is there any data on what test dates have the "easiest" curves?

The problem is that the test date is not correlated to curve difficulty or ease. Sure certain tests have an easier curve but whether or not you get one is not dependent on when you take the exam.

Hmm. haven't studied at all but the May 10th MCAT allegedly has a suuuuuper low SD and mean! Quick register now!!

baby methuselah said:

I've searched every thread on the mcat forum and nothing directly addresses this. I could have sworn I read a post on this last year, but maybe it's too old to stay on the board. Anyway, I was wondering if anyone has opinions on this. I don't know that AAMC would make that available, but anything other than "Just take it when you're ready" would be appreciated.

Click to expand...

Someone recently told me the August/Sept curves are pretty bad since most people have 3+ months to study without having to be in school. If I had to take a guess, I'd say January has the best curve. Most people will be focusing on finals during Nov/Dec, and only have a month of Christmas Break to really focus on studying for the MCAT.

BigJuicyWart said:

Someone recently told me the August/Sept curves are pretty bad since most people have 3+ months to study without having to be in school. If I had to take a guess, I'd say January has the best curve. Most people will be focusing on finals during Nov/Dec, and only have a month of Christmas Break to really focus on studying for the MCAT.

Click to expand...

The problem is that the MCAT is not curved just to the current year. The AAMC reuses test questions and to my knowledge they judge against the entire history of test takers...not just the current year. Your curve is more or less set before you walk into the test center. The current year is added to modify the curve but with the number of years that the MCAT has been administered, it probably does not have that great an effect.

Vihsadas said:

The problem is that the MCAT is not curved just to the current year. The AAMC reuses test questions and to my knowledge they judge against the entire history of test takers...not just the current year. Your curve is more or less set before you walk into the test center. The current year is added to modify the curve but with the number of years that the MCAT has been administered, it probably does not have that great an effect.

Click to expand...

Well, everyone keeps mentioning that "they judge against the entire history of test takers and not just the current year" as if you guys know for sure that such is the case. It is nothing but speculation, no one really knows how these tests are curved. Maybe it is based on the test takers on that particular day, or maybe it is not. I personally believe that it is based on the test takers of the particular test date, which in my opinion explains why it takes a whole month to get the score back. The writting section is just not important enough for it to merit an entire month of scrupulous grading. Furthermore, yes AAMC might reuse a couple discretes or passages here and there, but it is difficult for me to believe that the whole test is comprised of old recycled test questions from past years.

mterp45 said:

Well, everyone keeps mentioning that "they judge against the entire history of test takers and not just the current year" as if you guys know for sure that such is the case. It is nothing but speculation, no one really knows how these tests are curved. Maybe it is based on the test takers on that particular day, or maybe it is not. I personally believe that it is based on the test takers of the particular test date, which in my opinion explains why it takes a whole month to get the score back. The writting section is just not important enough for it to merit an entire month of scrupulous grading. Furthermore, yes AAMC might reuse a couple discretes or passages here and there, but it is difficult for me to believe that the whole test is comprised of old recycled test questions from past years.

Click to expand...

I have spoken to one AAMC administrator. This is what he led me to believe. It would still take a month to get your scores back with the system that I mentioned. They would still factor in the curve from the current year or test date, so it would take some time to factor how the current year affects the testing history. The reason I believe that the day doesn't really matter that much is because the effect on the curve of the current years test takers wouldn't have that much effect on the testing history.

From the reporst of people that I personally know who have taken the exam over the past 4 years, there are definitely multiple passages reused every year. The 'experimental' passages in the exam might (and this is speculation) be how the AAMC develops the curves for new material being introduced.

If they curve by the national average for that particular day, i would guess that the may dates may reflect a more favorable curve. Its during the heat of finals time for most undergrads. Pure speculation though. Who really knows?

baby methuselah said:

does anyone know if the july, aug, sept curves are less lenient than apr, may?

Click to expand...

My guess is that August has the most harsh curve of them all, and January probably has the easiest.

