Not so sure about that. We have 50ish so far and they seem to be all over the placeI sure do love posts like this. There will absolutely be a correlation between these scores. The question is, what are you trying to get from it? How will you interpret it?
Ah, I see now. You're using the word in a loose sense. Just because the correlation isn't high doesn't mean there isn't a correlation. Further, a non-significant correlation (if that's what happens in your data) doesn't mean that there is no correlation/relationship; absence of proof is not proof of absence, and the method you use doesn't necessarily cover other kinds of relationships that might be present. Even if there was a high correlation between the two exam scores, how would one go about comparing the scores from totally different exams? Imagine the correlation is -0.9. What does this mean to you? How would you use it to compare two candidates?Not so sure about that. We have 50ish so far and they seem to be all over the place
Finding a correlation (or showing there isn't one) is something that needs to be done if residency programs are going to continue to allow people to only take the comlex. Without a correlation there is no way to compare applicants based on their test scores
Ah, I see now. You're using the word in a loose sense. Just because the correlation isn't high doesn't mean there isn't a correlation. Further, a non-significant correlation (if that's what happens in your data) doesn't mean that there is no correlation/relationship; absence of proof is not proof of absence, and the method you use doesn't necessarily cover other kinds of relationships that might be present. Even if there was a high correlation between the two exam scores, how would one go about comparing the scores from totally different exams? Imagine the correlation is -0.9. What does this mean to you? How would you use it to compare two candidates?
I'm not trying to troll, but I am curious about the reasoning and practical benefits of even showing there is any kind of correlation.
I definitely know where you're coming from with the personal curiosity in the matter (that's why I'm asking you questions on it). One of the biggest obstacles I foresee is the interpretation of any results. Let's assume we got a very strong correlation like 0.95; what should that mean to a PD looking at someone's score?I'm doing this for my own curiosity. But I do think it would be beneficial for someone to get real data.
Right now you can't compare someone with a 230 Usmle to someone with a 650 comlex. Needs to be some way to compare the two. Can possibly be done by collecting this data
I mean if they find over and over that people who score 600's are scoring around 230, then they'll be able to compare an MD student with a 230 to a DO student with a 600. I'm by no means a statistician so I won't even pretend to have any real ideas lol. I just know it needs to be done.I definitely know where you're coming from with the personal curiosity in the matter (that's why I'm asking you questions on it). One of the biggest obstacles I foresee is the interpretation of any results. Let's assume we got a very strong correlation like 0.95; what should that mean to a PD looking at someone's score?
I would be interest to see what people have in the way of solutions for this. If there is a movement towards just the main testing bodies working together, what would people think of a "core" test which is STEP 1, and then an exam that's additional (but shorter) to incorporate the different focus for D.O. students?
That makes sense now. It sounds like you're looking for a way to find a comparable score (correlation more in a colloquial and less technical sense). I definitely think it would be cool to look at whatever information you get, but I agree with you that it'll be pretty involved for it to actually be done. I think the testing bodies need to agree that the tests are similar enough to actually equate scores, but I see that as a big obstacle (unfortunately). Can't hurt to try though. Let us know what you come up with!I mean if they find over and over that people who score 600's are scoring around 230, then they'll be able to compare an MD student with a 230 to a DO student with a 600. I'm by no means a statistician so I won't even pretend to have any real ideas lol. I just know it needs to be done.
One test would be great, but I don't see it happening. Tests = money
Any correlation you find is meaningless considering the vast difference in the caliber of the populations taking the exam.
