The role of sampling in qualitative research.

Abstract

Many qualitative researchers state that sample size and sampling are not issues in qualitative research. However, we argue that making sampling and sample size considerations is central to qualitative research. First, we refute arguments made by qualitative researchers who claim that sampling and sample size considerations are not relevant. Second, we contend that sampling represents a multidimensional construct. Third, we posit that most qualitative studies involve some type of analytical generalization. Thus, choosing a sample size and sampling scheme represent an active process of reflection.

In quantitative research, sample size and sampling considerations usually are made with the goal of making statistical generalizations, which involve generalizing findings and inferences from a representative statistical sample to the population from which the sample was drawn. Conversely, because most qualitative research does not involve making statistical generalizations, many qualitative researchers state that sample size and sampling are not issues in qualitative research and that sampling does not explain what is undertaken in qualitative inquiries (Onwuegbuzie & Leech, in press-d). These beliefs have been echoed by many members of a leading qualitative research listserv that has a membership of more than 1,500 researchers (i.e., [e-mail]). Interestingly, using the keywords "qualitative research" and "sampling", as well as "qualitative research" and "sample size," in searching the ERIC (i.e., Educational Resource Information Center) and PsycINFO databases, yielded only four journal articles (i.e., Crowley, 1994/1995; Jones, 2002; Merriam, 1995; Sandelowski, 1995) that discussed the issue of sampling and/or sample size in qualitative research. Thus, it is evident that the concept of sample size in relation to qualitative research has not been discussed or considered as a valid addition for qualitative researchers.

While quantitative researchers use complex mathematical formulae to make sample size considerations, and they promote the use of random sampling (even though the overwhelming majority of studies utilize non-random samples), sample size considerations in qualitative studies are neither mathematical nor systematic. Rather, they involve making a series of decisions not only about how many individuals to include in a study and how to select these individuals, but also about the conditions under which this selection will take place. These decisions are extremely important and, as stated by Curtis, Gesler, Smith, and Washburn (2000), "It seems essential to be explicit about these [decisions], rather than leaving them hidden, and to consider the implications of the choice for the way that the ... study can be interpreted" (p. 1012).

Several reasons have been given by these proponents to support their claims that sampling and sample size considerations are not relevant in qualitative research. In particular, some researchers associate sampling and/or sample size considerations with an obsession with (logical) positivism, which virtually all qualitative researchers reject (Lincoln & Guba, 2000; Schwandt, 2000). Yet, rejecting positivism should not lead qualitative researchers automatically to reject considerations, such as sampling, that are made by quantitative researchers.

Another reason provided for downplaying the importance of making sample size and sampling considerations is that they represent "methodolatry," which refers to having "a preoccupation with selecting and defending methods to the exclusion of the actual story being told" (Janesick, 2000, p. 390). However, it can be argued that providing information to readers about sample size and sampling schemes adds more richness to the story telling. Further, delineating these considerations helps to ensure that qualitative reports are public, as recommended by Constas (1992).

Because qualitative researchers typically are not interested in making generalizations to underlying populations, it is not unusual for qualitative researchers to conclude that sampling is not an issue. Yet, sampling also is important in interpretive research because many qualitative studies, if not most, involve making generalizations. Specifically, qualitative researchers tend to make analytic generalizations (Miles & Huberman, 1994), which are "applied to wider theory on the basis of how selected cases 'fit' with general constructs" (Curtis et al., 2000, p. 1002). In order for analytic generalizations to be richer, the qualitative researcher should collect data that reaches data saturation (Flick, 1998; Morse, 1995), theoretical saturation (Strauss & Corbin, 1990), or informational redundancy (Lincoln & Guba, 1985). If one does select appropriate cases to study (i.e., sampling issue) and an appropriate number of cases (i.e., sample size issue) this should yield thick rich data that reach saturation point.

Consistent with this, Maxwell (1992) defined generalizability in qualitative research as the extent to which a researcher can generalize the account of a particular situation or population to other individuals, times, settings, or context. Maxwell differentiated internal generalizability from external generalizability: the former referring to the generalizability of a conclusion within the setting or group studied, and the latter relating to generalizability beyond the group, setting, time, or context. According to Maxwell, internal generalizability is typically more important to qualitative researchers than is external generalizability.

Interestingly, Stake (2000) noted that "In intrinsic case study, researchers do not avoid generalizations--they cannot. Certainly they generalize to happenings of their cases at times yet to come and in other situations" (p. 439). This suggests that sample size and sampling considerations always are pertinent in qualitative research.

For most qualitative studies, it appears that there are two major issues with sampling, (a) sampling the entire population (e.g., there are only two people in the world who have experienced the phenomenon and interviews will be conducted with both thus, there is no sampling occurring), or (b) taking a sample of the population from which to make generalizations. Whenever our doctoral students inform us that they want to conduct a qualitative study of, say, two administrators, we ask them, "Why two? Why not three or five?" In most cases, the population of administrators is much larger than this, and we question why such a small number of participants has been chosen. Typically, such a question takes our students by surprise. If they are unable to answer this question, we conclude that they have not considered or not reflected sufficiently on their sample size and sampling (e.g., type of purposive sampling scheme). In rare circumstances, we conclude that they have made sample size and sampling considerations if they provide answers such as "because these are the only two administrators in which 1 am interested," "because I think that these two administrators are likely to provide data that will lead to saturation," or "because previous inquiries in this area studied two administrators that led to data saturation." These responses indicate that sampling has been considered; the researcher is not planning on undertaking what is convenient or easy, they have reasons for their sampling schema.