I have spoken to one AAMC administrator. This is what he led me to believe. It would still take a month to get your scores back with the system that I mentioned. They would still factor in the curve from the current year or test date, so it would take some time to factor how the current year affects the testing history. The reason I believe that the day doesn't really matter that much is because the effect on the curve of the current years test takers wouldn't have that much effect on the testing history.

Click to expand...

That doesn't make any sense to me. For one, the MCAT was offered two times a year until recently. So with 22 test administrations, you're looking at overlap within the same test year in some cases (let's face it, they're not going to use Bio passages from ten years ago). I doubt they'd do that.

Second, the history of test takers for every individual questions vs. curving within the test taking day? It just doesn't sound reasonable.

Vihsadas said:

I have spoken to one AAMC administrator. This is what he led me to believe. It would still take a month to get your scores back with the system that I mentioned. They would still factor in the curve from the current year or test date, so it would take some time to factor how the current year affects the testing history. The reason I believe that the day doesn't really matter that much is because the effect on the curve of the current years test takers wouldn't have that much effect on the testing history.

From the reporst of people that I personally know who have taken the exam over the past 4 years, there are definitely multiple passages reused every year. The 'experimental' passages in the exam might (and this is speculation) be how the AAMC develops the curves for new material being introduced.

Click to expand...

Vihsadas is right. I spoke with an administrator as well who said the test curve is generated from the previous 3 years of test-takers. He did not give the specifics, but basically said that they compare examinees who have taken similar difficulty passages/questions to assign raw-to-scaled scores.

bluemonkey said:

Vihsadas is right. I spoke with an administrator as well who said the test curve is generated from the previous 3 years of test-takers. He did not give the specifics, but basically said that they compare examinees who have taken similar difficulty passages/questions to assign raw-to-scaled scores.

Click to expand...

To be fair, I'm only half-right. I thought that the testing history would be much longer than that, but the AAMC guy I spoke too wouldn't give specifics either...

Vihsadas said:

To be fair, I'm only half-right. I thought that the testing history would be much longer than that, but the AAMC guy I spoke too wouldn't give specifics either...

Click to expand...

Once upon a time, they used to put more information on their website about the grading and scoring. It's a shame they don't list more information, but that would likely open them to unwanted scrutiny and a bombardment of questions. At the AAMC conferences each year, they usually have a workshop (or two) on the MCAT. I've been lucky enough to attend a couple of times.

What you mentioned earlier sounds pretty much identical to what the speaker was saying. If I'm interpreting it correctly, then each administration of the CBT includes several questions (and passages) that have appeared on the MCAT before. Current test-takers are curved against this continually building data pool. The new questions are scored and hyper-analyzed. The ones they keep are incorporated into the curve established by the field tested questions. Once they are done, based on whatever questions remain standing, a curve is set and scores are awarded.

As for one day being harder than another, we (TBR) have tracked how our students have scored versus where, what da,y and what time they took their exam. There seems to be a pattern and based on it, we recommend certain conditions in terms of location, weekday versus weekend, and AM versus PM, but they are meerly suggestions. The ultimate measure of when you should take it lies with when you've completed the necessary coursework, when you'll have time to prepare and most importantly when you'll be applying.

BerkReviewTeach said:

...
As for one day being harder than another, we (TBR) have tracked how our students have scored versus where, what da,y and what time they took their exam. There seems to be a pattern and based on it, we recommend certain conditions in terms of location, weekday versus weekend, and AM versus PM, but they are meerly suggestions...

Click to expand...

That's really intriguing =D

BerkReviewTeach said:

Once upon a time, they used to put more information on their website about the grading and scoring. It's a shame they don't list more information, but that would likely open them to unwanted scrutiny and a bombardment of questions. At the AAMC conferences each year, they usually have a workshop (or two) on the MCAT. I've been lucky enough to attend a couple of times.

