"repeated" questions within exam?

[Fermi]

I posted this in an unrelated thread, so I'm creating a new one here to solicit responses from those with the similar experiences:

On my Step 1, taken 7/7, I had at least two questions that were, for all intents and purposes, repeated later on in the exam! They weren't verbatim, but they were asking the exact same thing with slightly different wording and the answer choices were similar as well. I'm not talking about questions on the same disease or biochemical pathway--I guess USMLE policy prohibits me from saying anything specific about the questions, but they were both biochem/mol bio. I actually didn't know either of the questions at first, but the cool thing was that when I saw one the second time, only two of the answer choices were the same as before, so I could narrow it down immediately! BTW, I'm happy to report that I looked them up afterward and ended up getting all 4 right!

---

Members don't see this ad.

 

[LoneSEAL]

Yup same here. I had a repeated biostats question. Everything was the same except a number was changed.

 

[docuw]

I had three questions about Ondasteron and 5HT3 receptors - all said the exact same things with maybe a change in the order of the sentance. Two of the questions were in the same block!

 

[JP2005]

keep on dancing !!!!

 

[Jaded Soul]

I got 2 of the same biostats questions with just the numbers changed.

Doesn't matter how easy the questions are, nothing should be repeated like that on the exam because it's double jeopardy.

 

[BigBadBix]

i had three questions on mechanism of action of digoxin...very similar wording on all of them.

it makes no sense to me that they don't standardize this test more. we have to know all the systems equally, so why not test them equally? weird.

 

[LoneSEAL]

A friend of mine actually had a question where a word was misspelled. I can't fathom why questions should repeat. I wonder what their algorithm is for picking questions. Do they have say 10 different forms or is each question generated randomly from a pool? And if the NBME people do make a mistake, no one would ever know...

Also, I have to echo another poster's remark on how inherently fair this exam is since in talking with other people, people are getting tested on totally different things. I guess that's the nature of the beast of such a computerized testing format. But in the end, I believe knowledge is important but also if you want to ace, you better be lucky too. Of the 500+ drugs, you better hope that the 30 drugs they're testing you are the ones you've studied for.

 

[mvalento]

i had two nearly identical questions on Angelman / Prader-Willi...good stuff

bud

 

[TulaneKid24]

I got four, yes 4, questions that had to do with the loop diuretic furosemide! Good think I knew all about that, otherwise I would have been screwed!

I think that the NBME tries to randomize questions, but there will always be some repeats as randomization is never 100% perfect.

 

[Kosmo]

I have a somewhat different take on this. I too had a few questions that seemed to be virtual repeats, like having to do with the side effects of a specific anti-neoplastic drug. A friend of mine noted that he had five questions based on Lineweaver-Burke plots on his test, I didn't have a single one.
My guess is that while the questions are randomized somewhat, it is likely that repeated questions represent, to some extent, the experimental questions that are inevitably part of the exam. Otherwise these repeats could give an inaccurate representation of one's overall grasp of the material -- either positively or negatively with respect to the examinee. So by asking nearly identical questions to the same person, the quality of these questions can be gauged for use on future tests.

Anyhoo, that's just my theory.

 

[8744]

Originally posted by Kosmo
I have a somewhat different take on this. I too had a few questions that seemed to be virtual repeats, like having to do with the side effects of a specific anti-neoplastic drug. A friend of mine noted that he had five questions based on Lineweaver-Burke plots on his test, I didn't have a single one.
My guess is that while the questions are randomized somewhat, it is likely that repeated questions represent, to some extent, the experimental questions that are inevitably part of the exam. Otherwise these repeats could give an inaccurate representation of one's overall grasp of the material -- either positively or negatively with respect to the examinee. So by asking nearly identical questions to the same person, the quality of these questions can be gauged for use on future tests.

Anyhoo, that's just my theory.

You are too cool.

 

[Lara]

Just an idle question - is Step 1 similar to the MCAT in that a given number of students have the exact same exam, and those are who you are compared to when scores are given out? I'm finding it a bit disturbing that there's been such a vast discrepancy between exams in terms of which subjects are emphasized. Of course one should study everything as best as possible, but there almost seems to be too much "luck" involved. Or am I just being neurotic here?

In any case, back to summer vacation and congrats to all who have this behind them

 

[Fermi]

Originally posted by Lara
is Step 1 similar to the MCAT in that a given number of students have the exact same exam, and those are who you are compared to when scores are given out?

I'm not sure how the 350 questions are picked. They might be randomly picked and ordered individually, or there may be 50-question blocks, of whihc 7 are picked for any particular student. I only mention the latter because I noticed, as I was taking my exam, that the status/time-remaining counter marked my test blocks as complete in a non-sequential order. I don't know what was up with this.

I do know that they weight each individual exam based on its overall difficulty, though, so there is some normalization.

 

[Fermi]

Ha! I found what I was looking for. The NBME website (here) says:

"Blocks of items are constructed to meet specific content specifications. As a result, the combination of blocks of items creates a form of the Step that is comparable in content to all other forms."

So it's not a random selection of 350 questions--it's a random selection of 7 50-question blocks, and the questions don't vary within the blocks. This agrees with my observation that my test blocks were marked "complete" in a random order in the status window. Now we know...

 

[Maire]

Last year I actually had a question with mixed-up categories. The pH range was 22-28 and the HCO3 range was 7.2 to 7.6! The worst part was there was no way to report the problem - the Prometrics people just shrugged their shoulders when I asked.

Maire

 

[kcrd]

Originally posted by Fermi
Ha! I found what I was looking for.

Good find, Fermi!

Since we're bitching about usmle questions, I actually had a grammar/syntax error in mine. It was a neuro question, and it described a patient who had motor deficits in her "right arms and legs." Very Kafka-esque.

 

[Kosmo]

Originally posted by kcrd
...I actually had a grammar/syntax error in mine. It was a neuro question, and it described a patient who had motor deficits in her "right arms and legs." Very Kafka-esque.

Were they talking about Siamese twins?

 

[Anasazi23]

Try taking the COMLEX. On my Step 1 COMLEX, there were at least 10-15 questions about male erectile dysfunction EACH DAY (the COMLEX is almost 900 questions spanning two full days of testing).

I won't even broach the topic of misspellings in the COMLEX. You're lucky if the question has correct grammar.



*checks for spelling mistakes*

 

[Blade28]

I had that happen to me too - no less than 7 questions all asking about polycystic ovarian syndrome.