National Shelf Exam Scores

[siimpleton]

Is there somewhere that states the national averages for shelf exams? My school doesn't tell us.

---

Members don't see this ad.

 

[QuinnB]

the short answer is no...nbme gives this to your school to use at their discretion

 

[siimpleton]

Then, if anyone would like to offer their scores and percentiles, I'd appreciate it. I really have no sense of where I fall and am trying to make a decision about step 2 based on how I have been doing so far.

 

[siimpleton]

OR.. if anyone knows of a thread that has this that would be great. I would imagine there is one, but I couldn't find it.

 

[SaraMac13]

Our school gave us the stats from 2004-2005. I'll give you the means for the total year. Here ya go


Medicine: Mean 74.4, SD 7.9
Neurology: Mean 71.6, SD 7.5
OB/Gyn: Mean 71.3, SD 8.2
Pediatrics: Mean 74.0, SD 8.4
Psychiatry: Mean: 75.8, SD 8.4
Surgery: Mean: 70.9, SD 8.8

Sorry I don't have Family Med - our school doesn't make us take it

 

[zeloc]

Maybe someone can clarify how the scoring works. I thought all shelf exams were set such that the national mean is 70 and the standard deviation 8. Was this done for one particular year and then after that there are different means and standard deviations depending on the strength of the class?

 

[Syranope2]

Maybe I don't understand statistics, but wouldn't the average grade be the 50th percentile?

 

[kito]

the numbers might be raw, not percentiles.

 

[QuinnB]

Maybe I don't understand statistics, but wouldn't the average grade be the 50th percentile?

A lot of people confuse this, especially with board scores, etc...

You can't calculate a true percentile from a mean. Percentiles are based on a median and the associated quartiles. Therefore, you can't say that an average of 70 is the "50th percentile" b/c the mean is altered by outliers whereas the median is not. As the number of individuals involved increases, using standard deviations will give you an approximate idea of how many people are above/below a certain level (e.g., +/- 1 SD has 16.67% above, +/- 2 SD has 2.5% above, etc...) which has been used as a surrogate for percentile by most non statistics oriented people b/c the distribution of scores is thought to be "normal about the mean" (= normal bell curve).

I hope this helps.