<p>I have a question. Does anyone know what percent is needed on the AP Calculus AP Test.. to get a 5?</p>
<p>There is no percentage which is stable. They change it each each according to the exam grades of students in college calculus class. The more of them pass, the harder they grade us.</p>
<p>
[quote]
There is no percentage which is stable. They change it each each according to the exam grades of students in college calculus class. The more of them pass, the harder they grade us.
[/quote]
</p>
<p>This is mostly true, but not exactly.</p>
<p>Undergrad students are given early versions of some of the AP questions, but their purpose is to determine AP appropriateness, not to determine the scoring curve.</p>
<p>The scoring curve is determined each year using some overlapping multiple choice questions. There are six questions on the AP Calculus test in 2008 that also appeared on the 2007 test. (This is why, even though the 2008 test is scheduled to be a released exam, the test will not be released until after the 2009 test is given and scored.)</p>
<p>The idea is that students who answer those six questions correctly, as a group, should roughly score the same from year-to-year. (Keep in mind that while this may not be true for individual students, it's true enough for the large group of students who take the exam as a whole.)</p>
<p>For instance, suppose the group of students in 2007 who scored 6/6 scored a 5 on the exam 40% of the time, a 4 on the exam 30% of the time, a 3 on the exam 20% of the time, a 2 on the exam 8% of the time, and a 2 on the exam 2% of the time.</p>
<p>Then suppose in 2008, the group who scored 6/6, using the same cut scores, would score a 5 on the exam 55% of the time, a 4 on the exam 35% of the time, a 3 on the exam 6% of the time, a 2 on the exam 3% of the time, and a 1 on the exam 1% of the time.</p>
<p>Notice that the student scores have increased dramatically, suggesting that the 2008 test is easier than the 2007 test. The cut scores would then be adjusted upward to compensate.</p>
<p>Keep in mind that this isn't the only group looked at, but also the students who scored 5/6, 4/6, 3/6, etc. I'm not sure if they look at which questions were missed and scored correctly.</p>
<p>So why does the percentage of 5's, 4's, 3's, 2's, and 1's change each year? Well, for one thing, it's not possible to do a perfect statistical fit each and every time. For instance, you might find that in order to make a match of percentages, you might need to adjust by one amount for one group of students, but by another amount for the second group of students. So the College Board does the best they can to accomodate this.</p>
<p>There's limited data on how successful you need to be for a given year, as the cut scores are only released when exams are, and the last time that occurred was in 2003. For that exam, you only needed in the low 60% range to get a 5. That number seems surprisingly low to me, but I would venture that if you can get a 70% on the actual exam, you'll most likely be relatively safe.</p>
<p>I encourage my students to shoot for 75%. Some of the free response questions from year-to-year have been increasingly more common (with some noting that 2007's questions seemd a bit more algebraically intensive than the years preceding it), and students tend to be performing better on those questions. That will force the curve upwards (remember that the curve is based on the multiple choice), and accordingly, means that students have to do better to earn the same grade.</p>
<p>That being said, I've only had one student who estimated they got "at least a 75%" in the last few years who actually got a 4, and that student tended to consistently overestimate his abilities. I've found that students who thought they got "about 2/3rds" tended to be about 50-50 on the 4/5 border.</p>
<p>Hope that helps.</p>
<p>MathProf, I am taking BC Calc this year (I took AB last year and got a 5, long story), and I heard from my teacher that in order to get an AB subscore of 5 on the BC exam, you must get a very-near-perfect score on the AB parts (like only missing one question part or something) to ensure a 5 subscore. Is this true?</p>
<p>rb, I don't know the answer to that question, but I have to imagine that it's not true.</p>
<p>What I do know is that the AB subscore is comprised of a certain number of multiple choice questions (I don't know how many), multiplied by a different multiplier than the main AB exam (I believe the multiplier is larger), and that there are three full free response questions and part of a fourth to comprise the other half of the score.</p>
<p>But since I'm pretty sure they're trying to predict how well you would have done on the AB test had you taken it instead, I would imagine you'd have some of the same latitude that the AB students have.</p>
<p>That's just what I would expect to happen, and I don't know for certain.</p>
<p>OK, thanks!</p>