@wizzed101 I read through this whole thread twice and I'm still unsure of what anyone stands to gain from this elaborate scheme.
I do have some questions though...your system has people estimating their ability to score within a certain percentile before they take the exam. How can one reasonably be expected to know their ability to score in a specific range on a new exam with no historical score data available to them? And even if they could, how would you expect schools to handle this ladder system of data? Doesn't a system like this select for people with enough resources and time to sit for multiple exams? I barely had time and resources for one exam myself...I scored well, but had I been forced to predict my ability or face a penalty I would have been irresponsible not to aim lower than my ability to ensure that I wouldn't need to retake and lose further time away from work. Did you consider this burden in the development of this idea?
Sent from my iPhone using SDN mobile app
You just kinda know where you stand via the practice tests and your understanding of the materials. Like I said, this is very hard to code/quantify so I leave it to the power of the human brain. Again, the problem is that of self-serving bias. If you did better than your practice test, you will say that it is your "true" range and vice versa. To force test takers to be honest, wagers must be put in place. Now the concern is that my proposal is full of punishments with no reward. The rewards are the safeguarding of lower brackets and the validation of your decision making and self-awareness, both of which should be looked upon very favorably.
If Kayla has a 512 by picking the 510-515 bracket once, I would argue that she would be looked upon more favorably than Jack going from 475-500, 500-510 then failing 510-515 getting a 514.
As for resources, you don't have to pay if you pass the lower brackets. As for time, it is up to you how many exams you want to take. Plus, the time you spent in the lower bracket, despite not knowing exactly how you performed, are productive. In terms of pure number of hours, I don't see it any differently than taking an official practice test. If you hadn't taken the MCAT and the AAMC released a new scored exam tomorrow, I bet that you will have time to sit through it.
Lastly, for your concern of forcing to pick a lower bracket, I would not worry about it because the 475-500 range is HUGE as are the 500-510 and 510-515. The upper brackets 515-519, 519-523, and 523-528 are very risky, as they should be. A prudent person should be fine, just like "nice" pre-meds...
I hope this will assuage some of your concerns.
Are you stupid? More tests means more questions for us to complete. Even if you're not simply adding the questions to one assessment, you're increasing the number of opportunities "n" that a person has to score within the range "x" that they should be scoring at. My gripe and possibly the gripe of many others on here is you give no support for why your proposal to increase "n" will lead to a more accurate assessment "x" (value) to an extent that will justify multiple administrations of the exam (cost). Keep in mind multiple administrations means more staff so there's no way costs can stay the same. In addition, multiple administrations means students will have to continually prime themselves for the exam. While you make the quasi-purist argument that a true assessment should never be primed, but what else do you think students will attempt to do if your system was implemented?
Also your "stick" actually causes more problems. It will lead to some students risking it, getting poor scores by chance, and then ADCOMs will know they don't deserve that low of a score and then you'll hypothetical system would be trashed. Your whole system sounds ingenuous. Instead of making the assessment more accurate, it sounds more like you're trying to punish those who dared to be good enough to score within a high range by upping the ante on them to perform in a tight window or else be *****-slapped by you back down to a score below yours (which is a terrific score by the way).
To add more questions per section to produce the desired effect, based on the scope and depth of the MCAT, I would estimate about >200~ questions per section. And that wouldn't do for obviously reason.
You should read my previous post again. I didn't say that increasing n would improve anything statistically. Because realistically, I cannot expect an n>3, and that is no better than n=1. As for your point about priming, if by priming you meant they would spend even more time studying, I don't see anything wrong with that. If you meant they would just attune to the test without more preparation, I would disagree. I trust the test. I think every passage and questions and beautifully written and unless you have both the knowledge and the analytical skills, you won't do well. If only knowledge is required, the MCAT would be no different than the GRE and so. If only analytical skills are required, we would be looking at ACT 2.0. Let me make it clear. I would agree with the AAMC that if a test taker had done excellently in 5 physics passages and 3 acid-base passages, they have excellent analytical skills and physics and chem knowledge. But to say that the same person would possess the same level of knowledge in O.chem and biochem is kinda iffy if not absurd. Are we looking the MCAT or ACT 2.0?
The MCAT is like a puzzle except that you have to bring your own pieces. And that is what makes it good. If the goal is to ensure that a test taker has almost all the necessary pieces to solve any puzzle, I don't believe it has been achieved.
As for cost, nothing really changes. As far as an individual is concerned, they just walk in a test center as they used to do. The staffs do the same thing. A 523-526 takes the same exam as a 500-510 in the same room. I don't think creating a new UI and some pamphlets explaining the new system would cost that much. I think the AAMC cares about the integrity of the test, evidently by their counting voids attempts toward the lifetime number when they could've let people void without limit till they get the topics they want and get lots money.
And if a student makes a bad decision while the proposed system is equipped amble safety check, that is on them. The system is designed to counter "chance" to put it crudely. Plus, 475-500, 500-510 and 510-515 and arguably 515-519 are not "tight windows."
You see I have nothing to gain from this. My weakest section was CARS and my system is not designed to touch CARS. I averaged 131 in the other 3 under a self-inflicted condition. Call me haughty, but if I gave anyone the stuffs I gave myself after the C/P section (I realized too late 1 passage into CARS,which I would be anything in the world that I did not miss a single question in that passage-everything was so clear, I was still high from how good I felt after C/P
), I doubt that they would've performed better. I think out of the 4, CARS is the only one that can truly measure what is intended.
I think your system adds too many problems for really no reason at all. But then again, I still have no idea what you're trying to solve, since I'm just seeing more problems than solutions.
Well, the fact is that we'd rather not take the MCAT if we don't have to. So when you asked why, the answer depends on who you ask.
If you asked premeds who just want to get to any medical school and become a doctor, they would rather do away with the test or just replace it with a P/F version.
If you asked the more competitive ones, the MCAT is one way to distinguish themselves, to show that they are the ****.
But I don't have to tell you whose opinions are prevailing the eyes of those that matters. So, on the very surface level, the change should be good because it makes the test every
more competitive. For adcoms, more data are always welcomed. And that's pretty much it. And if you have any doubt about the trend of this whole process,
@gyngyn mentioned in another thread that the AAMC has already conducted research for a different kind of test.