CONCLUSIONS AND GENERAL RECOMMENDATIONS
Regardless of the phase along the pipeline for the development of an intervention, behavioral intervention research involves dealing with samples. It is too costly, not feasible, and unnecessary to include entire populations to develop and evaluate an intervention. Thus, the selection of a sample is critical at each juncture along the pipeline and needs careful consideration and planning as it can impact the conduct of the research as well as the external validity of the study outcomes. Unfortunately, consideration of the pros and cons of the various methods for selecting a sample is often a neglected part of designing and evaluating an intervention.
The primary issues in sample selection as it concerns behavioral intervention research include the composition of the sample and assuring it conforms to the purpose of the intervention; the size of the sample and assuring it enables an appropriate test of the intervention particularly in comparative studies; and the method of obtaining the sample. Of course, the decisions concerning these three areas should be informed by existing theory, the research goals and questions, the availability of the target population, and feasibility constraints including resources, budget, staffing, and know-how. In this chapter, we summarized various methods for obtaining a sample and highlighted some of the advantages and disadvantages associated with key approaches.
At any stage of the pipeline, it is important to define the target population; review the relevant literature to identify population characteristics that are aligned with the purpose of the intervention (e.g., in a caregiver study, the living arrangement of the caregiver and care recipient may be important; thus, an investigator might want to include only those who live together); determine the necessary sample size; specify the sampling frame (where participants are likely to located); determine the sampling method; review available resources and constraints; and implement the recruitment plan (see Chapter 10). Finally, it is important to be aware of the limitations inherent in a chosen sampling approach and ultimately the sample included in a study, as these will impact the extent to which the findings from a study can be generalized to the target population.