90% correct on STEP 1 = 265+? If you're hoping for a high score, speculate here!

[Asp]

My working hypothesis, based on talking to some others, and my own observations on medical student ability, is that a 90% on STEP I translates to about a 265+.

Unfortunately I don't have enough data, so I'm asking for your input. If you scored 80%+ on UWorld (or equivalently, 265 on the UWorld practice tests), I'd like to know how many questions you thought you missed per section, on average, on STEP I.

I'm making this thread because I think some of the percentages being tossed around are unreasonably high. I think very high scores on STEP I do not require stratospheric percentages correct (the curve is more generous than many think). If this proves correct, hopefully this can prove assuring to some people who are trying to obtain high scores for competitive specialties.

My own scores on UWorld Qbank / UWSA1 / UWSA2 are 85% / 265 (85% correct) / 265 (90% correct). I feel like I missed about 10% of questions on the real test. I think I scored a 265+, but could see myself getting a 255+ or less likely, 275+ (one can always dream =)). Anyways, if you're also bored waiting for June 14, I'd like to hear your speculations.

----------------

95% = 275+
MY INTERPRETATION: Missed 2 questions per section. You knew your stuff and you have high test-taking IQ (MCAT Verbal = 15) and flawless calibration. 1/200.

90% = 265+
MY INTERPRETATION: You missed 4-5 questions per section. You knew your stuff and have high test-taking IQ (MCAT Verbal = 13,14), but your calibration is suboptimal -- you tend to over-think some questions (i.e. overestimate their difficulty and go for the "tricky" answer) and are your worst enemy. 2/100 achieve this; these scores aren't that rare at all, especially at top schools, where it's more like 5/100 achieve this.

85% = 255+
MY INTERPRETATION: You knew your stuff, but your test-taking IQ (MCAT Verbal = 12) is limiting you. I think the best advice to people at this level is to stop focusing only on minutiae. You need to focus on the meta-test, in order to score higher. That is, not just what the question is asking, but whether the question is a "hard question" or not, and why the question is being asked (what concept is being tested?). If you can predict the % correct and the main point in the Educational Objective in UWorld, you probably have good meta-test skills.

---

Members don't see this ad.

 

[gatoradedrinker]

chiming in about the MCAT VR & USMLE correlation.

while this too is somewhat anecdotal, i think it's better than the massive amounts of speculation circulating on here. the dean of my medical school told us that the strongest/best/(only?) correlating factor to step 1 scores is your MCAT BS score, not MCAT VR.

 

[knuckles]

To be honest, I think people trying to correlate these things based on scores in this forum, will be hard pressed to find causation. The selection bias of this forum is too extreme. (God, I feel like I've studied too much biostats.)

I'm sure when a much larger sample is considered, you'll find the MCAT VR (and possibly BS) are pretty good indicators. I think the VR is more of a general IQ/reading comprehension/test taking ability indicator, but of course if you do well on BS, it shows you're good at the sciences and that could prove beneficial too. All things considered, I know the Step 1 can be studied for, and it doesn't take a genius to get a great score. Hard work really does pay off for this test, but being really smart doesn't hurt either.

 

[badasshairday]

the dean of my medical school told us that the strongest/best/(only?) correlating factor to step 1 scores is your MCAT BS score, not MCAT VR.

I believe this. It is definitely true for me.

VR 8, BS 13

Step 1 = 250

 

[slowbutsteady]

chiming in about the MCAT VR & USMLE correlation.

while this too is somewhat anecdotal, i think it's better than the massive amounts of speculation circulating on here. the dean of my medical school told us that the strongest/best/(only?) correlating factor to step 1 scores is your MCAT BS score, not MCAT VR.

Stop worrying about correlation and STUDY!!! That is the only way to conquer this beast!!!

 

[2012mdc]

Stop worrying about correlation and STUDY!!! That is the only way to conquer this beast!!!

`
True, but some ppl need to know how to study the right way. I wish I would have spent a little bit more time improving my test taking ability b/c near the end of my studying there wasn't a real knowledge gap and I was struggling to improve my score.

 

[VOP]

I believe this. It is definitely true for me.

VR 8, BS 13

Step 1 = 250

The Biological Sciences section most directly correlates to success on the USMLE Step 1 exam, with a correlation coefficient of .553 vs .491 for Physical Sciences and .397 for Verbal Reasoning.

http://www.ncbi.nlm.nih.gov/pubmed/12377692?dopt=AbstractPlus

was in an NIH study, and makes the most sense to me. I've looked at some practice board questions and they felt more like Mcat bio than anything else.

