USMLE Official 2017 Step 1 Experiences and Scores Thread

Your USMLE Scores Increased, Guaranteed!
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
WeedForLunch

WeedForLunch

Full Member
7+ Year Member
Joined
Nov 26, 2015
Messages
51
Reaction score
25
I know this is quite early but most American Students have finished giving the test for this year.
I am an IMG and have been prepping for the steps since quite some time and have seen Phloston, Transposony's and others' threads for their respective years and how helpful they have been.

I intend on giving step in Jan.. let's share timetables, plans and other stuff on how everyone intends on taking on this beast.

P.S. : I think it is not that early.. the 2015/2016 threads were started in September/October.. but in true SDN gunner style..i wanna start it in August.. :)
Members don't see this ad.
 
  • Like
Reactions: 10 users
Sort by date Sort by votes
do NOT avoid NBME 19!! why avoid the most recent NBME? just don't take it last (i.e. take 19 before 18) so that your confidence is still decent before the test. Just tell yourself that the form under-predicts
 
Upvote 0 Downvote
Members don't see this ad :)
LLCoolK said:
do NOT avoid NBME 19!! why avoid the most recent NBME? just don't take it last (i.e. take 19 before 18) so that your confidence is still decent before the test. Just tell yourself that the form under-predicts
Click to expand...
He said he already took 18. Nbme 19s curve is wicked and mess with people. if hes got time then ya but 2 weeks, that a crunch if hes gonna get through the other ones
 
Upvote 0 Downvote
Dr.Jekyll75 said:
avoid 19 everyone says 16s good. UWSA 2 if you wanna go in with confidence
Click to expand...
UWSA2 will not help with confidence for everyone. Personally, I got a 200. Really messed with me. Two days later I did a couple blocks of UWSA1 and got an 80% average. Granted, things were off for me when I took UWSA2, I don't think it's safe to say it will make you feel great.
 
Upvote 0 Downvote
chromuffin said:
I've seen that 80 = 240, 90 = 250 but that's about it. Notably, the R^2 was like 0.24
Click to expand...
Keep in mind that the usual R-square doesn't do much in the way of telling us about predictive utility. It more so details how the model explains the in-sample variation of the dependent variable. Other statistics like predicted R-square can be found with various methods such as hold-out sample, k-folds, or jackknifing to give us a better indication of out-of-sample performance, which is what we really care about. The model standard deviation (estimate of the sd(errors of prediction)) is also useful if available, but can also benefit from out-of-sample methods. I think this is one of the major issues with all of these "analyses" floating around on the net, but there are other more important problems that are probably unaddressed. Namely, people are looking for prediction, and no one seems to actually be into making a good attempt at building and validating that kind of model. The current analyses aren't useless, but they have much room for improvement.

Dr.Jekyll75 said:
hm so what about 18 then, i heard 18 and 19 were similar
Click to expand...
I thought 18 was more challenging and the curve was more forgiving. Form 18 had me less sure of my answers during the exam, but it was my best NBME form in terms of a score. Form 19 had shorter stems and more straight forward questions, in my opinion, but I made a few extra dumb mistakes which cost me 20 points on the scaled score for about 4 more incorrect on the raw score (compared to form 18); form 19 was the 2nd lowest NBME I had, second only to my first during dedicated. I have seen many other people do slightly worse than other NBMEs and other people still kill NBME 19 (260+). The same percent correct gets you a lower scaled score, which is one of the few objective statements I can offer.

Army guy said:
Is nbme 19 difficult? or is it one of the easier nbme's that's why the scaling is messed up?
Click to expand...
Summarizing my above remarks, NBME 19 seems more difficult in that a given percent correct resulted in a lower scaled score for me and my friends who took it.
 
Last edited by a moderator:
Upvote 0 Downvote
yankees527 said:
anyone know the difference between gliosis (2 weeks in) vs gliotic scar (>1 month)?
Click to expand...

