This month, College Board officials released the latest data on the Advanced Placement program,  noting record increases in the numbers of students taking AP courses and scoring well enough on the exams to get college credit. The AP program saves students "time and tuition," said Gaston Caperton, president of the College Board. The Bush administration is climbing on the AP bandwagon as well, calling for more students  to take the courses in high school.
There's just one problem, according to research presented Friday in St. Louis at the annual meeting of the American Association for the Advancement of Science: AP courses -- whatever their merits -- may be poor substitutes for college courses in the sciences.
The study looked at 18,000 students in introductory biology, chemistry and physics courses in college. The students were at 63 randomly selected four-year colleges and universities and their performance in the courses was correlated to various factors. The researchers found that students who had taken AP courses -- even those who had done well on the AP exams -- did only marginally better than students who had not taken AP courses. Other factors, such as the rigor of mathematics taken in high school, were found to have a strong impact on whether students did well in college-level work in the sciences.
"Our survey, the largest ever of its type, suggests that AP courses do not contribute substantially to student success in college," said Philip M. Sadler, director of science education at the Harvard University-Smithsonian Center for Astrophysics and a senior lecturer in astronomy at Harvard. "Even a score of 5 on an AP test is no guarantee of a college grade of A in the same subject," he said.
Sadler conducted the study -- which is being disputed by the College Board -- with Robert H. Tai, assistant professor of science education at the University of Virginia.
They found that students who received a 5 on an AP exam in science and then took the same material in college averaged only 90 -- even after the additional year of college study in an area in which they had presumably already demonstrated A-level work. Students who received an AP score of 4 averaged 87 and those who received an AP score of 3 received an average grade after a full year of college science of only 84 -- compared to 82 as the average for students who had taken honors science courses in high school, but who had not taken AP science classes. Sadler said that if the AP system lived up to its billing (in which scores of 3 are recognized by many colleges as indicating successful completion of college-level work), the students with AP scores of 3-5 should have done much better in college.
Other factors, having nothing to do with AP, did have a strong impact on success in college science: completing strong math courses and high school science courses that were not "drained of math," a high school curriculum that stressed "depth over breadth," and laboratory experience in high school in which the outcome of experiments is not known in advance and students must write up their results.
Of course, some of those qualities may well be present in AP courses, and Sadler stressed in an interview that he did not want his findings to be viewed as "anti-AP." His son is taking AP courses now, and Sadler is pleased that he's doing so. He's just dubious that his son should skip any college courses on the basis of AP -- and that gets to the central strengths and weaknesses with the program, he said.
"AP courses are very helpful to students, and they give a lot of students some heavy duty things to do in high school and they learn a lot," Sadler said. For many high schools, he said, AP courses have become a highly effective way of engaging talented students in their senior year -- something that has been a struggle for many high schools.
But while AP may function well as enriched content, it doesn't equal college-level work, Sadler said, and shouldn't be promoted as such. If the College Board wants to promote the AP curriculum as a way to allow students to receive credit for some college courses, Sadler saw two options: Make the tests significantly more difficult, or create new scores of 6 or 7, above the current top score, and let only the absolutely top performers with such scores earn college credit. Either way, he said, his research suggests that the vast majority of those now achieving scores indicating that they have done college-level work shouldn't be receiving such scores.
Jennifer Topiel, executive director of public affairs for the College Board, questioned the size of the study, noting that the 18,000 students included many who had never taken an AP course and that many of those who had didn't receive high scores. She said that College Board data paint a different picture, in which students who score well on AP exams do well not only in introductory courses, but for those who place out of intro courses, in the more advanced classes. "There's a lot of research that shows the exact opposite of what they are saying," Topiel said.
Topiel added that the College Board works to keep AP test questions at the right level by giving sample questions to college students in introductory courses, and studying which questions A students, B students and so forth get correct. She also noted that the test questions are developed by college professors.
Sadler said, however, defended his study. He said that if students with 4's and 5's on the AP exam were not acing introductory courses, there was no reason to think they would be doing better in more advanced courses.
While his research challenges some conventional wisdom about AP courses, Sadler said that he has been struck by who has not been surprised at all by the results: professors who teach science. "They've known this for a long time," Sadler says, and if they had their way, they wouldn't award credit for AP (or would do so much more minimally). "They just haven't had the data."