AAMC statement…Any thoughts on this?

[700003]

I was reading the MCAT 2015 pdf document on how the test is scored and I came across this:

'Because the 2016 application cycle will open before the 2015 testing year is done, we will use data from the first few administrations of the new exam to create the percentile rank tables that your admissions committee can use for 2016 selection. We recognize that the first test takers on the new exam may be different from test takers who sit for the exam later in the year. It is possible that early test takers will be higher scoring than those who test later. We also recognize that 2015 may be an unusual testing year. Some test takers may decide to wait until 2016 to sit for the new MCAT exam.'

Why would AAMC think the early test-takers would score higher than the later test-takers? I mean, shouldn't it be the opposite way around? We hardly have any practice materials to work on! Is it possible that the AAMC assumes the first few administrations of the new MCAT will be easier and more leniently scored than the others?

---

Members don't see this ad.

 

[AspiringERMD]

I was reading the MCAT 2015 pdf document on how the test is scored and I came across this:

'Because the 2016 application cycle will open before the 2015 testing year is done, we will use data from the first few administrations of the new exam to create the percentile rank tables that your admissions committee can use for 2016 selection. We recognize that the first test takers on the new exam may be different from test takers who sit for the exam later in the year. It is possible that early test takers will be higher scoring than those who test later. We also recognize that 2015 may be an unusual testing year. Some test takers may decide to wait until 2016 to sit for the new MCAT exam.'

Why would AAMC think the early test-takers would score higher than the later test-takers? I mean, shouldn't it be the opposite way around? We hardly have any practice materials to work on! Is it possible that the AAMC assumes the first few administrations of the new MCAT will be easier and more leniently scored than the others?

Huh. Maybe they'll get a more generous scale to make up for the "unknown" factor?

 

[jurisburger]

Maybe AAMC is just trying to account for self-selection bias, in that the first wave of examinees taking the new exam are likely the group who (as a whole) feels more confident and prepared.

 

[700003]

Maybe AAMC is just trying to account for self-selection bias, in that the first wave of examinees taking the new exam are likely the group who (as a whole) feels more confident and prepared.

I don't understand…how are the first group of examinees supposed to feel more confident and prepared? If anything, we feel more terrified because we don't know how the exam is going to be.

 

[jurisburger]

I was going on a limb and assuming that surely a proportion of examinees taking the new exam on the first few dates (rather than waiting for later) are choosing to be guinea pigs because they are confident in their abilities to perform...hence self-selection bias and thus providing credence for AAMC's claims that the "early takers may score higher".

But I'm just theorizing. Because there are surely many other test takers who are forced to sit for these exams because they have no other choice in the matter, regardless of how prepared they feel.

 

[mehc012]

Perhaps part of the error comes in from the trial section data. I know they are withholding scores for the first few cycles of the new MCAT in order to gather information to establish a scale, but it is less clear whether the data from the trial section testers will be implemented in any way. At the very least, it seems that the material used on the new exam has to have been vetted by looking at trial section data...and pretty much everyone taking that section either winged it or answered mostly randomly. It wouldn't surprise me at all, then, if the first few rounds of the test are slightly easier because the scores of the guinea pigs they've had thus far are skewed (the apparent difficulty of the trial material was inflated because the testers weren't invested in their scores).

 

[johnnytest]

New mcat is stupid..

 

[Aloft085]

Perhaps part of the error comes in from the trial section data. I know they are withholding scores for the first few cycles of the new MCAT in order to gather information to establish a scale, but it is less clear whether the data from the trial section testers will be implemented in any way. At the very least, it seems that the material used on the new exam has to have been vetted by looking at trial section data...and pretty much everyone taking that section either winged it or answered mostly randomly. It wouldn't surprise me at all, then, if the first few rounds of the test are slightly easier because the scores of the guinea pigs they've had thus far are skewed (the apparent difficulty of the trial material was inflated because the testers weren't invested in their scores).

I doubt AAMC is that terrible with statistics. Most likely the trial data is used to better differentiate easy questions from hard questions. That kind of data would be valid even if 90% of the test takers guessed randomly, given a sufficiently large sample size.

As for AAMC's assumption that easy test takers may score higher, I agree with jurisburger's post above. People taking the test in the latter half of 2015 are not going to have really any accurate preparatory material either. The ones willing to take the leap first are probably disproportionately smart and prepared dudes.

 

[AspiringERMD]

I doubt AAMC is that terrible with statistics. Most likely the trial data is used to better differentiate easy questions from hard questions. That kind of data would be valid even if 90% of the test takers guessed randomly, given a sufficiently large sample size.

As for AAMC's assumption that easy test takers may score higher, I agree with jurisburger's post above. People taking the test in the latter half of 2015 are not going to have really any accurate preparatory material either. The ones willing to take the leap first are probably disproportionately smart and prepared dudes.

Great point.

 

[justadream]

The ones willing to take the leap first are probably disproportionately smart and prepared dudes.

This may be somewhat true but I think it's really not that big of a factor.

The most compelling reason why people are taking it in April/May as opposed to later this year is that many people want to apply for this upcoming cycle. Many people don't want to have to wait another year just to apply.

Even if you are the smart and prepared dude, if you have the choice, you would rationally choose to take it later when there is more prep material and a better understanding of the test.