 

[DdrumbumD]

VR 9 BS 14

step 1 = 261

 

[gatoradedrinker]

VOP - thanks for the post.

i hope everyone reads what you cited.

mcat BS NOT mcat VR

 

[dozitgetchahi]

Congratulations on that score. However, you are not limited by verbal intelligence at 250. I actually have a couple friends who scored in the 250-255 range. They will be great doctors, far better clinicians than I, but one of them I know also had a low verbal MCAT.

Verbal intelligence becomes important in the jump from 255 to 265. It becomes almost the dominant factor past 270.

I think all this talk about "verbal intelligence" in the context of the USMLE is pretentious, pseudoscientific rubbish. I haven't taken the USMLE yet, but consider the following: I "only" got an 11 on the MCAT verbal but managed a 780 on the SAT I verbal section. Consequently, I doubt I'm "lacking" much of anything in terms of verbal intelligence. The SAT, after all, was considered an intelligence test that you "couldn't study for" for decades, and it still has far stronger correlations with formal intelligence tests than the MCAT or any other discipline-specific examination.

Feel free to attack me all you want, but unless you're able to produce some sort of bona fide evidence that what you're saying is true here I think you're talking out of your ass.

 

[kryptik]

^have to agree with you 100%, that whole verbal/step thing is such BS.

 

[inkerman]

I also did 10,000 questions. It's a very effective way of increasing your calibration, versus reading texts (i.e. Taus method). I would highly recommend this strategy for people who are not naturally great test takers.

If your verbal ability is sky high, you can get away with just focusing on knowledge. Reading texts is of course the fastest way to learn raw facts.

85% = 255+
MY INTERPRETATION: You knew your stuff, but your test-taking IQ (MCAT Verbal = 12) is limiting you. I think the best advice to people at this level is to stop focusing only on minutiae. You need to focus on the meta-test, in order to score higher.

Oh dear. I just think one starts to hallucinate things after one goes through all of ten thousand practice questions. At that point after weeks of doing hundreds of problems a day you start to see all kinds of meta-things that may or may not be a figment of your imagination. Perhaps it is like a higher state of consciousness. I certainly wish I had done more practice q's before I took the exam.

 

[Asp]

I think all this talk about "verbal intelligence" in the context of the USMLE is pretentious, pseudoscientific rubbish.

Verbal explains a non-trivial amount X of the variance in STEP scores, while having nothing to do with the subject matter of STEP. Think about that for a second.

On the one hand, it's surprisingly high if STEP I is just a "knowledge test." I think this X represents verbal ability -- i.e. general passage- or vignette-based test taking ability.

On the other hand, it's low enough that people can score highly despite having just mediocre test taking ability. Don't be fooled by SDN selection bias. Most people scoring 255+ have high MCAT Verbals as well.

A lot of the posts above are attacking strawmen.

 

[WellWornLad]

Don't be fooled by SDN selection bias. Most people scoring 255+ have high MCAT Verbals as well.

Assuming your hypothesis about verbal ability and step scores to be true, how would SDN membership bias that association?

FFIW, MCAT verbal = 11, step 1 = 259

 

[Asp]

A MCAT verbal of 11 is not low -- it puts you at 90th+ percentile.

A lot of people with actually low verbals (less than 9) who also got low STEP scores, however, probably wouldn't be posting. Hence the selection bias.

 

[2012mdc]

A MCAT verbal of 11 is not low -- it puts you at 90th+ percentile.

A lot of people with actually low verbals (less than 9) who also got low STEP scores, however, probably wouldn't be posting. Hence the selection bias.

After reflecting on how I could have done a little bit better and thinking back to how I felt after the test - I think more of the questions I got wrong were due to "what the hell are they asking/want???" vs "I know what they want, I just can't answer it".

I was not weak in verbal (11) but I feel that if I was in the 13-14 range it would have been possible for me to crack 270.

 

[gatoradedrinker]

Verbal explains a non-trivial amount X of the variance in STEP scores, while having nothing to do with the subject matter of STEP.

speculation.

Most people scoring 255+ have high MCAT Verbals as well.

speculation times ten.

 

[DrBowtie]

Did anyone take the extended form NBMEs? I think this is the closest thing we have to approximate or even give you a sense of the % correct -> score conversion.

Obviously somewhat flawed with fewer questions and no experimental questions but here is a data point to crunch on:

180/200 (90%) correct was 249 on NBME 6.

 

[2012mdc]

Did anyone take the extended form NBMEs? I think this is the closest thing we have to approximate or even give you a sense of the % correct -> score conversion.

Obviously somewhat flawed with fewer questions and no experimental questions but here is a data point to crunch on:

180/200 (90%) correct was 249 on NBME 6.

On the two extended ones I did
176/200 (88%) was 244 on 6
190/200 (95%) was 264 on 7

 

[physics junkie]

I'm definitely inclined to agree with you guys about verbal reasoning being the primary factor in all standardized tests.

