I like your response. I was trying to link my last comments to that newsletter but it wasn't working for me.
You are right, it seems that repeat test takers would be included for percentile purposes. And I implied that this wasn't the case, because that is what I was thinking. That is probably wrong, but either way things are still a little murky and here why I say this:
Interestingly, last year the mean, sd, and passing score were all exactly the same as this year. I am basing that on old post on this website from last year which i went back and re-read earlier today. So does that 209 passing score in 2006 match up to the 26% percentile or the 12th percentile?
The calculated percentiles are a little closer to 26 but not much. It is almost right in between which means either the distribution is either much tighter or much looser than one would calculate.
And certainly if re-takers are included in the percentiles this makes the comparison to the ITE difficult if not impossible.
Most of the repeaters fail the exam, most score less than 209. So one would expect the scoring percentiles to jump from when one takes the exam from starting CA-3 (the ITE) to finishing CA-3 (the real thing) just because you've added a bunch of people to the group who do quite poorly on the exam. This would be true even if one has stayed at the relative same percentile within their peer group. i.e. if you have tended to score around 50th percentile or 75th or 95th on your ITE, and you have actually done sone when compared to the FIRST-TIME test takers, your percentile on the real exam would be significantly higer IF ALL SCORES are in included. Based on my relatively small N this doesn't seem to be the case, which has led me to further believe that the numbers are skewed, but that may have something to do with my sample being at the higher end of the score range.
Anyway, i dont have an answer to all this. I just think it could be more clear. The ABA is a bit cryptic in my opinion.
In 1999 the NBME stopped reporting percentiles for the USMLE. I don't really know why exactly. Seemed like they wanted to make the information a little less clear and people would then focus more on pass and fail. I think the ABA exam is made in conjuction with the NBME. In the defense of the NBME and the USMLE, I think the pass rates are a little more appropriate (95% range). It seems kinda harsh that in 1 in 8 finishing CA-3's failed the exam last year. Hopefully fewer failed this year.
I must say, I never worried about the written exam because the ITE let me know it wouldn't be a problem but quite honestly, the fail rate for the orals is somewhat worrisome especially given the length of the exam (short) and the subjectivity. Why in the world would 25% of people who just passed their written exam need to fail this test? Again, seems pretty harsh.