Correlation Uworld% to USMLE Score

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

TheSeanieB

Full Member
10+ Year Member
Joined
Dec 17, 2010
Messages
263
Reaction score
1
Approximately percentage correct in UWorld before taking the USMLE and then actual USMLE score.

Members don't see this ad.
 
This is probably one of the weakest correlations you can attempt to make
 
Members don't see this ad :)
Also, For This To Work Out, The SDN Names @least ShouldBe Displayed.
That Way Can Double CheckThat It MatcheS With The Scores in "the Step1 Scores & Experiences Thread"
 
Also, For This To Work Out, The SDN Names @least ShouldBe Displayed.
That Way Can Double CheckThat It MatcheS With The Scores in "the Step1 Scores & Experiences Thread"

I made it anonymous so that users can participate without showing their score. It is pretty obvious which should be disregarded.
 
I see the trolls have done their part. :laugh:

< 50% in UW and 260+

That wasn't even worth putting up.

I made it anonymous so that users can participate without showing their score. It is pretty obvious which should be disregarded.

Not that obvious, b/c then CDI's example occurs (amongst many others, like >90% UW and <188). Then the poll loses credibility.

I think there are plenty of people who post their scores on the "scores & experiences thread" to make this poll worth it only with their credible data.
Heck, you could probably gather all this data yourself by going thru the "scores & experiences" threads for the last few years. That would be more credible AND you could also compare any trends seasonally and/or annually.
 
Not that obvious, b/c then CDI's example occurs (amongst many others, like >90% UW and <188). Then the poll loses credibility.

I think there are plenty of people who post their scores on the "scores & experiences thread" to make this poll worth it only with their credible data.
Heck, you could probably gather all this data yourself by going thru the "scores & experiences" threads for the last few years. That would be more credible AND you could also compare any trends seasonally and/or annually.

Maybe. But I would remove that entry before doing any sort of analysis.
 
FWIW, did a quick and dirty linear regression on the 16 legit votes (61% corelation). According to it, a 59% on UWorld approximates getting a actual score of 225 and a 71% on UWorld equates to a 240. If more people legitimately report their stats, I'll put together some better information.
 
FWIW, did a quick and dirty linear regression on the 16 legit votes (61% corelation). According to it, a 59% on UWorld approximates getting a actual score of 225 and a 71% on UWorld equates to a 240. If more people legitimately report their stats, I'll put together some better information.

Not sure how you're determining which are "legit". Additionally if you don't think such scores are legit, why put them as an option on the poll?
 
I think its better to do like last 15 tests or something.. Some ppl start in the 30s and end in 60+ but their overall
mean will be low..

Yes, I agree. The UWorld% should be an approximation of where you are on test day with UWorld rather than your cumulative performance on the entire database.
 
With 63 votes, new quick and dirty estimate is 54.6% equates to a 225 on the actual test and 67% equates to a 240. I did not include outliers in this estimate.
 
Top