With science education in the United States entering a period of greater accountability, this study investigated how student learning in science was assessed by educators within one state, asking what systemic assessment approaches existed and how the information from them was used. Conducted during the 20o6-2007 school year, this research developed and piloted a network-model case study design that included teachers, principals, administrators, and the state test development process, as well as several state-level professional associations. The data analyzed included observations, interviews, surveys, and both public and private documents. Some data were secondary. This design produced an empirical depiction of practice with a web of related cases. The network model expands on the hierarchical (nested) models often assumed in the growing literature on how information is used in educational contexts by showing multiple ways in which individuals are related through organizational structures.

Seven case study teachers, each employing assessment methods largely unique and invisible to others in their schools, illustrate one set of assessment practices. The only alternative to classroom assessments that could be documented was the annual state accountability test. These two assessment species were neither tightly coupled nor distinct. Some teachers were partners in developing state test instruments, and in some cases the annual test could be seen as a school management resource. Boundary practices—activities where these two systems connected—were opportunities to identify challenges to policy implementation in science education. The challenges include standards, cognition, vocabulary, and classroom equipment. The boundary practices, along with the web of connections, provide the outlines of potential (and often unrealized) synergistic relationships.

This model shows diverse indigenous practices and adaptations by actors responding to pressures of change and persistent historical tensions of diversity and control. It provided evidence of a broadening instructional agenda and rapid deployment of information infrastructures for collection, dissemination, and analysis of student information. The model became a lens to view these changes and paths that policy for science education may take for implementation. It also became a lens to evaluate accountability policies to see how models embedded within policies may fit with current practice.