Desktop version

Home arrow Engineering arrow Behavioral Intervention Research: Designing, Evaluating, and Implementing


For different quantitative and qualitative strands of mixed methods study designs mentioned earlier, a mix of quantitative and qualitative sampling methods can be used in a single research inquiry. When using mixed methods, investigators need to be aware of different expectations for the rigor of methods used for sampling across quantitative and qualitative approaches. Assumptions about the nature of reality (ontology) and how we know what we know (epistemology), assumptions not typically examined critically by investigators, determine sampling, data collection, and data analysis methods that need to be made explicit for the effective combination of quantitative and qualitative approaches (Creswell & Clark, 2010; Creswell & Miller, 2000; Dellinger & Leech, 2007).

As with any sampling method, sampling in mixed methods must consider not only how sampling was carried out, but why a particular sampling strategy was used, how the sampling strategy was supported by theory, and how the sampling strategy was consistent with the aims of the study (Glaser & Strauss, 1967). As a general rule, sampling strategies for quantitative studies aim to recruit a number of participants to provide sufficient statistical power for the primary outcomes. In addition, quantitative studies seek to sample randomly and systematically with a view to generalize the findings to the population of interest (see Chapter 10 for discussion on sampling). In contrast, the goal of sampling for qualitative strands will not be gen- eralizable to the population; instead there is purposive recruitment of participants who will likely bring valuable insights and perspectives that maximally inform the research question. When considering sample size in qualitative work, keep in mind that often the goal is to obtain a breadth of views (e.g., the scope of a domain), not generalization. Sampling strategies with both goals (representativeness, informativeness) may be deployed in mixed methods in the designs discussed earlier. For example, random sampling may be the foundation for selecting participants for a large-scale survey or intervention study, whereas purposive sampling may be based on participants with specific characteristics (e.g., minority groups, or persons who did or who did not respond to a treatment).

< Prev   CONTENTS   Source   Next >

Related topics