Sampling and Data Collection in Qualitative Studies

Describe the logic of sampling for qualitative studies

Identify and describe several types of sampling in qualitative studies

Evaluate the appropriateness of the sampling method and sample size used in a qualitative study

Identify and describe methods of collecting unstructured self-report data

Identify and describe methods of collecting and recording unstructured observational data

Critique a qualitative researcher’s decisions regarding the data collection plan

Define new terms in the chapter

Key Terms

Data saturation

Diary

Field notes

Focus group interview

Key informant

Log

Maximum variation sampling

Participant observation

Photo elicitation

Photovoice

Purposive (purposeful) sampling

Semistructured interview

Snowball sampling

Theoretical sampling

Topic guide

Unstructured interview

This chapter covers two important aspects of qualitative studies—sampling (selecting informative study participants) and data collection (gathering the right types and amount of information to address the research question).

SAMPLING IN QUALITATIVE RESEARCH

Qualitative studies typically rely on small nonprobability samples. Qualitative researchers are as concerned as quantitative researchers with the quality of their samples, but they use different considerations in selecting study participants.

The Logic of Qualitative Sampling

Quantitative researchers measure attributes and identify relationships in a population; they desire a representative sample so that findings can be generalized. The aim of most qualitative studies is to discover meaning and to uncover multiple realities, not to generalize to a population.

Qualitative researchers ask such sampling questions as, Who would be an information-rich data source for my study? Whom should I talk to, or what should I observe, to maximize my understanding of the phenomenon? A first step in qualitative sampling is selecting settings with potential for information richness.

As the study progresses, new sampling questions emerge, such as, Whom can I talk to or observe who would confirm, challenge, or enrich my understandings? As with the overall design, sampling design in qualitative studies is an emergent one that capitalizes on early information to guide subsequent action.

TIP Like quantitative researchers, qualitative researchers often identify eligibility criteria for their studies. Although they do not specify an explicit population to whom results could be generalized, they do establish the kinds of people who are eligible to participate in their research.

Types of Qualitative Sampling

Qualitative researchers avoid random samples because they are not the best method of selecting people who are knowledgeable, articulate, reflective, and willing to talk at length with researchers. Qualitative researchers use various nonprobability sampling designs.

Convenience and Snowball Sampling

Qualitative researchers often begin with a volunteer (convenience) sample. Volunteer samples are often used when researchers want participants to come forward and identify themselves. For example, if we wanted to study the experiences of people with frequent nightmares, we might recruit them by placing a notice on a bulletin board or on the Internet. We would be less interested in obtaining a representative sample of people with nightmares than in recruiting a group with diverse nightmare experiences.

Sampling by convenience is efficient but is not a preferred approach. The aim in qualitative studies is to extract the greatest possible information from a small number of people, and a convenience sample may not provide the most information-rich sources. However, convenience sample may be an economical way to begin the sampling process.

Example of a convenience sample

Wise (2015) explored pregnant adolescents’ beliefs about healthy eating and food choices. The convenience sample of 14 adolescents was recruited from teen parenting programs.

Qualitative researchers also use snowball sampling (or network sampling), asking early informants to make referrals. A weakness of this approach is that the eventual sample might be restricted to a small network of acquaintances. Also, the quality of the referrals may be affected by whether the referring sample member trusted the researcher and truly wanted to cooperate.

Example of a snowball sample

In a focused ethnography, Martin and colleagues (2016) studied family health concerns from the perspective of adult tribal members residing on an American Indian reservation. A snowball process was used to recruit tribal members.

Purposive Sampling

Qualitative sampling may begin with volunteer informants and may be supplemented with new participants through snowballing. Many qualitative studies, however, evolve to a purposive (or purposeful) sampling strategy in which researchers deliberately choose the cases or types of cases that will best contribute to the study.

