even that hard was the weirdest part - it involved p450 induction and a sulfonamide cephalosporin. Sometimes I wonder how such a low % is possible but then I remember that many of my errors come on questions that 80% of UWORLD seems to know so I really have no idea.

WTF is a sulfonamide cephalosporin?

 

[psychforme]

WTF is a sulfonamide cephalosporin?

THAT IS WHAT I WONDER AS WELL WTF??? I am doomed to do FP for the rest of my life or what the ??? I thought cephalosporins where beta-lactam antibiotics oh my god

 

[joeylux]

WTF is a sulfonamide cephalosporin?

Turns out I may have gotten that question right as a guess... by "sulfonamide-cephalosporin" (which doesn't exist) I was trying to say a cephalosporin with a sulfa group that may induce P450 (the S in SIC KEG). I understand that sulfonamides are different (SMX, sulfonaureas, celecoxib, etc) but for some reason I thought some cephalosporins were sulfas as well, I apologize, was the night before the USMLE.

Cephalosporins that include a methylthiotetrazole group (cefamadole, cefmetazole, etc) can cause hypoprothrombinemia and bleeding disorders - maybe some how my brain converted that to some sort of warfarin interaction (I know it doesn't make sense).

Point being, I guessed and got a question right that 9% of UWORLD did the same.