ABSTRACT
Implementation fidelity in experimental settings is a priority for funders of educational research. However, there is little guidance on how to include fidelity-related data in efficacy or effectiveness studies conducted in classrooms and schools. In this chapter, we describe strategies for measuring fidelity and including fidelity-related data in randomized trials of innovative educational practice. We illustrate our recommendations using recent experiences from the Promoting Acceleration of Comprehension and Content through Text (PACT) project, conducted by the Meadows Center for Preventing Educational Risk at the University of Texas with funding from IES’ Reading for Understanding initiative (see https:// files.eric.ed.gov/fulltext/ED551445.pdf). PACT comprised a series of studies, many of which were efficacy-focused and group-randomized. For this subgroup of studies, we describe the program model, the study design, and the measurement strategy. The PACT work operationalized fidelity as treatment dosage, which we define in terms of adherence and crossover (or diffusion), where adherence is teachers’ use of the intended program when in treatment-assigned classes and crossover is teachers’ use of the intended program in business-as-usual classes. Finally, we discuss the use of structural equation models to estimate dosage as a mediator of the effect of assignment on outcomes.
