Efficacy studies of educational programs often involve mediator analyses aimed at testing empirically appropriate theories of action. In particular, in the studies of professional development programs, the intervention targets primarily teachers' pedagogical skills and content knowledge, while the ultimate outcome is the student achievement measured by a standardized test. In this case, teaching practices affected by the professional development program in question can be measured in the process of classroom observation, as well as using one or more additional instruments including peer and stakeholder surveys. The main objective of this study is to develop a methodology for creating summative teacher performance metrics from item scores for use as mediators in the analyses of student outcomes. The authors describe using one particular observation rubric, "PLATO" (PLATO, 2014), to answer the question of how a summative teacher performance metric, aligned with a selected measure of student achievement, can be constructed from item scores. [An earlier version of this paper was presented at the Spring 2013 Conference of the Society for Research on Educational Effectiveness (SREE), Washington, DC.]