ABSTRACT
Behavioral scientists can utilize a range of quantitative and qualitative approaches to analyze data and make conclusions. They can be used to consider whether behavior change interventions work, how they work and when they work. In this chapter, we overview various methods of analysis highlighting when these different approaches can be used. In addition, practical guidance and issues relating to randomization, recruitment, attrition, linear and no-linear relations, clustered data, choosing the right sample size and how best to handle missing data and outliers are also covered. As well as considering whether health behavior change interventions work, how they work is also a critical question. Process evaluations provide additional information relating to how well the intervention is implemented, potential mechanisms, and context that can influence, positively or negatively, the effectiveness of the intervention. As such, in this chapter we cover process evaluations, mediation, moderation and RE-AIM as a potentially useful framework to consider effectiveness and process-related factors to estimate the potential public health benefits that may accrue from a particular health behavior intervention.