What you mentioned earlier sounds pretty much identical to what the speaker was saying. If I'm interpreting it correctly, then each administration of the CBT includes several questions (and passages) that have appeared on the MCAT before. Current test-takers are curved against this continually building data pool. The new questions are scored and hyper-analyzed. The ones they keep are incorporated into the curve established by the field tested questions. Once they are done, based on whatever questions remain standing, a curve is set and scores are awarded.

Click to expand...

It makes sense that they have a curve set before based on the historical data, it has to be like that. This means they estimate the percentages of score, for example they estimate how many people gonna get 14s, 13s,.... This tells me that each test more or less has to have a similar curve may be 1 or 2 questions different. Imagine one day they give really easy test, many people score high( schewed to the right) they are screwed, to prevent such a problem they have field tested passages and they incorporate them into the test. that way they have an idea of the curve. Question is, these AAMC people are experienced, do they take into account different times of year how population performed at a given month based on past? I mean would they give a harder test in August because people who take it in Aug has shown to be smarter and score higher?

baby methuselah said:

and i bet that the only way to learn this pattern is to enroll in a tbr class? if i bought a couple of tests from you guys, would you tell me?

Click to expand...

Unless you are taking it in California, our data probably don't apply to you. Factors such as traffic on a weekday morning and external noise at certain locations play a major role in the suggestions we make in terms of where and when to take the exam.

If you happen to be in California, PM me and I will gladly make some suggestions as to when and where to take the exam.

BerkReviewTeach said:

If I'm interpreting it correctly, then each administration of the CBT includes several questions (and passages) that have appeared on the MCAT before. Current test-takers are curved against this continually building data pool. The new questions are scored and hyper-analyzed. The ones they keep are incorporated into the curve established by the field tested questions. Once they are done, based on whatever questions remain standing, a curve is set and scores are awarded.

Click to expand...

I feel it has to be more complicated than that... There is really no "continually" building pool if they're just evaluating new scores using the old data - the new data points will have no effect on the curve because the curve is already set. Then they'll have to use the OLD DATA to come up with a curve for a set of new questions... So basically, all of this goes back to a few set of data that they used to curve, and all subsequence curves are generated using those as standards? That seems a bit too extreme...

But I definitely agree that the curve is not ONLY based on the test-takers for a certain day. What I wonder is to what degree will the test-takers on a given day actually be able to influence the curve. Does anyone know about what percentage of passages are reused from old tests - close to 100% or much less?

con1988 said:

I feel it has to be more complicated than that... There is really no "continually" building pool if they're just evaluating new scores using the old data - the new data points will have no effect on the curve because the curve is already set. Then they'll have to use the OLD DATA to come up with a curve for a set of new questions... So basically, all of this goes back to a few set of data that they used to curve, and all subsequence curves are generated using those as standards? That seems a bit too extreme...

Click to expand...

But that's not exactly what's happening. There is a continually building pool because the curve is set based on old data and the new data. The bell curve has been generated over some number of past years (the actual number of past years that are considered is not known), and then the scoring from the current year and testing day is also used to regenerate the bell curve. Your score is released from the new bell curve. The point of contention is that noone can really say how large the historical data pool that is actively used in curving the MCAT is. It could be that it's only a few years, then the current testing year would have a significant effect on the exam. It could be that its more than a few years, then the current administrations would have less effect...

But I definitely agree that the curve is not ONLY based on the test-takers for a certain day. What I wonder is to what degree will the test-takers on a given day actually be able to influence the curve. Does anyone know about what percentage of passages are reused from old tests - close to 100% or much less?

Click to expand...

It's unknown how many experimental passages and questions there are...what is known is that every exam has some of these. If the experimental passages and questions generate good statistical data then they are kept and curved. If not, they are thrown out.
So the MCAT is made up of previously used passages and questions that already have a decent curve attached to them, and then a group of experimental questions that are statistically analyzed to see if they would be good MCAT questions (generate a good bell curve).

Well I agree for september being the hardest curve. I will be taking my MCAT this september. On last test i had 12+ in sciences, but it is verbal dragging me down. I have two friends that are in the same boat, who are also writing it in the september. Just a heads up