III DOES THE INTERVENTION WORK? SELECTING OUTCOMES AND ANALYTICS
Behavioral interventions are intended to have an impact on specified outcomes. Thus, in Part III, we examine the role of measurement and data analyses in building the evidence for an intervention. Chapter 14 discusses the use of measures as outcomes, covariates, mediators, or moderators. Chapter 15 explores the relative advantages of objective measurement strategies and the challenge of demonstrating change in daily functioning, a key outcome for many behavioral interventions. We next discuss data analytic techniques and examine traditional and novel approaches (Chapter 16). Finally, we consider the importance of clinical significance (Chapter 17) and economic evaluations (Chapter 18) to optimize the impact of interventions.
The key “take home” points of Part III include the following:
- ? Measurement serves multiple purposes in behavioral intervention research including evaluation of treatment outcomes, mediators, moderators, covariates, and descriptors.
- ? Reliance on subjective measures alone has disadvantages whereas objective measures can enhance an understanding of important clinical benefits, particularly in the areas of cognition and daily function.
- ? Traditional and emerging novel analytic strategies are important to consider for evaluating intervention effects.
- ? Determining the clinical significance of an intervention is of critical importance and should be evaluated in addition to statistical significance.
- ? Interventions must also be evaluated for their economic value if they are to be implemented in community and clinical settings.