So gliosis would be the reactive process to damage where a gliotic scar would be after the fact. Probably just a fibrotic scar with very little reactive rocked going on.


Sent from my iPhone using SDN mobile
 
  • Like
Reactions: 1 user
Upvote 0 Downvote
dempty said:
Keep in mind that the usual R-square doesn't do much in the way of telling us about predictive utility. It more so details how the model explains the in-sample variation of the dependent variable. Other statistics like predicted R-square can be found with various methods such as hold-out sample, k-folds, or jackknifing to give us a better indication of out-of-sample performance, which is what we really care about. The model standard deviation (estimate of the sd(errors of prediction)) is also useful if available, but can also benefit from out-of-sample methods. I think this is one of the major issues with all of these "analyses" floating around on the net, but there are other more important problems that are probably unaddressed. Namely, people are looking for prediction, and no one seems to actually be into making a good attempt at building and validating that kind of model. The current analyses aren't useless, but they have much room for improvement.
Click to expand...
Not sure if you're just flexing your statistics knowledge or what, but I have no idea what you just said. Regardless, I simply stated the R2 for people to make their own judgement.
 
  • Like
Reactions: 2 users
Upvote 0 Downvote
Anybody have an idea of how high yield memorizing the chromosomes # the diseases are on is?
 
Upvote 0 Downvote
Members don't see this ad :)
chromuffin said:
Not sure if you're just flexing your statistics knowledge or what, but I have no idea what you just said. Regardless, I simply stated the R2 for people to make their own judgement.
Click to expand...
perhaps one of the best comments I've read on SDN in a while
 
  • Like
Reactions: 7 users
Upvote 0 Downvote
I thought I'd share a nice quote I got from a 4th year. He said during dedicated prep you should avoid being efficiently inefficient... that is so damn relatable hahaha


Sent from my iPhone using SDN mobile
 
Upvote 0 Downvote
chromuffin said:
Not sure if you're just flexing your statistics knowledge or what, but I have no idea what you just said. Regardless, I simply stated the R2 for people to make their own judgement.
Click to expand...
Not flexing, but trying to be useful. I'm pointing out that whatever the r-square is there, it doesn't matter a whole lot for our purposes, since people want out-of-sample prediction (the given number is purely in-sample). I'm also pointing out that even that reddit post, which has received a lot of praise, is pretty lacking in that it doesn't really live up to it's stated purpose either through incomplete work or failing to disclose information that would be helpful to the end user.

I asked this question a lot in the past when people kept asking for the correlation of scores, so it's similar to this case regarding an r-square value. What does that number mean to you in terms of prediction (you did say it was notable)?

There's a right way to interpret these statistics and there's limitations to each statistic, and unfortunately, medical school doesn't really teach this stuff well. I'm not trying to be difficult or anything like that, just helpful as I mentioned earlier. Sorry if it comes off as the former.
 
Last edited by a moderator:
Upvote 0 Downvote
Brorthopedic said:
perhaps one of the best comments I've read on SDN in a while
Click to expand...
Candidly, I think it's a funny reply, but I'm not sure why it seems like my post is out of place in a discussion of predictive value and citing certain statistics used. My main point was the value given, which I believe was from reddit, should be taken very lightly and doesn't really help anyone much. I wasn't a traditional premed, so my background is a bit different, but I don't see my contribution as out of left field in this context.
 
Upvote 0 Downvote
dempty said:
Not flexing, but trying to be useful. I'm pointing out that whatever the r-square is there, it doesn't matter a whole lot for our purposes, since people want out-of-sample prediction (the given number is purely in-sample). I'm also pointing out that even that reddit post, which has received a lot of praise, is pretty lacking in that it doesn't really live up to it's stated purpose either through incomplete work or failing to disclose information that would be helpful to the end user.

I asked this question a lot in the past when people kept asking for the correlation of scores, so it's similar to this case regarding an r-square value. What does that number mean to you in terms of prediction (you did say it was notable)?

