Quota sampling is similar to convenience sampling, but in this case the goal to select a certain quota or number of individuals with a particular characteristic (age group, race/ethnicity, gender, socioeconomic status). It is the nonprobability version of stratified random sampling. As described earlier in the REACH II study, the goal was to enroll equal numbers of Black/African American, Caucasian/White, and Latino/Hispanic caregivers. This approach has some utility in intervention research and can be applied when the research question seeks to examine treatment effects on different discrete populations.
Purposive sampling is a type of nonprobability sampling that serves a specific need or purpose and involves sampling a specific group of individuals such as those who did not do well in a particular training program or caregivers who have participated in select community support groups. It may not be possible to specify the entire population and access to the entire population may be difficult; thus the investigator attempts to include whoever is available from the target group. This type of sampling is referred to as “purposive sampling” because individuals included in the sample fit a specific purpose or description. Purposive strategies are used to enhance understanding of the opinions/experiences of selected individuals or groups participating in an intervention study. It is frequently used when applying a mixed methods approach to understand the processes of adopting an intervention, using imparted strategies, or examining underlying mechanisms (see Chapter 11 on mixed methods).
Generally, three types of cases are optimal in purposive sampling: typical cases, those who are “average”; deviant or extreme cases, those who are at the high or low end of some phenomenon of interest; and negative or disconfirming cases, those who represent exceptions to the rule (e.g., experience an opposite reaction to an intervention) (Devers & Frankel, 2000). For example, in an evaluation of a basic computer skills-training program designed for older adults (Czaja, Lee, Branham, & Remis, 2012) and implemented at several sites, some class participants performed less well than others. In this case, purposive sampling can be used to interview this group of poor performers to gain an understanding of the challenges and difficulties they experienced and how the training program might be redesigned to better serve their needs. As with any sampling strategies, the research questions and goals must be clearly understood prior to the selection of a purposive sample.