Test taking advice needed

[ProtonatedAlchy]

Hey all... quick advice would be very appreciated, test few days away, here's the score breakdown which has me baffled-

5 weeks out: NBME 7 - 228
4 weeks out: Uworld 1 - 245
3 weeks out: NBME 12 - 226
2 weeks out: Uworld 2 - 260
1 week out: NBME 13 - 224

Unfortunate progression here...am I just a huge idiot for sucking onthese NBMEs?? Wondering if I should just pound out more NBMEs this last week to get used to the questions as opposed to continuing with Uworlds "amazing" questions...anyone else had this happen to them?
Thanks all!

---

Members don't see this ad.

 

[ipizzy]

It seems like your NBMEs are all predicting around the same score, so I think that's what you should expect for your real deal at this point in your preparation. The UWSAs are a bit all over the place in terms of score prediction judging from others' experiences that they've shared on sdn. In terms of what you should do from here, I think you are the best person to answer that question. Do you feel like your test taking skills (timing, rapid question reading, etc) need work? Or do you feel like you miss questions because your knowledge base is lacking?

 

[rocketbooster]

An easy answer to your question.

The NBMEs are the only accurate scores compared to the real Step 1. UWorld overestimates your score by +15-20. These statements are supported by nearly every post on SDN. They are also supported by your results.

UWorld is a great learning tool, not a good indicator of your score progress. You should keep taking UWorld, though, because it will increase your overall Step 1 score more than NBME tests since the UWorld explanations are so good. UWorld = learning tool, NBMEs = Step 1 score indicator.

 

[herewego]

An easy answer to your question.

The NBMEs are the only accurate scores compared to the real Step 1. UWorld overestimates your score by +15-20. These statements are supported by nearly every post on SDN. They are also supported by your results.

UWorld is a great learning tool, not a good indicator of your score progress. You should keep taking UWorld, though, because it will increase your overall Step 1 score more than NBME tests since the UWorld explanations are so good. UWorld = learning tool, NBMEs = Step 1 score indicator.

UWSA overestimates by FIFTEEN to TWENTY? Browse through the 2012 experience thread, I've been following it for weeks. If taken relatively close to the exam they seemed to be within 5-8 points.

 

[rocketbooster]

UWSA overestimates by FIFTEEN to TWENTY? Browse through the 2012 experience thread, I've been following it for weeks. If taken relatively close to the exam they seemed to be within 5-8 points.

The ppl in that thread were saying +15-20, and every single person says UWorld for sure overestimates it. The take home message is UWorld definitely overestimates your score and only trust NBME scores, period. And even for NBME, you need to take more than one to make sure your NBME scores are consistent. Consistent NBME scores = best predicator it seems. If your NBMEs are like 245, 220, 235, 220, 240, then obviously there are some big gaps in your knowledge that are only exposed on certain tests. However, NBMEs of 245, 255, 240, 250, 260, etc. I would still say is consistent because once you get 240+ the curve gets much narrower. The curve between 220-240 is pretty big, though. 240=~mid 80th percentile, 220=50th percentile, 250=~90th+ percentile.

 

[usmleq]

The ppl in that thread were saying +15-20, and every single person says UWorld for sure overestimates it. The take home message is UWorld definitely overestimates your score and only trust NBME scores, period.

You are wrong. Here are just a few people posting in the 2012 experiences thread with very accurate predictions (and there are a ton more like this if you actually look):

Kaplan q bank = 68% finished before school ended (march 18)

UWorld 77% average

Cbsse = 235 8 weeks
NBME 13 254 4 weeks
Free 150 at site 90% 260
UWSA 1 265 3 weeks
NBME 12 254 1 week
.UWSA 2 259 2 days,

Real thing 252

.

Got scores today, 259/89! Very happy. Let me know if u have any questions, but i think my initial post is still applies. Interestingly uwsa 1 predicted exactly (right before the real thing.) good luck everyone! Back to the OR

Hi, all. My score is back and I figured I would follow up.

My goal pre-study: 240
My practice scores:
UWSA 1: 226 about 6 weeks before the test
UWSA 2: 236 about 5 weeks before the test
School administered NBME CBSE: 235 4 weeks before the test
NBME 12: 233 3 weeks before the test
NBME 13: 228 2 weeks before the test-almost had a heart attack taking this one. HARD
NBME 11: 233 1 week before the test

Real deal: 249/87

Just got my scores back this morning from the April 16th test date -- 262/89.


