<p>Why do score distributions seem to follow a bell curve for some subjects, like Studio Art-Drawing, whereas other subjects like Comparative Gov have a more or less an even distribution (about 20% get a 5, 20% a 4, 20% a 3, etc...); whereas the sciences tend to have 1/3 of test takers with scores of 1? I can understand AP language score distributions are lopsided on the high end due to native speakers...but what about the other distributions that are not even? And why are there consistently smaller percentages of 5's in certain subjects like AP English Lit? How does CB set scores? </p>
<p>Not sure if this is what you’re looking for, but here’s my theory …</p>
<p>The reason some tests have higher overall scores is because students self-select. At ds’s school, only kids who know they like math are going to take BC Cal and then the test. Not surprising they do so well on the test. But at his school everyone is required to take AP English Lit and Lang, so you end up with a “weaker” pool of test-takers. Not surprisingly, our English AP scores aren’t nearly as impressive as the BC Cal scores.</p>
<p>Well, AP scores aren’t “fixed” like SAT II’s…For APs, if you get a certain score, you get a 5, another score-a 4…and so on. I think some subjects just lend themselves to higher grades</p>
<p>CB sets the score ranges by comparison with the performance of students at selected universities on the same exam questions. Within the past few years, my university was asked to administer a CB exam to our students enrolled in a course that corresponds to an AP subject, make the CB exam scores “count” toward the grades, and then provide information on the grades actually awarded in the class. The cut-off for a 5 is set at the median performance of students who received a grade of “A” in the course. A score of 4 corresponds to the lower half of the “A” grades and the upper half of the “B” grades, etc. The score distributions are therefore affected by the actual performance of college students on the exams.
So, for example, a relatively high fraction of Calc BC students receive scores of 5. There are several reasons for this:
- The students who take Calc BC in high school are, on the whole, stronger math students than those who are taking introductory calculus in college. I realize that a number of students who have taken Calc BC in high school will repeat it in college, but still I think the generalization holds overall.
- The students who take Calc BC in high school are often specifically prepped for the type of questions that appear on the BC exam, in a way that students in a college couse are not. Particularly with the free response, but also with some of the multiple choice, the Calc BC high school students can be helped by “pattern recognition,” beyond the general understanding that would be acquired in a college course that is not geared to the exam. A couple of examples I have in mind are: volume of revolution questions, related rates, and the sorts of fundamental theorem of the calculus questions that often appear as free response question #1. These questions may stump “A” college students if they are unprepared–I think that after a time lapse to allow understanding to sink in, they would be able to handle them all, but prep is an advantage.<br>
Another example is provided by the AP Stat test. There are some formulaic expectations in the free response that an “A” student in statistics would not necessarily adhere to, if not prepped.
With regard to AP Lit, my guess is that the additional maturity of college students and their broader reading has a bigger impact on performance in this subject than most, and also there may be fewer “canned-type” questions.</p>
<p>Youdon’tsay makes a good point, also.</p>
<p>I should clarify–obviously CB asks multiple universities, when they are norming each test.</p>
<p>youdontsay- I never really thought about it, but that seems to be the same here. The HS wants every student to at least “try” an AP class and English Lang seems to be the one thsy push to the kids. The pool of students taking this it is a much more academically diverse group of students than in some of the other AP’s.</p>
<p>In math and science, you either know what you’re doing, or you don’t. This explains the polarity of a lot of math and science grade distributions - lots of 5’s and lots of 1’s. In subjective subjects like history and literature, there is a bigger bulge in the middle (2’s, 3’s, and 4’s) because only really strong history/literature students will be able to write high quality essays that incorporate all of their knowledge, and therefore a small amount of students will be able to earn a 5. This is coming from someone who got 4’s in English Lit and USH, by the way.</p>
<p>Who does the grading on the AP tests?</p>
<p>Kajon, from what I’ve been told by a teacher who did the grading, a combination of high school AP teachers and college professors meet in Louisville, KY to grade AP exams sometime in May/June.</p>
<p>
</p>
<p>That is correct. My husband has done it a few times.</p>
<p>I’m surprised that Human Geo has such a low mean score. I took the test without taking the class, studied intensely for about 3 weeks, and received a 5.</p>