From professional development for science teachers to student learning in science

This study investigates the effects of professional development for science teachers on student learning. It is usually expected that professional development programs positively impact student learning, however this dimension is not commonly incorporated in the programs evaluation. It is simply assumed that students will be indirectly impacted through their participating teachers in the work with their students. Two main research questions are addressed: 1) Are professional development programs effective in enhancing student learning in science? 2) What are the characteristics of the most and least effective programs? To answer these questions a meta-analysis of 37 professional development programs reporting their impact on student learning was performed. Program characteristics have been defined according to the categories defined by Loucks-Horsley et al (1998), the National Science Education Standards (NRC, 1996), as well as new categories developed by us analyzing other variables such as the programs length. A significant impact of professional development for science teachers on student learning has been found in the form of an overall correlation effect size of r = 0.22 (p<0.001). Moreover, a Fixed Effects Model was used to differentiate between the impacts of the different characteristics of professional development programs for science teachers. In particular, programs emphasizing work on curriculum development, replacement, or implementation, scientific inquiry, pedagogical content knowledge, lasting over 6 month and with a total duration of at least 100 hours have been identified as having a larger impact on student learning. To enhance the findings vignettes have been developed based on the attained effect sizes describing possible professional development programs. Recommendations for present and future professional development programs are made based on what works best in order to maximize their impact on student learning.