Practice test scores:
NBME 12, 5 weeks out -- 228
UWSA 1, 4 weeks out -- 247
NBME 12, 3 weeks out -- 250 (I took the same form again by accident -- but I hadn't done extended feedback the first time so I don't think it dramatically changed my score)
NBME 11, 3 weeks out -- 245
NBME 13, 2 weeks out -- 254
NBME 7, 2 weeks out -- 247
UWSA 2, 1 week out -- 262
NBME 3, 6 days prior -- 259
NBME 6, 2 days prior -- 254

 

[rocketbooster]

You are wrong. Here are just a few people posting in the 2012 experiences thread with very accurate predictions (and there are a ton more like this if you actually look):

You're joking, right? You're in denial. Don't rely on UWorld scores to predict your score. Let me break down every example you gave and point out how they do nothing to support your argument.

Quote:
Originally Posted by nemostar19
Kaplan q bank = 68% finished before school ended (march 18)

UWorld 77% average

Cbsse = 235 8 weeks
NBME 13 254 4 weeks
Free 150 at site 90% 260
UWSA 1 265 3 weeks
NBME 12 254 1 week
.UWSA 2 259 2 days,

Real thing 252

.

This person already maxed out their scores. Obviously if you are getting near 260 on NBMEs then UWorld would be similar because it's nearly impossible to go above 260+. This example in no way supports your claim. If you are getting 260 on NBMEs, you will ace any kind of practice exam you take. And look at his test results. At the beginning after 1 week of studying, he jumped 20 pts? You believe that? Half of the gain is from the UW overestimation. Now look at his last few tests, he had 247, 259, 254 NBME and 262 UW. He got a 252 on the real thing. Is 252 closer to his NBME scores or UW? Do I need to answer that for you too?

Quote:
Originally Posted by murdoc9
Got scores today, 259/89! Very happy. Let me know if u have any questions, but i think my initial post is still applies. Interestingly uwsa 1 predicted exactly (right before the real thing.) good luck everyone! Back to the OR

What I already said applies here, again. He was already maxing out his score most likely. Ask him what his NBME scores are and then you try to argue me using this example. Based on this, it is not usable for either side of the argument.

Quote:
Originally Posted by Tatastrophy
Hi, all. My score is back and I figured I would follow up.

My goal pre-study: 240
My practice scores:
UWSA 1: 226 about 6 weeks before the test
UWSA 2: 236 about 5 weeks before the test
School administered NBME CBSE: 235 4 weeks before the test
NBME 12: 233 3 weeks before the test
NBME 13: 228 2 weeks before the test-almost had a heart attack taking this one. HARD
NBME 11: 233 1 week before the test

Real deal: 249/87

Why did you even post this? If anything, this likely supports me. He only took UWorld at the very beginning of his studying. Substract 15-20 and he was probably realistically getting 210-220 before studying. After 3 weeks of studying, his score then increased to 230s on NBMEs. Do you really think his score did not change at all after the 1st 3 weeks of studying? If anything, you improve the most during the first half of your studying since that's the first time you're going over all the material.

Quote:
Originally Posted by sanityonleave
Just got my scores back this morning from the April 16th test date -- 262/89.


Practice test scores:
NBME 12, 5 weeks out -- 228
UWSA 1, 4 weeks out -- 247
NBME 12, 3 weeks out -- 250 (I took the same form again by accident -- but I hadn't done extended feedback the first time so I don't think it dramatically changed my score)
NBME 11, 3 weeks out -- 245
NBME 13, 2 weeks out -- 254
NBME 7, 2 weeks out -- 247
UWSA 2, 1 week out -- 262
NBME 3, 6 days prior -- 259
NBME 6, 2 days prior -- 254

Again, this person started very high. He was getting 240+ from the beginning. He's going to do well on every test. You can't see the +15 when the score is already super high since the curve is so narrow at that point.

I hope you use better statistical logic on your Step 1 or else you'll be missing lots of easy questions. Just in case you didn't get it...the ends of a curve have narrow distribution, i.e. very few ppl. The middle of a curve has wide distribution, i.e. majority of ppl. Your examples are at the very high end of the curve = narrow distribution = not indicative for most ppl.

Your argument that UWorld is a good predictor right before the test is unfounded. You only used examples of people who aced the Step 1, which is a minority of ppl and of course these ppl will get high scores on every test since they maxed out the score. Okay, I'm done.

 

[aspire13]

Of all the assessment tests which one do you think are closest to the real thing? what books should we must have...i m just a beginner!!

 

[usmleq]

Your argument that UWorld is a good predictor right before the test is unfounded. You only used examples of people who aced the Step 1, which is a minority of ppl and of course these ppl will get high scores on every test since they maxed out the score. Okay, I'm done.