Dozens of purposive sampling strategies have been identified (Patton, 2002), only some of which are mentioned here. Researchers do not necessarily refer to their sampling plans with Patton’s labels; his classification shows the diverse strategies qualitative researchers have adopted to meet the conceptual needs of their research:

Maximum variation sampling involves deliberately selecting cases with a wide range of variation on dimensions of interest.

Extreme (deviant) case sampling provides opportunities for learning from the most unusual and extreme informants (e.g., outstanding successes and notable failures).

Typical case sampling involves the selection of participants who illustrate or highlight what is typical or average.

Criterion sampling involves studying cases who meet a predetermined criterion of importance.

Maximum variation sampling is often the sampling mode of choice in qualitative research because it is useful in illuminating the scope of a phenomenon and in identifying important patterns that cut across variations. Other strategies can also be used advantageously, however, depending on the nature of the research question.

Example of maximum variation sampling

Tobiano and colleagues (2016) studied patients’ perceptions of participating in nursing care on medical wards. Maximum variation sampling was used to recruit patients who varied in terms of age, gender, and mobility status.

Sampling confirming and disconfirming cases is another purposive strategy used toward the end of data collection. As researchers analyze their data, emerging conceptualizations sometimes need to be checked. Confirming cases are additional cases that fit researchers’ conceptualizations and strengthen credibility. Disconfirming cases are new cases that do not fit and serve to challenge researchers’ interpretations. These “negative” cases may offer insights about how the original conceptualization needs to be revised.

TIP Some qualitative researchers call their sample purposive simply because they “purposely” selected people who experienced the phenomenon of interest. Exposure to the phenomenon is, however, an eligibility criterion. If the researcher then recruits any person with the desired experience, the sample is selected by convenience, not purposively. Purposive sampling implies an intent to choose particular exemplars or types of people who can best enhance the researcher’s understanding of the phenomenon.

Theoretical Sampling

Theoretical sampling is a method used in grounded theory studies. Theoretical sampling involves decisions about where to find data to develop an emerging theory optimally. The basic question in theoretical sampling is What types of people should the researcher turn to next to further the development of the emerging conceptualization? Participants are selected as they are needed for their theoretical relevance in developing emerging categories.

Example of a theoretical sampling

Slatyer and colleagues (2015) used theoretical sampling in their grounded theory study of hospital nurses’ perspective on caring for patients in severe pain. Early interviews and observations in a renal/hepatology unit provided data on caring for patients who had problems tolerating analgesic medications. The emerging category, labeled “medication ineffectiveness,” guided the researchers to observe in an orthopedic ward where nurses cared for older patients who continued to experience severe pain for months after hip surgery. This theoretical sampling led the researchers to notice differences in nurses’ responses to patients with acute and chronic pain conditions. In turn, this prompted the researchers to sample in the eye/ear/plastic surgery ward where patients were treated for long-term pain.

Sample Size in Qualitative Research

Sample size in qualitative research is usually based on informational needs. Data saturation involves sampling until no new information is obtained and redundancy is achieved. The number of participants needed to reach saturation depends on various factors. For example, the broader the scope of the research question, the more participants will likely be needed. Data quality can affect sample size: If participants are insightful and can communicate effectively, saturation can be achieved with a relatively small sample. Also, a larger sample is likely to be needed with maximum variation sampling than with typical case sampling.

Example of saturation

Van Rompaey and colleagues (2016) studied the patients’ perception of a delirium in a Belgian intensive care unit (ICU). Adult patients in the ICU were interviewed at least 48 hours after the last positive score for delirium. Data collection continued until “data saturation was achieved after interviewing 30 patients” (p. 68).

TIP Sample size adequacy in a qualitative study is difficult to evaluate because the main criterion is information redundancy, which consumers cannot judge. Some reports explicitly mention that saturation was achieved.

Sampling in the Three Main Qualitative Traditions

There are similarities among the main qualitative traditions with regard to sampling: Samples are small, nonrandom methods are used, and final sampling decisions take place during data collection. However, there are differences as well.

