August and September MCATs Scored Harder?

[DrBTS]

---

Members don't see this ad.

I think it makes sense that there is a significant percentage of test takers that have begun studying in Early April/March to take their tests beginning in May or so on, who have either postponed, or voided, or are retaking the test to score a higher score, and thus have scheduled for the end of the summer. Since our numerical scores are based off how well we do with respect to other test takers, wouldn't it follow suit that these groups of test-takers would be more seasoned, theoretically better prepared, and less-stressed (not in school, summer relaxation) than those who take it form January to May? I'm basing this purely from the way the school calendar is set up, application deadlines, and general habits of pre-meds taking the exam twice. What do you guys think?

Disclaimer: Please take my rumination with a grain of salt, I really dont want this devolving into a personal attacks on my thought process, remarks on competition or being competitive...
just simply trying to get a light-hearted discussion out there to think of something that qualitatively has substantial logic behind it. Thanks

 

[DrBTS]

So I was doing a little bit of searching on previous SDNers who may have had the same thought process:

ok heres the truth:

the "curves" are NOT based on how your fellow MCATers, who take the test a at the same time as you, perform. But rather, is based off how previous students responded to the questions when the same questions were given on previous years MCATs.

therefore, it does NOT matter whether you take it in april or august

That raises another question of whether or not this is true? It would actually solve this problem drastically by increasing the N number for the curve and thus standardizing things across various test-dates.

 

[Melomare17]

answer to thread title: NO

 

[DrBTS]

I understand that the thread title/subject comes off a bit...idk, neurotic? But, again, I would really appreciate a legitimate discussion instead of the unconscious response that you may or may not have any real thought-process backing up. Thanks.

Mark Twain once said: "when you find yourself standing on the side of the majority, it's time to pause and reflect." 🙂 😉 8)

 

[wjs010]

I take on aug 3 and have been trying to trick myself into thinking it will be harder, for some of the same reasons. The answer is that it probably won't matter or make any different. You could actually say people are beginning school again so it should be easier in August. But there's no way to know

 

[gettheleadout]

I understand that the thread title/subject comes off a bit...idk, neurotic? But, again, I would really appreciate a legitimate discussion instead of the unconscious response that you may or may not have any real thought-process backing up. Thanks.

Mark Twain once said: "when you find yourself standing on the side of the majority, it's time to pause and reflect." 🙂 😉 8)

I have done extensive research into this personally and I can guarantee you that it does matter when you take the test and that the questions that actually contribute to your score are not evaluated based on the performance of other test takers on that day.

It's really quite ridiculous that this question gets brought up here every other week.

 

[grburst]

I have done extensive research into this personally and I can guarantee you that it does matter when you take the test and that the questions that actually contribute to your score are not evaluated based on the performance of other test takers on that day.

It's really quite ridiculous that this question gets brought up here every other week.

Could you elaborate? 🙂

 

[gettheleadout]

Could you elaborate? 🙂

Specifically, every administration of the MCAT contains a mixture of (predominantly) scored items and a small number of "experimental" items, referred to by the AAMC as "try-out" items. These latter items on the exam are scored but not factored into your actual score. They are scored for the purposes of establishing a difficulty rating for the items; the percentage of test takers that answer the items correctly corresponds to the AAMC's rating of the question difficulty. When a particular exam form is assembled, the distribution of question difficulty is factored in to the raw-to-scaled score conversion, so that tests containing different proportions of different difficulty questions yield the same scaled scores based on test taker ability. Thus, the only questions that you are scored on have already been rated for difficulty, and the performance of your peers on the day of your test does not influence your score. It will, however, influence the difficulty rating the "try-out" items on your particular test will receive, and will contribute to future test form scales.

 

[grburst]

Specifically, every administration of the MCAT contains a mixture of (predominantly) scored items and a small number of "experimental" items, referred to by the AAMC as "try-out" items. These latter items on the exam are scored but not factored into your actual score. They are scored for the purposes of establishing a difficulty rating for the items; the percentage of test takers that answer the items correctly corresponds to the AAMC's rating of the question difficulty. When a particular exam form is assembled, the distribution of question difficulty is factored in to the raw-to-scaled score conversion, so that tests containing different proportions of different difficulty questions yield the same scaled scores based on test taker ability. Thus, the only questions that you are scored on have already been rated for difficulty, and the performance of your peers on the day of your test does not influence your score. It will, however, influence the difficulty rating the "try-out" items on your particular test will receive, and will contribute to future test form scales.

I see, that makes sense. And there's no way to predict when the exams are harder? Based on what I've read here, it sounded like the exams this past May were tough and I didn't see a lot of really high scores. But I saw lots of very high scores last September. Who knows.

