Qualitative purposive selection

How we chose samples for research projects tends to fall into two groups. Probability sampling "is based on the notion that a sample can be selected that will mathematically represent subgroups of some larger population" (Berg & Lune, 2020, p. 38), a method which tends to be used in quantitative research. "Probability samples require that every member of the overall population have a known chance of being included in the sample" (Given, 2008, p. 681). Researchers know they have enough data when they reach saturation, or when a sample is representative of the population being studied. 

Whereas in qualitative research, researchers tend to need participants or situations which will yield rich data. A purposive selection is form of nonprobability sampling (Schwandt, 2007), where skilled participants, well-informed about the phenomenon being researched, are recruited for a research project (Etikan et al, 2016). Researchers use their field understanding to find participants "who represent [the desired] population" (Berg & Lune, 2020, p. 39). This may happen through networks or interest groups where interviewees are specifically sought for what knowledge and expertise they can share on the research topic. 

Purposive selection is "sometimes called judgmental sampling" (Berg & Lune, 2020, p. 39), providing "rich and textured descripti[ve results]" (p. 39). Interestingly, if research gaps are found once data has been collected, a purposive selection can help to "ensure that certain types of individuals or persons displaying certain attributes are included in the study" (p. 39). There are several types of purposive selection: 

• "maximum variation sampling—[...] where a wide range of cases or incidents" are chosen in order "to get variation on the concepts of interest" (Schwandt, 2007, p. 271) 
• "emblematic sampling—choosing a case or incident because it is extreme or deviant, typical or average, or emerging or novel" (p. 271). This is also known as "Extreme or Deviant Case Sampling" (Given, 2008, p. 697)
• "Paradigmatic Case Sampling: A case is paradigmatic when it is considered the exemplar for a certain class", such as using the All Blacks to study elite rugby performance (Given, 2008, p. 697)
• "Typical Case Sampling: Sometimes researchers are interested in cases simply because they are not unusual in any way" - a 'vanilla' sample (Given, 2008, p. 697)
• "Stakeholder Sampling: Particularly useful in the context of evaluation research and policy analysis, this strategy involves identifying who the major stakeholders are who are involved in designing, giving, receiving, or administering the program or service being evaluated, and who might otherwise be affected by it" (Given, 2008, p. 697)

I tend to think of the word 'sample' as needing to be representative, and therefore more probability-oriented. I wonder if that is because so much research is quantitative in nature... ?

Sam

References:

Berg, B. L., & Lune, H. (2017). Qualitative Research Methods for the Social Sciences. Pearson.

Etikan, I., Musa, S. A., & Alkassim, R. S. (2016). Comparison of convenience sampling and purposive sampling. American Journal of Theoretical and Applied Statistics, 5(1), 1-4. https://doi.org/10.11648/j.ajtas.20160501.11

Given, L. M. (Ed.). (2008). The Sage Encyclopedia of Qualitative Research Methods (Volumes 1 & 2). SAGE Publications, Inc.

Meyer, M., & Mayrhofer, W. (2022). Chapter 17: Selecting a Sample. In U. Flick (Ed.), The SAGE Handbook of Qualitative Research Design (pp. 273-289). SAGE Publications.

Schwandt, T. A. (2007). The SAGE Dictionary of Qualitative Inquiry (3rd ed.) SAGE Publications.