• children's mental health;
  • evidence-based treatments;
  • service system;
  • treatment integrity

This article presents a broad-based framework for evaluating integrity, defined as the structured comparison of observed values (i.e., what did happen) with expected values (i.e., what should happen), across various domains (e.g., practices, outcomes) for the purposes of informing key questions in system contexts (e.g., Is a practice being done properly? How well is it working?). We discuss the application of integrity in serving the goals of the scientific and service systems and potential sources for observed and expected values. An expanded framework regarding integrity analysis is presented detailing multiple domains of measurement (e.g., outcomes, activities, and resources), types of analysis (e.g., across vs. within units), and levels of analysis (e.g., event, episode, multi-episode). Other possible applications of integrity analysis in service systems and ongoing dissemination and implementation efforts are also discussed.