There's a right way to interpret these statistics and there's limitations to each statistic, and unfortunately, medical school doesn't really teach this stuff well. I'm not trying to be difficult or anything like that, just helpful as I mentioned earlier. Sorry if it comes off as the former.
Click to expand...

I interpret R2 as a poorly conducted study result. I'm not going to act like I know the ins and outs because I don't. You were definitely helpful. Not trying to bash you!
dempty said:
Candidly, I think it's a funny reply, but I'm not sure why it seems like my post is out of place in a discussion of predictive value and citing certain statistics used. My main point was the value given, which I believe was from reddit, should be taken very lightly and doesn't really help anyone much. I wasn't a traditional premed, so my background is a bit different, but I don't see my contribution as out of left field in this context.
Click to expand...
Your contribution was great! It was just littered with jargon and hard to interpret without a stats background. There was no hard feelings.
 
  • Like
Reactions: 1 users
Upvote 0 Downvote
dempty said:
Candidly, I think it's a funny reply, but I'm not sure why it seems like my post is out of place in a discussion of predictive value and citing certain statistics used. My main point was the value given, which I believe was from reddit, should be taken very lightly and doesn't really help anyone much. I wasn't a traditional premed, so my background is a bit different, but I don't see my contribution as out of left field in this context.
Click to expand...
To be fair, I had to reread your post just to truly digest the information haha. Don't take it as a bad remark, if anything it's a compliment. It felt as if I was reading a Uworld question all of the sudden haha, trying to decipher it and all
 
  • Like
Reactions: 1 users
Upvote 0 Downvote
FMSIIM1 said:
Congrats on completing the test. I know you were in a bit of a crunch at the end. Were you able to take another NBME after UWSA1? How far did you manage to get on UWorld?
I think my scores are close to yours and just wanted some perspective. I test Friday.
Click to expand...

Yeah, I really felt like I was. I ended up taking the weekend before off due to my fiance's graduation from her masters program and helping her move. Coming back the week of I was getting consistent mid-70s in UW, which was high for me. I got a 74 or 75% on the free 120. I only completed 60% of UW or so. Really regret that. But as long as I pass, oh well. I ended up getting a 210 on another NBME after UWSA1 of 215. I'm hoping I'll hit a 220 from the real deal. Best of luck to you on Friday, my man. A few of your comments really helped me clear my head and spoke some sense to me in the last week or two, so thank you.

chromuffin said:
Anyone have a good source for understanding the physiology behind RTAs?

Edit: specifically, why type 4 RTA has low urine pH if aldosterone deficient/defective..
Click to expand...

I would hit up Goljian.
 
  • Like
Reactions: 3 users
Upvote 0 Downvote
Are the NBME exams supposed to be an overestimate? People at my school make it seem that way, claiming NBME 15 in particular overestimates by ~15. Hit a 269 on both NBME 15 and 19. Wondering if I am about where I need to be to break 260 on the real thing.
 
Upvote 0 Downvote
chromuffin said:
I interpret R2 as a poorly conducted study result. I'm not going to act like I know the ins and outs because I don't. You were definitely helpful. Not trying to bash you!
Click to expand...
Understood! This where I find it's helpful to ask those kinds of questions (what it meant to you). Don't let the r-square guide you down path of "good" or "bad" model or study judgements, because it doesn't really give you that kind of insight. The study and model might be excellent, but a low r-square can generally be interpreted as telling us that the independent variable(s) we are using in the model isn't "explaining" a whole lot of the movement in the dependent variable (other things we failed to include can probably improve the model from this point of view). Conversely, even if a study had an R-square of .99 (99%), there are many reasons why that could be the case aside from the reason we hope for, but it still could be an incredibly poor study. One final thing to note is that GPAs and test scores are usually difficult to obtain r-square values close to 1, so a "good" value is typically relative to the previous studies in the given area of research.

