Evaluators are continually tasked with making value decisions in the course of study design. In our decisions about implementation fidelity, we place value on specific observations (e.g., self-report, trained observer ratings) and measurement indicators (e.g., dosage, environment, observed use). Each value judgment can strongly impact how a study's implementation fidelity is conceptualized and understood. Yet across fields and within our field, researchers and evaluators tend to hold opposing values in the conceptualization and use of fidelity. As a consequence, there are different relations between fidelity and outcomes, leading to null or significant effects. Drawing on the literature and experiences from two elementary reading and mathematics program efficacy studies, this paper explores differences in fidelity measurements, fidelity variables, and offers recommendations for measuring fidelity. #ProgramTheoryandTheoryDrivenEvaluation #HumanServicesEvaluation #2011Conference #IndependentConsulting #Implementation_Fidelity #GovernmentEvaluation