<p>* So if you really have issues with the chart, go ahead and e-mail him.*</p>
<p>I don't have issues with the overall thrust of Jay's message that challenging high school classes are the best way to prepare for college. </p>
<p>And, as a practical matter, for the typical public high school, AP or IB courses are probably the most accessible route for challenging students. (The private schools that are abandoning AP have a faculty with very high levels of education and a low student-faculty teacher ratio that frees up time for inventing their own advanced curricula, tailored to the needs, interests, and backgrounds of their students. It's also interesting to note that their students typically take the AP exams ANYway, even though they may not take AP courses to prepare for them. Their generally strong intellectual preparation along with some self-study often enables them to do quite well, despite not taking classes targetted specially towards the test. Homeschoolers often do the same thing, by the way. They often read very widely and write very prolifically in a self-designed program that doesn't particularly relate to APs and then, down the road, a bit of structured study aimed at filling in a few gaps, may enable them to do very well on the exams.)</p>
<p>And I don't have a huge problem with Jay's use of the chart--he's a journalist, not a psychometrician, and his readers will presumably take that into account. </p>
<p>What I DO have a problem with is the head of the College Board's AP program uncritically endorsing this analysis in his memo publically posted on the web in several places, without any mention of the statistical caveats that any responsible psychometrician should attach to such an analysis.</p>
<p>It seems to me that the College Board has a responsibility to set a better example of the responsible use of statistics to draw conclusions from data!</p>
<p>Again, I want to stress that I don't necessarily disagree with the end conclusion that APs (or IBs) are a valuable preparation for college, just the uncritical use of statistics to support that conclusion.</p>
<p>And, by the way, it is entirely possible that a more complex study might demonstrate that students who take IBs have higher success rates than students who take APs (or vice versa)....and I suspect that whichever organization came out on the bottom of such a study would be the first to point out the statistical caveats that need to be applied in interpreting the data. The populations of students who CHOOSE to take a given course or exam may be different a priori.</p>