% correct to get a 240 on step 1?

[Quetiapine]

Original thread title was bad

This of course is all speculation and based on what people got before and how many they think they missed. What do you need to break a 240 (of course it's dependent on the difficulty of the exam but the % swing should be no more than 3-5%). 80-85% or higher?

---

Members don't see this ad.

 

[hyrule]

no way to know

 

[jfgavina]

80% first time

 

[winkleweizen]

im 98% sure its 95%

 

[Quetiapine]

no way to know

yea we know there's no way to know but does anyone know of anyone who guesstimate what % they got correct n their score, we can kinda get a ball park number maybe...

 

[britishmafia]

There's actually a formula for this.

Basically you need to take your target USMLE score and multiply it times the quantity (e^(i*pi)+1). This should equal a number that gives you the percent chance of knowing whether or not you can predict your USMLE score based upon % correct, based on that target USMLE score.

 

[jfgavina]

im 98% sure its 95%

ahahah don't freak him out

 

[Quetiapine]

There's actually a formula for this.

Basically you need to take your target USMLE score and multiply it times the quantity (e^(i*pi)+1). This should equal a number that gives you the percent chance of knowing whether or not you can predict your USMLE score based upon % correct, based on that target USMLE score.

didn't know md/phds students knew that e^(ix) = cos(x) + isin(x), i'm quite impressed, maybe you should give me an equation in the S domain, i'm more familiar with Laplace transforms, you can also give me the values of the system so the feedback is stable. I want the expressions/values for P, I, and D.

 

[winkleweizen]

Ok ill bite...ill guess somewhere bw 83-86. The NBMEs require about an 88 so I assume the real thing would be a bit tougher.

 

[britishmafia]

didn't know md/phds students knew that e^(ix) = cos(x) + isin(x), i'm quite impressed, maybe you should give me an equation in the S domain, i'm more familiar with Laplace transforms, you can also give me the values of the system so the feedback is stable. I want the expressions/values for P, I, and D.

haha I kid. Don't take offense.

Honestly, any answers you get would be wrong. It's so tough to say, especially since we barely know how the thing is scored, plus you have a problem with recall bias. It'll just come down to taking practice exams and figuring it out for yourself based upon "gut" feelings or whatever.

It's also like asking people how many questions they marked per block after hearing their score. You will get wildly different answers based upon test-taking strategies and errors in recall, plus it may be possible that a % correct on one test could give a 240 whereas in another it gives a 220.

Sadly, my scores don't come out until July 13th (took it on 5/23, so this is just awful). If you want, send me a PM at that time and I'll give you a sense of how many questions I was sure I got wrong and how many I narrowed down to 2-3 options. Even then, take it with a mound of salt, and don't believe for one second that what I'm reporting to you in any way corresponds to good data.

This is one of those "I really really wish there was an answer, but there just isn't" type of thing that plagues these exams.

 

[Quetiapine]

haha I kid. Don't take offense.

Honestly, any answers you get would be wrong. It's so tough to say, especially since we barely know how the thing is scored, plus you have a problem with recall bias. It'll just come down to taking practice exams and figuring it out for yourself based upon "gut" feelings or whatever.

It's also like asking people how many questions they marked per block after hearing their score. You will get wildly different answers based upon test-taking strategies and errors in recall, plus it may be possible that a % correct on one test could give a 240 whereas in another it gives a 220.

Sadly, my scores don't come out until July 13th (took it on 5/23, so this is just awful). If you want, send me a PM at that time and I'll give you a sense of how many questions I was sure I got wrong and how many I narrowed down to 2-3 options. Even then, take it with a mound of salt, and don't believe for one second that what I'm reporting to you in any way corresponds to good data.

This is one of those "I really really wish there was an answer, but there just isn't" type of thing that plagues these exams.

Yea I understand that too. Just being neurotic like some around here since I was in the 250-260 in practice tests and just praying for a 240+ now that I've counted from 25-30 known gimme mistakes (poor 50/50 luck and changing right to wrong on several occassions). Trying to see if anyone knew how many they got wrong (after looking it up) for sure and with a score they got, we can make some type of a guess (not accurate of course but seeing that you have a math/engr background, you can see where i'm going with the statistics and limits; someone who know they got like 10% wrong for sure, factor in the experimentals, etc and with their score we can guess a % correlate, etc).

I'm just being neurotic and in sadness since I feel like I prolly fell below my expected score and expectations.

 

[ATrim7]

I don't understand why there isn't more transparency in the scoring process of the test. It's not like it gives people who understand the scoring an unfair advantage.

 

[britishmafia]

Agreed.

I actually wrote a letter to the NBME about the one problem of not knowing after the test whether or not one had failed or passed. It could be possible to devise a method whereby students know immediately after the exam whether they had passed, failed, or were in some wide "borderline" region. This could actually add to the neurotic tendencies, but it could possibly help with those students that are pulled from the wards because they found out, a month later, that they failed the exam and have to take it again.

Just wanted to say that I completely agree with your desire to know this. It will really ease worries on test day as you estimate your performance block to block. The opposite could also be true. Hearing that a person who scored an x% received your target score and then NOT feeling confident you achieved that because you were given a "difficult test" or whatever could work against you.

The lack of transparency is very confusing, but we might be even more neurotic if we actually knew the formulae and methods they used.

 

[Sheldor]

I feel like they could modify this process a great deal to be kinder to us. Off the top of my head I don't see why they couldn't let us know our percent correct when we leave the center, along with a wide range of what the three digit score correlation could be. Heck, even don't include the range of score correlation, just knowing the percentage for a few weeks would help maintain sanity.

Also, based on the way they score posted in a different thread they should be able to tell us if we passed right away as well. They said that something about how every question counts the same towards the minimum needed to pass. Still, some transparency would be nice as the process seems quasi shady.

 

[Mattabet]

I absolutely agree with above that the process could be kinder to us.

The reason that these percentage estimates get tricky though, is the significant amount of unscored questions that are in the test. Although it would be unlikely, you could miss 20 (or more) questions and still get a perfect 280, as long as the questions you missed were new questions that were still in the process of being validated. (or study questions, or questions that are unscored for some other reason)

Furthermore, you're probably more likely to miss a question that still has some kinks in it than you are a tried-and-true scored question. And the only people with the information to untangle that whole ball of wax are sitting in their comfy NBME chairs at NBME central*.

Probably better to just say Hakuna Matata and get as many questions as you can.


*Nether chairs nor place actually exist.