No your statement that "every single person says UWorld for sure overestimates it" and that it over estimates by "15-20" are unfounded.

I'm not going to go through all the posts in the 2012 exp thread or all the posts in the 2011 exp thread (I do have to study), but as you know, the people who post here all tend to have high scores so that's why the examples (which I told you I just randomly found) all have high scores. i didn't do a damn scientific study.

Anyway take a look at this thread (2 years old, but these exams don't change much in 2 years) http://forums.studentdoctor.net/showthread.php?t=750266

-"I took UWSA2 and NBME7 5 days before my exam. UWSA2 was 2 pts higher than my real score, and NBME7 was 4 pts higher. Did UWSA1 a little less than halfway thru my dedicated study period and it was 8 pts lower than my real score"

-"I'm pretty confident UWSA1 underestimated my score. I took it a month after I took an NBME and I got the exact same score. I took UWSA2 about 4 weeks after that and jumped up 38 points. So I think UWSA1 underestimated me by 10-20 points at that time. "

-"USWA 1 - 3 weeks out...18 pts lower than real thing
USWA 2 - 1 week out...exactly the same as the real thing "

-"UWSA 1 - two weeks out - 2 points lower than the real thing
UWSA 2 - one week out - exactly the same"

-"UWSA1 (2 weeks out) = 3 points below actual score
UWSA2 (1 week out) = 4 points above actual score"

-"I took UWSA2 about 4.5 weeks in (a few days before my test) - it had climbed 20 points over UWSA1 and it ended up being 8 points lower than the real thing."

-"UWSA1 (4 weeks before) - 12 points lower than actual score
UWSA2 (2 weeks beofre) - 4 points higher than actual score"


And here's one for you: someone who hasn't maxed out their score:


-"UWAA1 (1.5 weeks out) 247
UWSA2 (.5 weeks out) 234

Actual- 248"

Again there are alot more, this is just the first posts from that thread

 

[rocketbooster]

No your statement that "every single person says UWorld for sure overestimates it" and that it over estimates by "15-20" are unfounded.

I'm not going to go through all the posts in the 2012 exp thread or all the posts in the 2011 exp thread (I do have to study), but as you know, the people who post here all tend to have high scores so that's why the examples (which I told you I just randomly found) all have high scores. i didn't do a damn scientific study.

What you're failing to analyze is every post that compares NBMEs and real score. There's way more variation in UW and real score compared to NBMEs and real score. I've read countless threads on this forum. UW is all over the place. Yes, for some ppl it may have been accurate but also for many it was 10-15 off. However, almost everyone had a small variation between their NBMEs and real score. Which do you think is more accurate then? NBME or UW? UW works for some apparently, NBME works for basically everyone. I'm gonna go NBME.

What were your NBME scores? I'm willing to bet that your NBME scores were also similar...

 

[usmleq]

What you're failing to analyze is every post that compares NBMEs and real score. There's way more variation in UW and real score compared to NBMEs and real score. I've read countless threads on this forum. UW is all over the place. Yes, for some ppl it may have been accurate but also for many it was 10-15 off. However, almost everyone had a small variation between their NBMEs and real score. Which do you think is more accurate then? NBME or UW? UW works for some apparently, NBME works for basically everyone. I'm gonna go NBME.

What were your NBME scores? I'm willing to bet that your NBME scores were also similar...

I agree that NBME scores seem to be accurate (possibly more accurate that UW). I think in general UW might be +/- 10-15 and NBME might be a little more accurate (but again it is +/-).

The point is that there is variation in both UW and NBME practice scores and it is +/-, UW doesn't exclusively over estimate (which is what you were saying earlier).

 

[rocketbooster]

I agree that NBME scores seem to be accurate (possibly more accurate that UW). I think in general UW might be +/- 10-15 and NBME might be a little more accurate (but again it is +/-).

The point is that there is variation in both UW and NBME practice scores and it is +/-, UW doesn't exclusively over estimate (which is what you were saying earlier).

Why did you argue with me to begin with? The point of my post was to tell the OP that NBME is the best predictor while UW has a lot of variation. You just agreed with me. Glad that settles that.

 

[usmleq]

Why did you argue with me to begin with? The point of my post was to tell the OP that NBME is the best predictor while UW has a lot of variation. You just agreed with me. Glad that settles that.

I was saying your statement below is incorrect - and now you now agree with on it. You just don't like to admit when you are wrong.

The take home message is UWorld definitely overestimates your score

 

[johndoe3344]

.