Moreover, what is often ignored is that the concepts of sample size and sampling are multidimensional. Not only do they pertain to cases, but they also pertain to units of data (e.g., interview data, observational data). Thus, for instance, a one-hour interview will yield different amounts and quality of data, and, in turn, should extract more meaning than will a one-minute interview. Therefore, one would expect that a longer interview would be more appropriate if a researcher was interested in a person's life history, than if the researcher was interested in the person's account of a specific event. In fact, in addition to reflecting about how many cases to sample and how to select this sample, qualitative researchers should make sampling decisions such as how many interviews or focus groups to conduct, how long each interview or focus group should be, how many sets of observations to conduct, and how long each observation period should be. These decisions should not be automatic but should result from adequate reflection. Further, these decisions should be made with the goal of attaining prolonged engagement and persistent observations, as advanced by Lincoln and Guba (1985). Thus, prolonged engagement and persistent observations represent sampling concepts. If we do not sample enough of these observational units or textual units, the quality of our data will be affected, and our data will not be sufficiently rich and thick, making it more difficult to find meaning.

Consistent with our assertions regarding the importance of making sample size considerations, a few methodologists have provided sample size guidelines for several of the most common qualitative research designs and techniques. Specifically, Creswell (2002) has recommended that 3-5 participants be used for case study research. Also, with respect to phenomenological studies, sample size recommendations range from 6 (Morse, 1994) to 10 (Creswell, 1998). For grounded theory research, sample size guidelines have ranged from 15-20 participants (Creswell, 2002) to 20-30 participants (Creswell, 1998). With regard to ethnographic research, Morse (1994) has recommended that 30-50 interviews be conducted. Finally, with regard to the use of focus groups, the following recommendations have been made: 6-9 participants (Krueger, 2000); 6-10 participants (Langford, Schoenfeld, & Izzo, 2002; Morgan, 1997); 6-12 participants (Johnson & Christensen, 2004); 6-12 participants (Bernard, 1995); 8-12 participants (Baumgartner, Strong, & Hensley, 2002). In general, as noted by Sandelowski (1995), sample sizes in qualitative research should not be too small that it is difficult to achieve data saturation, theoretical saturation, or informational redundancy. At the same time, the sample should not be too large that it is difficult to undertake a deep, case-oriented analysis.

Moreover, it should not be assumed, as many researchers mistakenly do, that qualitative research studies always involve the use of small samples. In fact, qualitative research can utilize large samples, as in the case of program evaluation research. Moreover, to associate qualitative data analyses with small samples is to ignore the growing body of literature in the area of text mining--the process of analyzing naturally occurring text in order to discover and capture semantic information (see for example, Del Rio, Kostoff, Garcia, Ramirez, & Humenik, 2002; Liddy, 2000; Powis & Cairns, 2003; Srinivasan, 2004).

Further, as contended by Onwuegbuzie and Leech (in press-d), in most qualitative investigations, one or more of the following generalizations are made: (a) from the sample of words to the voice; (b) from the sample of observations to the truth space; (c) from the words of key informants to the voice of the other sample members; (d) from the words of sample members to those of one or more individuals not selected for the study; or (e) from the observations of sample members to the experience of one or more individuals not selected for the study. Each of these types of generalizations necessitates sampling decisions.

Within any particular qualitative study, sampling often may represent an iterative process, as is particularly the case in grounded theory and ethnographic studies. Indeed, according to The American Heritage College Dictionary (1993, p. 1206), a sample is "a portion, piece, or segment that is representative of a whole." Also, sampling is "an act, process, or technique of selecting an appropriate sample" (p. 1206). Thus, sampling is a concept that transcends research studies in general and research paradigms in particular (Onwuegbuzie & Leech, in press-a, in press-b, in press-c, in press-d).

Although there are some qualitative researchers who are uncomfortable with the use of the word "sampling," we believe that we should continue to use this term in qualitative research. First and foremost, getting rid of this term will not get rid of the fact that sampling takes place in qualitative research, as we have attempted to illustrate above. Second, replacing this term with another term likely would lead to more confusion among researchers. Third, this term has a long tradition in qualitative research. For example, the concept of purposive sampling has been around for a long time. Also, for decades, grounded theorists have used the term "theoretical sampling," which involves the sampling of additional people, incidents, events, activities, documents, and the like, in order to develop emergent themes, to assess the adequacy, relevance, and meaningfulness of themes, to refine ideas, and to identify conceptual boundaries (Charmaz, 2000). Fourth, as mentioned previously, the dictionary definition of the term "sampling" is consistent with how it is operationalized in qualitative research.

Choosing a sample size and sampling scheme should represent an active process of reflection that is based on many factors, including the context, method of collecting data, and type of generalization (if any) needed. As such, we believe that making sampling and sample size considerations should be an integral part of the qualitative research process.

