An Analysis of the Relationship Between School-Level AP Professional Development Activity and Subsequent Student AP Performance

Skip to Main Content
Publication Information
Publication Thumbnail: 
Abstract: 

The overarching purpose behind this evaluation was to gauge the impact of AP professional development (PD) on AP student outcomes in a state with a significant rate of PD implementation. The evaluation attempted to predict the level of student AP performance by the number of AP professional development events attended by teachers in that school in the prior year, while controlling for some socioeconomic status (SES), teacher, and school effects. The outcomes predicted by the number of PD events attended were defined as the average AP score obtained for that school as well as the percentage of AP Exam takers scoring 3 or above. A similar analysis was also performed for AP courses comprising the STEM disciplines (Biology, Chemistry, Computer Science, Environmental Science, Calculus AB, Calculus BC, Physics B, Physics C: Mechanics, Physics C: Electricity and Magnetism, and Statistics). The controlling factors (covariates) used in the analyses were average household income (a proxy for SES), the percentage of students taking AP in the school (school effect), and the average number of years teaching AP (teacher effect).

The results were as follows:
  • After controlling for average household income (SES), level of AP activity, and teacher experience, schools with higher levels of teachers participating in AP PD were more likely to have higher levels of overall average AP performance (average exam score and average percentage of exams with scores of 3 or above) the following year.
  • In addition to the number of PD events attended, teacher experience was also a statistically significant predictor of subsequent overall AP performance.
  • For STEM-related AP Exams only, the level of AP PD attended by teachers in the school was also a statistically significant predictor of subsequent AP performance. Teacher experience was also significant and more predictive for AP STEM course performance than overall AP performance (reported in the previous bullet).

Given some data access restrictions that did not allow for teachers to be directly linked to their students, drawing direct causal inferences between AP PD and positive student outcomes is not warranted. In addition, there was no information collected regarding implementation of skills learned from the PD. However, the results do point to the potential unique contribution that a school’s level of AP PD activity may have on subsequent student success on AP Exams while accounting for a set of school environmental factors.

Publication Topic: 
Publication Type: