<p>Are curves for all exams predetermined before they are administered or are they curved after the exams are scored and then make the bell curve?</p>
<p>If they do it after, how will they accommodate those curves with late-testers?</p>
<p>I am pretty sure that CB calculates scores after each test is graded since they usually want a certain percentage of 5’s etc. And for late testers, there is a form B and a new curve.</p>
<p>The curve is indeed determined after the fact. Late testers have a different curve for their test.</p>
<p>As I understand it, each test has a certain number of control questions in the multiple choice that are in the exact same place on the test, with the exact same answer choices and worded identically. For instance on the AP Calculus exam, there are 6 multiple choice questions that are identical on the 2009 and the 2010 exams. (It’s one of the reasons that they didn’t release the 2008 Calculus exam until after the 2009 exam took place.)</p>
<p>They then use performance on those six questions as one of the factors to determine what the curve should look like.</p>
<p>^Yeah, but if that’s true, then why did they release the 2009 Euro Test long before the 2010 Euro Test took place?</p>
<p>Here’s a scoring guide for 2002 AP Bio: <a href=“Supporting Students from Day One to Exam Day – AP Central | College Board”>Supporting Students from Day One to Exam Day – AP Central | College Board;
<p>On the right, it says:
</p>
<p>What does this mean, and does it have anything to do with the curve?</p>
<p>Can someone explain how performance on previous years’ questions would help determine the curve? Like, if students this year do better, then would the curve be higher?</p>
<p>If students do WORSE, the curve is greater. If everybody did extremely well, there would be a much lower curve.</p>
<p>seekinguni. That means the weights of each section are predetermined, but the cutoffs are not and that those are determined by the chief reader.</p>
<p>
</p>
<p>That’s the gist of it, but the actual analysis they do with the numbers is fairly complex.</p>
<p>
</p>
<p>Not sure. Maybe only certain exams do their comparative analysis this way.</p>
<p>The only other thing I can think of is that there are so gosh-darned many multiple choice questions for some of the exams (like for some of the history AP exams, definitely not for calculus) that they figure no one’s going to take the time to memorize them all. :)</p>
<p>Why would CB determine the curve for the AP test AFTER the test if for all other functions (SATs and SAT IIS) they PREDETERMINE the curve? It doesn’t make any sense.</p>
<p>who cares???
dude the curve is prlly gonna give u like 4-5 extra raw points max so i doubt its gnna affect you that much
i mean honestly the curve is what it is. if i told u its 100 points or -30 points it wouldnt make a difference because ur still gona perform the same way and study the same amount.
you cant change the curve so dont worry about it
=)))</p>
<p>^^ There are a number of reasons why this might happen. The concepts on the SAT are simpler, and it’s significantly easier to generate multiple tests of approximately the same difficulty level. You’ve also got a larger sample size to work with. And there are significantly more score outputs for each individual test (any multiple of 10 from 200 to 800, rather than just 1 to 5). If they were reporting raw scores, they might not worry about it so much.</p>
<p>The folks making the AP Exam don’t worry about making their AP Exams of equal difficulty. They worry about testing the concepts that they want to test. Then they try to make a determination regarding the difficulty level of the exam. While the difference between a 64 and a 65 one year might not be huge in terms of ability, the difference between a reported 4 and a reported 5 very well could be huge in terms of credits earned. So it’s important to try to get that measurement right.</p>