The collection of baseline data from the experimental and confrol group teachers 3 months apart infroduced a potential insfrumentation threat to intemal validity. Ideally, baseline measures for both groups would have been collected at the same time; however, data collection from the control group teachers was not possible until the school year began. Additionally, intervening events, such as other summer inservice for teachers, could have infroduced a history threat to intemal validity. As Onwuegbuzie (2003) cautioned with most educational studies, population and ecological validity threats were concems. Specifically, the use of only two school disfricts and the small sample size of experimental (n = 30) and confrol teachers (n = 29) limit the extemal generalizability of the findings. Nevertheless, in light of the concems about elementary teachers and their failure to understand science processes deeply, the effects due to the STEM Starter intervention were noteworthy. As policy makers work to build a pipeline for science and mathematics talent, STEMStarters and similar programs that provide contentspecific professional development and in-classroom insfructional support are essential for needed improvements in elementary science classrooms.
The lead editors for this article were Anthony J. Onwuegbuzie and John R. Slate.