 

[gettheleadout]

I see, that makes sense. And there's no way to predict when the exams are harder? Based on what I've read here, it sounded like the exams this past May were tough and I didn't see a lot of really high scores. But I saw lots of very high scores last September. Who knows.

The only reason I can imagine it would matter (and I can see this making a difference, which does suggest a slight unfairness in the administration of the exam) is that taking a more difficult exam form could constitute an intimidation factor that affects examinee confidence. That said, if that's reflected in test takers' scores, it would affect the creation of the scale anyway. (i.e. if harder questions scare people and they do worse, the questions get rated as harder anyway and so the scales reflect that and everything's fair.)

All in all, I wouldn't worry about it. I might say I would prefer an "easier" exam for the sake of confidence during the test, but I honestly don't think it would affect the score I'd get simply because I trust the scaling to give me whatever score I deserve.

But no, short answer is there is no way to predict which exam administrations will be more skewed toward difficult questions and thus more lenient scales; not that it matters.

As far as SDN goes, keep in mind we have a biased population sample coupled with biased score reporting; I seriously wouldn't put any stock in the types of scores you see for a given test date on SDN. It's unlikely the score distribution nationwide varies much between test dates precisely because the scales are intended to even it out between tests.

 

[inycepoo]

omg people -____________________________________________________________-

You are lucky GTLO has the patience from God knows where to deal with this.

 

[inycepoo]

Disclaimer: Please take my rumination with a grain of salt, I really dont want this devolving into a personal attacks on my thought process, remarks on competition or being competitive...
just simply trying to get a light-hearted discussion out there to think of something that qualitatively has substantial logic behind it. Thanks

Sure, but if you used THE SEARCH FUNCTION, you would've found that said discussion occurs every 10 days with the same results as this thread shows.

 

[Melomare17]

Specifically, every administration of the MCAT contains a mixture of (predominantly) scored items and a small number of "experimental" items, referred to by the AAMC as "try-out" items. These latter items on the exam are scored but not factored into your actual score. They are scored for the purposes of establishing a difficulty rating for the items; the percentage of test takers that answer the items correctly corresponds to the AAMC's rating of the question difficulty. When a particular exam form is assembled, the distribution of question difficulty is factored in to the raw-to-scaled score conversion, so that tests containing different proportions of different difficulty questions yield the same scaled scores based on test taker ability. Thus, the only questions that you are scored on have already been rated for difficulty, and the performance of your peers on the day of your test does not influence your score. It will, however, influence the difficulty rating the "try-out" items on your particular test will receive, and will contribute to future test form scales.

Could you provide a source for where you found this information? Thanks.

 

[DrBTS]

Specifically, every administration of the MCAT contains a mixture of (predominantly) scored items and a small number of "experimental" items, referred to by the AAMC as "try-out" items. These latter items on the exam are scored but not factored into your actual score. They are scored for the purposes of establishing a difficulty rating for the items; the percentage of test takers that answer the items correctly corresponds to the AAMC's rating of the question difficulty. When a particular exam form is assembled, the distribution of question difficulty is factored in to the raw-to-scaled score conversion, so that tests containing different proportions of different difficulty questions yield the same scaled scores based on test taker ability. Thus, the only questions that you are scored on have already been rated for difficulty, and the performance of your peers on the day of your test does not influence your score. It will, however, influence the difficulty rating the "try-out" items on your particular test will receive, and will contribute to future test form scales.

Yeah this makes sense, I'm sure it's much more complicated IRL BUT it again it seems like a good way to decrease dimensionality between location and test date. That being said if this is the case, then verbal is probably would have the most drastic impact followed by BS.

 

[gettheleadout]

Could you provide a source for where you found this information? Thanks.

Published AAMC materials and personal communication with them. I don't have a list of citations on hand, sorry. Read MCAT Essentials and call them if you want to confirm.

 

[inycepoo]

Could you provide a source for where you found this information? Thanks.

What, you don't believe him? Look up your own stuff then. This is easily gathered from most published AAMC material.

 

[Melomare17]

What, you don't believe him? Look up your own stuff then. This is easily gathered from most published AAMC material.

When did I say I don't believe him?

I looked it up. It is not completely in there. If you want to find it yourself you can, https://www.aamc.org/students/download/63060/data.

Doesn't say anywhere there that that exam's experimental questions help determine the scale for for that exam. Unless I misread it.

 

[inycepoo]

When did I say I don't believe him?

I looked it up. It is not completely in there. If you want to find it yourself you can, https://www.aamc.org/students/download/63060/data.

Doesn't say anywhere there that that exam's experimental questions help determine the scale for for that exam. Unless I misread it.

Somewhere else AAMC says that the experimental questions are used to determine the scale for when those questions area actually counted towards a future scored exam.