Modern/Post 2015 Step 1 Scores and UWORLD correlation.

[XxSynapsexX]

Thought I would start this thread. Ive noticed a lot of the UWORLD % and STEP 1 threads are old and they make ridiculous notions like a 75% on UWORLD gets a 255.

I feel like Ive been getting around that on timed random mode just from knowing the basics on first aid and only get stuff wrong when it asks specific details or in a complex way and I feel like that level of knowledge correlates more to 230s-240s.

SO anyone who took the test post 2015, what were your STEP 1 scores and UWORLD %?

Also would anybody be able to say roughly what UWORLD scores in the 60s,70s,80s, would correlate to on the test?

---

Members don't see this ad.

 

[dempty]

With all of the data floating around, it's sad that none of these are really done properly. Even the reddit thread, which appears to be one of the better attempts, is pretty lacking.

Maybe someone will add a google excel sheet where people could start adding their information. It would be available on SDN and could be updated as needed. People can do what they want with it on their own computer to analyze it whichever way they feel comfortable, and at the end of the day, we might get some objective and higher quality data on the predictive utility of NBMEs/Uworld (% and percentile)/Kaplan and others, rather than the typical eyeball of people just saying "yeah it worked for me." This approach that everyone has been using isn't too scientific and is essentially no different than physicians only writing case reports or case series (to draw a parallel). The human mind often sees what we hope to see, so getting better numbers for this stuff would be a step in the right direction. Even though the sample will be biased, it may not be too far off from the population that is likely to use it (sampling from SDN users to use with SDN users isn't necessarily biased whereas sampling from SDN to use with anyone taking Step 1 may be an issue).

For your actual questions, I have no answers since the raw data is slim or the analyzed data leaves a lot to be desired.