Eh, equating PCAT performance to patient's lives is stretching, even equating NAPLEX to lives is questionable but certainly closer to the mark than PCAT.
But I can't believe anyone would defend a 7th percentile. How can you possibly believe that person would have the prerequisite knowledge to succeed in pharmacy school? It doesn't mean the person can never be a pharmacist but it does mean they don't have the knowledge required to start pharmacy school. At least, that is what I would have thought...
Sent from my SAMSUNG-SM-G920A using
SDN mobile
Not wanting a controversial argument, but depending on the admissions office, this might have been either a diversity or a legacy admit. I'm seen some truly atrocious scores. This might have not been with the consent of the College of Pharmacy but a University stipulation.
By the way, the PCAT is not a straight percentage test as some of you seem to be commenting. It's normalized to a group of people. So, what that 7% is saying is that this person performs at that level compared to an index of testing peers. (That means that there is an expectation that 93% of candidates test at or better than s/he does).
Illustrated:
100 question test: (Number of questions and percentages, with 100 questions, they are the same)
Calibration group:
Owlegrad - 96
Gwarm01 - 95
Lord999 - 90
PAtoPharm - 60
Scale rank score:
Owlegrad - 99% (only 1% or so are expected to score at 96 or higher, note that this is not 100% even though this was the exam setting score)
Gwarm01 - 75% (only 25% or so are expected to score at 95 or higher, though is a very small margin considering Owlegrad's example)
Lord999 - 50% (half of all expected examiners are supposed to get 90 or higher)
PAtoPharm - 25% (75% of all expected examiners are supposed to score above 60).
We have a candidate that scores:
a. 50 raw score, then the candidate would get a percentage no higher than 25% and is adjusted to 50/60 comparison.
b. 75 raw score, then the candidate would get 37% as 15 raw score puts that candidate between PAtoPharm and Lord999.
Depending on how the curve is set, I would consider accepting a low scaled score given other characteristics. The correct interpretation is if this candidate tested against the reference group, then 93% of that reference group would either score as well or better than this candidate, not that 7% of the questions were answered correctly. If 93% of people are expected to score better, then I would have some issues with thinking that I can draw from a better pool. But, if the other credentials are intact, possibly it's due to a bad exam. That said, it takes more than 50 of the raw score correct to even get to the 25% scale score.
Before the 2004 rescale, basically, you only needed to answer something like 40% of the questions correctly to get a 99% in the Chem section. That is why there were consistent 99%ages on the PCAT and why it was not used as admission criteria back then, because even if you scored in the 99%, the initial population setting score was so abysmal that the outperform score is easily achievable.
Random guessing is a combinatorial problem off either the Poisson (for large numbers of questions >50-60 and the assumption that every single question is independent) or the Negative Binomial (for any case, even small ones and also thinks that if you answer one question correctly, related questions should have a higher percentage of being answered correctly). If it were a straight percentage, this person would need to guess correctly around 40/200 for a five-choice multiple test to at least achieve that much due to random chance, but at 14 (7%), is way, way out of simple random guessing. If I got that kind of score, my guess would have been the student had the wrong underlying concept as that's far worse than naive guessing.
--------
NAPLEX doesn't work that way as it's a straight curve not a population scaled curve. That's why I'm harsh toward those who don't pass, because the scaling works definitely in the candidates favor and it's much less than a raw 75 percent to pass. But also in response to some other comments regarding that, no I don't consider the NAPLEX any more than the bare minimum standard, but it should be reasonably achievable consistently by us. MCAT and USMLE have similar approaches and do not use a reference group as a scoring standard.