I didn't know any biology for the MCAT(I didn't even know that humans have a 4 chambered heart ) and scored 11-13 on practice tests and 12 on the real deal. My VR was a 13.

On the free150/NBMEs I see the same verbal reasoning skills are a huge help. I read very fast and can pick out the key information to start thinking about from the writing style alone. The vignette authors are nudging you towards an answer choice--it is not simply a presentation of facts where you have to decide what is important.

For the record, I scored a 249 on my free150 before I "started studying".

 

[badasshairday]

VR probably plays a role. But still, a great score on Step 1 is achievable even with crappy VR abilities.

And for the record, my VR score was an 8, and I scored a 243 on the Free 150 "before" studying.

 

[kryptik]

step 1 is a strictly knowledge based exam, the more you read and retain the better you score, has little to do with verbal

 

[nshams]

I'm definitely inclined to agree with you guys about verbal reasoning being the primary factor in all standardized tests.

I didn't know any biology for the MCAT(I didn't even know that humans have a 4 chambered heart ) and scored 11-13 on practice tests and 12 on the real deal. My VR was a 13.

On the free150/NBMEs I see the same verbal reasoning skills are a huge help. I read very fast and can pick out the key information to start thinking about from the writing style alone. The vignette authors are nudging you towards an answer choice--it is not simply a presentation of facts where you have to decide what is important.

For the record, I scored a 249 on my free150 before I "started studying".

Hey physics junkie, can you break down this question for me:

A 30 year old overweight female is a school teacher who is known to be a strict disciplinarian in the classroom. She presents with angina pectoris, and coronary angiography reveals 75% narrowing of the anterior descending branch of the left coronary artery. Which of the following would be a major risk factor for coronary atherosclerosis in this patient?

A. Obesity
B. Type A personality
C. Diabetes Mellitus
D. Sedentary life style
E. Age

How does a vignette like this push you towards one choice? An explanation would be muchly appreciated

 

[kryptik]

interesting question, i know its not directed to me but i will go with age since thats the number one risk factor for cad in the general population

 

[aashkab]

she's 30 years old though. can't be age.

 

[buckets301]

pretty sure its DM. Btw I made an 8 on VR and a 268 on step 1. Yes, its anecdotal.

 

[Asp]

A 30 year old overweight female is a school teacher who is known to be a strict disciplinarian in the classroom.

...

A. Obesity
B. Type A personality
C. Diabetes Mellitus
D. Sedentary life style
E. Age

Classic distractor opening. Note how obesity (overweight), personality (strict disciplinarian), sedentary lifestyle (school teacher), and age (30 year old) are all directly mentioned in the first sentence. The odd one out is diabetes mellitus, so you should guess it even if you knew nothing.

 

[knuckles]

classic distractor opening. Note how obesity (overweight), personality (strict disciplinarian), sedentary lifestyle (school teacher), and age (30 year old) are all directly mentioned in the first sentence. The odd one out is diabetes mellitus, so you should guess it even if you knew nothing.

+1

 

[kryptik]

nshams do you happen to have the answer and explanation to your question?

 

[TexasTriathlete]

So should I just get a bunch of MCAT verbal reasoning prep books and study those or what? Got the step in a month.

 

[nshams]

it's DM - the question's out of Robbin's

DM is a significant risk factor forr early accelerated, and advanced atherosclerosis. If a premenopausal woman or a young man has severe coronary atherosclerosis, diabetes must be suspected a predisposing factor. "Soft" risk factors that can play a lesser role in the development of atherosclerosis include obesity, stress, & lack of exercise.

By the way Asp, that's a really good question style to know. Know any books/websites where I can pick up more of these helpful hints?

 

[knuckles]

So should I just get a bunch of MCAT verbal reasoning prep books and study those or what? Got the step in a month.

That could possibly be the biggest waste of time ever. You improve your question reasoning by doing more questions. Nothing more, nothing less.

 

[TexasTriathlete]

you might be right

 

[Asp]

Just answer 10,000 questions correctly. Understand why you got the ones you got wrong, wrong.

You don't need Goljan. You don't need First Aid. You don't need Robbins. You don't need BRS this or High Yield that.

10,000 questions. Do this, and you're guaranteed a score within 5-10 points of your maximal potential.

 

[princess peach]

Just answer 10,000 questions correctly. Understand why you got the ones you got wrong, wrong.

You don't need Goljan. You don't need First Aid. You don't need Robbins. You don't need BRS this or High Yield that.

10,000 questions. Do this, and you're guaranteed a score within 5-10 points of your maximal potential.

i can't really decide if this kid is serious. too ridiculous to be true and even more surreal that people actually listen. i would highly recommend that people do not do this.