References

Baumgartner, T. A., Strong, C. H., & Hensley, L. D. (2002). Conducting and reading research in health and human performance (3rd ed.). New York: McGraw-Hill.

Bernard, H. R. (1995). Research methods in anthropology: Qualitative and quantitative approaches. Walnut Creek, CA: AltaMira.

Charmaz, K. (2000). Grounded theory: Objectivist and constructivist methods. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 509-535). Thousand Oaks, CA: Sage.

Constas, M. A. (1992). Qualitative data analysis as a public event: The documentation of category development procedures. American Educational Research Journal, 29, 253-266.

Creswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage.

Creswell, J. W. (2002). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. Upper Saddle River, NJ: Pearson Education.

Crowley, E. P. (1994/1995). Using qualitative methods in special education research. Exceptionality, 5, 55-70.

Curtis, S., Gesler, W., Smith, G., & Washburn, S. (2000). Approaches to sampling and case selection in qualitative research: Examples in the geography of health. Social Science and Medicine, 50, 1001-1014.

Del Rio, J. A., Kostoff, R. N., Garcia, E. O., Ramirez, A. M., & Humenik, J. A. (2002). Phenomenological approach to profile impact of scientific research: Citation mining. Advances in Complex Systems, 5(1), 19-42.

Flick, U. (1998). An introduction to qualitative research: Theory, method and applications. London: Sage.

Janesick, V. J. (2000). The choreography of qualitative research design. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 379-399). Thousand Oaks, CA: Sage.

Johnson, R. B., & Christensen, L. B. (2004). Educational research: Quantitative, qualitative, and mixed approaches. Boston, MA: Allyn and Bacon.

Jones, S. R. (2002). Writing the word: Methodological strategies and issues in qualitative research. Journal of College Student Development, 43, 461-473.

Krueger, R. A. (2000). Focus groups: A practical guide for applied research (3rd ed.). Thousand Oaks, CA: Sage.

Langford, B. E., Schoenfeld, G., & Izzo, G. (2002). Nominal grouping sessions vs. focus groups. Qualitative Market Research, 5, 58-70.

Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.

Lincoln, Y. S., & Guba, E. G. (2000). Paradigmatic controversies, contradictions, and emerging confluences. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 163-188). Thousand Oaks, CA: Sage.

Liddy, E. D. (2000). Text mining. Bulletin of the American Society for Information Science & Technology, 27(1), 14-16.

Maxwell, J. A. (1992). Understanding and validity in qualitative research. Harvard Educational Review, 62, 279-299.

Merriam, S. B. (1995). What can you tell from an N of 1?: Issues of validity and reliability in qualitative research. PAACE Journal of Lifelong Learning, 4, 51-60.

Miles, M., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage.

Morgan, D. L. (1997). Focus groups as qualitative research (2nd ed.). Qualitative Research Methods Series 16. Thousand Oaks, CA: Sage.

Morse, J. M. (1994). Designing funded qualitative research. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 220-235). Thousand Oaks, CA: Sage.

Morse, J. M. (1995). The significance of saturation. Qualitative Health Research, 5, 147-149.

Onwuegbuzie, A. J., & Leech, N. L. (in press-a). Enhancing the interpretation of "significant" findings: The role of mixed methods research. Qualitative Report.

Onwuegbuzie, A. J., & Leech, N. L. (in press-b). On becoming a pragmatist researcher: The importance of combining quantitative and qualitative research methodologies. International Journal of Social Research Methodology: Theory & Practice.

Onwuegbuzie, A. J., & Leech, N. L. (in press-c). Taking the "Q" out of research: Teaching research methodology courses without the divide between quantitative and qualitative paradigms. Quality & Quantity: International Journal of Methodology.

Onwuegbuzie, A. J., & Leech, N. L. (in press-d). A call for qualitative power analyses. Quality & Quantity: International Journal of Methodology.

Powis, T., & Cairns, D. (2003). Mining for meaning: Text mining the relationship between social representations of reconciliation and beliefs about Aboriginals. Australian Journal of Psychology, 55, 59-62.

Sandelowski, M. (1995). Focus on qualitative methods: Sample sizes in qualitative research. Research in Nursing & Health, 18, 179-183.

Schwandt, T. A. (2000). Three epistemological stances for qualitative inquiry: Interpretivism, hermenueutics, and social constructivism. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 189-213). Thousand Oaks, CA: Sage.

Srinivasan, P. (2004). Generation hypotheses from MEDLINE. Journal of the American Society for Information Science & Technology, 55, 396-413.

Stake, R. E. (2000). Case studies. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 435-4454). Thousand Oaks, CA: Sage.

Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage.

The American Heritage College dictionary (3rd ed.). (1993). Boston, MA: Houghton Mifflin.

Anthony J. Onwuegbuzie, University of South Florida Nancy L. Leech, University of Colorado at Denver and Health Sciences Center 

Anthony J. Onwuegbuzie is an associate professor in the Department of Educational Measurement and Research. Nancy L. Leech is an assistant professor in the Department of Educational Psychology.