That said, it is important to note that many CBPR projects focus on a single site, which may limit generalizability [23, 24]. Moreover, population-level policy and practice interventions, which often have more “reach” than typical individually or group-focused interventions, often benefit from, or require, inclusion of multiple communities. For example, community trials often include a number of design features that enhance internal, external, and statistical conclusion validity (such as multiple sites, randomization, and explicit accounting for nested data) (e.g., [25, 26]; see Boruch, [27] for an overview of methodological issues). These trials may provide solid designs for assessing the effects of interventions aimed at community change and may involve some level of community participation in designing and implementing interventions. However, such participation typically does not extend to the choice of the problem, study design and methods, and interpretation and dissemination of results. Moreover, as typically practiced, CBPR seldom involves pursuit of policy change—particularly changes in state or local public policy—as an intervention strategy [28]. And when CBPR is used to pursue policy change, the focus is typically a single community, limiting generalizability [28, 29].