Long story short: keep r-square simple. If it's not a "predicted r-square" (which implies some kind of out-of-sample performance assessment), then it's just telling us how well we can explain the sample data on the dependent variable with that particular model (which doesn't let us know much about how it does with "new" data).


chromuffin said:
Your contribution was great! It was just littered with jargon and hard to interpret without a stats background. There was no hard feelings.
Click to expand...
The jargon is there in case someone wants to look it up for their own use, but it's also somewhat unavoidable when naming specific methods without a paragraph to describe them (which I did feel that descriptions weren't really necessary at this point, but I see what you mean about it being less accessible for casual reading). The 3 examples I summarized were ways to get "out-of-sample" performance, meaning how does our model work for prediction outside the data set used to make the model...is it valid on "external data"? But, those terms would be something for an interested reader, which I have had before (mainly through PMs with a few people).


Brorthopedic said:
To be fair, I had to reread your post just to truly digest the information haha. Don't take it as a bad remark, if anything it's a compliment. It felt as if I was reading a Uworld question all of the sudden haha, trying to decipher it and all
Click to expand...
I apologize for the UWorld flashback :D...Thanks to both of you for clarification. I have had unfortunate cases where researchers insist similar things are irrelevant and can be omitted or ignored because "this isn't statistics, it's X field" (which is where I thought this might be going). It's kind of like a chemist saying that physics is unimportant because he is a chemist :thinking:.
 
Upvote 0 Downvote
Hey so I need my score back by June 28. Does anyone know when I have to take the test by for that to happen? I know theres a score delay right now but does that mean that it goes back to the regular 3/4 wednesdays beginning in june?
 
Upvote 0 Downvote
JD_waynestudent said:
Are the NBME exams supposed to be an overestimate? People at my school make it seem that way, claiming NBME 15 in particular overestimates by ~15. Hit a 269 on both NBME 15 and 19. Wondering if I am about where I need to be to break 260 on the real thing.
Click to expand...

When you're scoring that high, it's hard for any NBME to underestimate. You're in very good shape to break 260 on the real deal.
 
Upvote 0 Downvote
doc-Oc said:
does nbme 15 actually overpredict? was gonna take it soon and didn't want any surprises.
Click to expand...

No. From what I've seen of those posting their NBMEs and their score on the real deal, NBME 15, 16, and 18 are all fairly accurate, and tend to underpredict (if anything). With NBME 13, 17, and 19, the curves are very tough and seem to underpredict by 5 to 10+ points. The thing is when you start scoring 250+ on these NBMEs, you are getting very few questions wrong and luck begins to play a bigger role in getting a test that plays into your strengths (since only missing a couple more questions separates a 250 from a 260).

It also depends on how close you take the said NBME before your test date. If you take NBME 15 four weeks from your test date, then yeah, your NBME 15 is likely to underpredict by a good amount since you have 4 weeks of additional studying under your belt by the time you take the real thing.
 
Upvote 0 Downvote
chromuffin said:
On this topic,
RBBB: bunny ears in V1 with S waves in I, V6
LBBB: inverted bunny ears V1
Both: prolonged QRS
Seem about right? I forget these all the time
Click to expand...
someone can correct me if I'm mistaken but I think asking an M2 to recognize L or R BBBs is out of scope for Step 1? I was under the impression that the ones in FA is as tough as it gets
 
Upvote 0 Downvote
Gorne said:
someone can correct me if I'm mistaken but I think asking an M2 to recognize L or R BBBs is out of scope for Step 1? I was under the impression that the ones in FA is as tough as it gets
Click to expand...
Maybe its just important in terms of wide and paradoxical splitting?
 
Upvote 0 Downvote
For those that have taken the exam, what pharm sections would be high yield for someone low on time? GI, cards, renal, repro, autonomics?
 
Upvote 0 Downvote
chromuffin said:
For those that have taken the exam, what pharm sections would be high yield for someone low on time? GI, cards, renal, repro, autonomics?
Click to expand...

Although I haven't taken the exam yet...it seems like autonomics and cardio would be the highest yield...followed by renal.