 

[Aclamity]

oh dear this thread is turning into a crapfest. I dont know how predictive UWSA is, but if your NBME scores are consistently lower than your UWSAs I would be safe and trust the NBME scores. If they're about the same then who cares?

 

[sanityonleave]

I had a similar problem to yours, but not as extreme. You can refer to my post here:
http://forums.studentdoctor.net/showthread.php?t=910249

Basically, we decided that I wasn't that good at answering "easy" questions, but I was better at reasoning through hard questions. In a typical NBME form, I'd have a small handful of questions I missed because I was being careless or tried to overthink the (apparently very easy) question. With the NBME's, even a few mistakes can easily drop your score by 10+ points.

If you find this happening to yourself, I think it's safe to say that on the actual exam you can expect a score somewhere between the NBMEs and the UWSAs, as these are both two extremes, with the actual exam being somewhere in the middle.

With regard to the predictive value of the UWSAs vs NBMEs, it is true that ON AVERAGE, the UWSA will overpredict while the NBMEs are traditionally better estimators. However, there are a LOT of people for which the vice versa is true -- the NBMEs will actually underpredict while the UWSAs will be more accurate (just go through the 2011 and 2012 score experiences thread yourself). I think it is possible that you could be in the latter category, especially with the addition of the new (read: more difficult) testing questions. In that sense, the curve on the actual exam will be more lenient (like the UWSAs) and less like the NBMEs.

I think johndoe did a great job of illustrating the value of the two tests. NBME exams have an extremely narrow curve where just a few questions either way can change your score +/- 10 points -- and there tend to be many more "one step" fact-based questions. UWorld questions tend to be harder but often offer more "hints" to allow you to reason to a 2- or 3-step answer.

Now the bad news -- there are lots of questions on the real thing that are simple factual recall, and if you continue to do poorly on those your exam score will reflect it. The good news -- it's a lot easier to bring up a score weighed down by missing simple facts than one lower due to inability to reason through hard questions.

My advice -- hit FA hard. Focus on pharm, micro, and HY lists -- either from the back of FA or the Goljan docs that can be found distributed around the internet. If you have micro or pharmcards, use them. Hit the simple high yield facts as much as possible. Your score isn't going to go up much from deepening your knowledge of physiology, but you may very well get a question on the test where a BM biopsy shows "tissue paper macrophages" (literally I had a question about a genetic disease that was this much of a giveaway on my test as long as you knew the keywords for that disease -- no images or anything, they gave you the buzzwords) and you have to pick Gaucher's disease.

Hit the high yield factual recall stuff hard and I think you're going to be fine for test day, but you should be prepared that with a gap that large between your UWSAs and NBMEs there's a big range for your potential score as well.

 

[HotLikeSauce]

Took the exam last year, but thought I could add some input.

I found that NBMEs underestimated my score routinely whereas UW probably overestimated my performance at the time. FWIW, I forced myself to take NBMEs but always dreaded taking them because I found them to be really hard (during the tests, I would flag EVERYTHING I felt like haha). I found the real test to be easier than an NBME as well as more straight forward.

UWSA1 (5 weeks out): 245
UWSA2 (4 weeks out): 260
NBME 5 and 6 (3 weeks out): 250, 251
NBME 11 and 12 (2 weeks out): 251, 245
NBME 7 (1 week out): 252

I ended up with a 258 on the real thing. UW is great and it was immensely helpful for prep, but do I think I was already at a 245 after 1 week of studying when I had taken it? Nah.

I would just look at the areas where you've seemed to struggle on past exams and try to hit those hard the last few days. Try to make your weakness your strength (sounds cliche, but it's true). If you're routinely missing most of your points in one area, it's worth it at this point to try to spend a bit of extra time and hopefully gain back some of those points on your real exam.

Overall, I'm glad I took the NBMEs. They expose you to a totally different type of question IMO that I hadn't really seen before, and I felt that having been exposed to them helped me a lot during the test, even though they were stressful to take. Like I said, I found the NBMEs (particularly the later ones like 11 and 12) to be a lot harder than my test. I know it's hard, but try not to get too caught up in the numbers right now so close to your test.

What was your UW%?

 

[ProtonatedAlchy]

thanks for the advice all- just took the test yesterday. it felt like an easier version of uworld, less nitpickety questions than the nbme practices I took which was very welcome. I'll let you all know how it ended up when i find out.

I took NBME 11 the week before my real test and just a day after my NBME 13 decline. Ended up with a 242 on nbme 11 when I took my time and split the sections up over two days. Final UWorld percentage was 72% after my first complete run through, I started out around 67% and by the end I was hitting ~80% on my blocks