 

[knuckles]

i can't really decide if this kid is serious. too ridiculous to be true and even more surreal that people actually listen. i would highly recommend that people do not do this.

Tell us how you got your 270

 

[buckets301]

I do agree to some extent that doing tons of questions is the best way to go (2nd only to studying hard during first two years of med school). I did somewhere around 5000-6000 questions between kaplan qbank, rx, and uworld. But, doing questions is how I study best so that just made sense to me.

 

[princess peach]

Tell us how you got your 270

lol by using many of the things asp just said not to use. did about 5000 questions over about 6 months time. beyond that you're heading into crappy useless non-boards style question territory. i did a second run-through of uworld that helped, but would have been useless without real books as well. but everyone's different i guess, do what works best for you.

 

[rem6775]

Tell us how you got your 270

You mean the one whose second post was how they got a 274 while saying they only got 79% on UW? Sounds reasonable to me....

 

[knuckles]

You mean the one whose second post was how they got a 274 while saying they only got 79% on UW? Sounds reasonable to me....

who knows, but I wouldn't be surprised if half the scores posted on this site are embellished or straight up lies. people still have good advice though.

 

[TexasTriathlete]

You mean the one whose second post was how they got a 274 while saying they only got 79% on UW? Sounds reasonable to me....

Classmate of mine I think did 79% on UW, and dropped a 260+ on the step.

 

[badasshairday]

Classmate of mine I think did 79% on UW, and dropped a 260+ on the step.

I believe it. I was around 71% overall and hit a 250.

 

[TexasTriathlete]

Everyone is going to be a little different though. I am sitting at 77% on World right now (my second pass through), but I don't think I'm nearly that good. I think I am reaping the benefits of tutor mode, and going system-by-system.

I think I'll get a little better feel for where I am once I do a diagnostic or two.

I don't think I'm in any danger of failing or anything, so that's nice. Goal is 240+, which I think is certainly within reach. Time will tell.

 

[knuckles]

Everyone is going to be a little different though. I am sitting at 77% on World right now (my second pass through), but I don't think I'm nearly that good. I think I am reaping the benefits of tutor mode, and going system-by-system.

I think I'll get a little better feel for where I am once I do a diagnostic or two.

I don't think I'm in any danger of failing or anything, so that's nice. Goal is 240+, which I think is certainly within reach. Time will tell.

I don't think 2nd time through is the same, with regards to predicting your score based on your average.

 

[rem6775]

People really focused on that second part I see. I agree it's possible to get only 79% correct and a 274, although I'd expect higher. (I had 85% correct and a 267.) The bigger point was that it was their second post.

 

[badasshairday]

People really focused on that second part I see. I agree it's possible to get only 79% correct and a 274, although I'd expect higher. (I had 85% correct and a 267.) The bigger point was that it was their second post.

I meant to add something about the point you made about it being their second post. There are two possible reason:

1) They are full of ****.

2) It shows the extreme selection bias of SDN.

 

[TexasTriathlete]

I don't think 2nd time through is the same, with regards to predicting your score based on your average.

No doubt. That's why I just bought Kaplan, and will be using it for a while so that I can put some distance between my second and third passes on World.

And I don't see why people bitch about Kaplan so much. I just did my first block on there, and I thought the questions are pretty good. Small sample so far, but I think the explanations are solid, and I don't see what people are talking about when they say that it is so nitpicky. Seems pretty straightforward to me. Not too different from World as far as I can tell so far.

 

[dreamix]

Can anyone predict what the USMLE 3 Digit Score would be with a 90% accuracy?

I know in terms of the practice NBME's, 90% = 250. Is this also the case for the real exam, or is the curve a little more "friendly" as someone mentioned earlier.

 

[dreamix]

I read the initial post couple of weeks ago, and I mistakenly thought that Asp was referring to 90% accuracy on UWorld for some reason...

Oops. my badd.

I guess no one knows the answer to my question...yet

 

[Asp]

A 90-95% on the real test is about a 270-275 range.

If a 90% on the real test were only a 250, then there's really not enough dynamic range to explain higher scores.

For instance, the high score in our class this year was 280. Even if 280 were the theoretical top score and corresponded to 100% correct (unlikely), a 95% couldn't be much lower than 270.

The best speculation I can offer is 90% = 270.

 

[dreamix]

A 90-95% on the real test is about a 270-275 range.

If a 90% on the real test were only a 250, then there's really not enough dynamic range to explain higher scores.

For instance, the high score in our class this year was 280. Even if 280 were the theoretical top score and corresponded to 100% correct (unlikely), a 95% couldn't be much lower than 270.

The best speculation I can offer is 90% = 270.

I see what you are saying, and I hope you are right.

I have no idea where I acquired this document from, but this is what I was following...