Sent from my iPad using SDN mobile
 
  • Like
Reactions: 1 user
Upvote 0 Downvote
Respiratory as well. Lots of chances to combine resp with cardio and lotsa bugs.

In other news, my UWorld %s just tanked 10% out of nowhere. That and the variation is all over the place. I've scored as low as 45% and as high as 80% with blocks of 20-40 questions... The 'average' for the questions wasn't all that different between them. Hoping it's just a blip and I'll see changes as I keep going on. This is definitely not fun though...
 
  • Like
Reactions: 1 users
Upvote 0 Downvote
docbsb2015 said:
Is it just a massive pool of questions and everyone gets just a random selection of those questions?
Click to expand...
That's my guess.

chromuffin said:
For those that have taken the exam, what pharm sections would be high yield for someone low on time? GI, cards, renal, repro, autonomics?
Click to expand...
The sections you're weakest in. Honestly, there was a little bit of everything and not heavy in anything in particular for me.
 
  • Like
Reactions: 1 users
Upvote 0 Downvote
ridethecliche said:
Respiratory as well. Lots of chances to combine resp with cardio and lotsa bugs.

In other news, my UWorld %s just tanked 10% out of nowhere. That and the variation is all over the place. I've scored as low as 45% and as high as 80% with blocks of 20-40 questions... The 'average' for the questions wasn't all that different between them. Hoping it's just a blip and I'll see changes as I keep going on. This is definitely not fun though...
Click to expand...

I had a week where I was averaging low 70s, then one day I got two 50's on my blocks for the day. Doesn't feel good, man (but don't let it phase you, and I know how hard it is to take that advice)
 
Last edited:
Upvote 0 Downvote
yankees527 said:
Literally cant study anymore for this test :hurting:

Friday cant come soon enough
Click to expand...
I know how you feel. I'm 10 days out from test day and I'm done with my 2nd pass of UWorld and complete pass of Pathoma. I have no idea what else to do. Can't.......read......First......Aid.......anymore.
 
  • Like
Reactions: 1 user
Upvote 0 Downvote
chromuffin said:
Wish I felt the same. Still need to review pharm, immunology, viruses, maybe some neuro.
Click to expand...
I feel like what i know by now i know, and if i dont know it....i tell myself it wont be on the exam :p
The Dayman said:
I know how you feel. I'm 10 days out from test day and I'm done with my 2nd pass of UWorld and complete pass of Pathoma. I have no idea what else to do. Can't.......read......First......Aid.......anymore.
Click to expand...
I dont ever want to open up that book anymore (besides when i read that rapid review thing tomorrow at the back) o_O
 
Upvote 0 Downvote
The Dayman said:
I know how you feel. I'm 10 days out from test day and I'm done with my 2nd pass of UWorld and complete pass of Pathoma. I have no idea what else to do. Can't.......read......First......Aid.......anymore.
Click to expand...

Why don't you take your test on friday and let me use the extra week? We can mess with the space time continuum to make that happen right?
 
  • Like
Reactions: 1 user
Upvote 0 Downvote
You must log in or register to reply here.
Next unread thread

Similar threads

slowthai
  • Question
Official 2021 Step 1 Experiences and Scores Thread
9 10 11 12 13
Replies
624
Views
58K
DrStephenStrange
DrStephenStrange
E
  • Question
USMLE Official 2020 Step 1 Experiences and Scores Thread
94 95 96 97 98
Replies
5K
Views
235K
ExplodingUlcers
ExplodingUlcers
A
  • Question
COMLEX Official 2019 Comlex 1 Experiences and Scores Thread
23 24 25 26 27
Replies
1K
Views
214K
Ellomate
E
libertyyne
  • Question
USMLE Official 2019 Step 1 Experiences and Scores Thread
57 58 59 60 61
Replies
3K
Views
277K
STUW18
S
S
  • Question
Communication and Experiment type questions on Step 1
Replies
0
Views
830
shoenberg3
S
Top