Sampling in Ethnography

Ethnographers often begin with a “big net” approach—they mingle and converse with many members of the culture. However, they usually rely heavily on a smaller number of key informants, who are knowledgeable about the culture and serve as the researcher’s main link to the “inside.” Ethnographers may use an initial framework to develop a pool of potential key informants. For example, an ethnographer might decide to recruit different types of key informants based on their roles (e.g., nurses, advocates). Once potential key informants are identified, key considerations for final selection are their level of knowledge about the culture and willingness to collaborate with the ethnographer in revealing and interpreting the culture.

Sampling in ethnography typically involves sampling things as well as people. For example, ethnographers make decisions about observing events and activities, about examining records and artifacts, and about exploring places that provide clues about the culture. Key informants often help ethnographers decide what to sample.

Example of an ethnographic sample

In their ethnographic study, Michel and colleagues (2015) studied the meanings assigned to health care by nurses and long-lived elders in a health care setting in Brazil. The data collection, which involved observations and interviews, relied on the assistance of 20 key informants: 10 nursing professionals and 10 elders.

Sampling in Phenomenological Studies

Phenomenologists tend to rely on very small samples of participants—often 10 or fewer. Two principles guide the selection of a sample for a phenomenological study: (1) All participants must have experienced the phenomenon and (2) they must be able to articulate what it is like to have lived that experience. Phenomenological researchers often want to explore diversity of individual experiences, and so, they may specifically look for people with demographic or other differences who have shared a common experience.

Example of a sample in a phenomenological study

Pedersen and colleagues (2016) studied the meaning of weight changes among women treated for breast cancer. A purposive sample of 12 women being treated for breast cancer at a Danish university hospital were recruited. “Variations were sought regarding age, initial cancer treatment, type of surgery and change in weight and waist” (p. 18).

Interpretive phenomenologists may, in addition to sampling people, sample artistic or literary sources. Experiential descriptions of a phenomenon may be selected from literature, such as poetry, novels, or autobiographies. These sources can help increase phenomenologists’ insights into the phenomena under study.

Sampling in Grounded Theory Studies

Grounded theory research is typically done with samples of about 20 to 30 people, using theoretical sampling. The goal in a grounded theory study is to select informants who can best contribute to the evolving theory. Sampling, data collection, data analysis, and theory construction occur concurrently, and so, study participants are selected serially and contingently (i.e., contingent on the emerging conceptualization). Sampling might evolve as follows:

1. The researcher begins with a general notion of where and with whom to start. The first few cases may be solicited by convenience.

2. Maximum variation sampling might be used next to gain insights into the range and complexity of the phenomenon.

3. The sample is continually adjusted: Emerging conceptualizations inform the theoretical sampling process.

4. Sampling continues until saturation is achieved.

5. Final sampling may include a search for confirming and disconfirming cases to test, refine, and strengthen the theory.

Critiquing Qualitative Sampling Plans

Qualitative sampling plans can be evaluated in terms of their adequacy and appropriateness (Morse, 1991). Adequacy refers to the sufficiency and quality of the data the sample yielded. An adequate sample provides data without “thin” spots. When researchers have truly obtained saturation, informational adequacy has been achieved, and the resulting description or theory is richly textured and complete.

Appropriateness concerns the methods used to select a sample. An appropriate sample results from the selection of participants who can best supply information that meets the study’s conceptual requirements. The sampling strategy must yield a full understanding of the phenomenon of interest. A sampling approach that excludes negative cases or that fails to include people with unusual experiences may not fully address the study’s information needs.

Another important issue concerns the potential for transferability of the findings. The transferability of study findings is a function of the similarity between the study sample and other people to whom the findings might be applied. Thus, in critiquing a report, you should assess whether the researcher provided an adequately thick description of the sample and the study context so that someone interested in transferring the findings could make an informed decision. Further guidance in critiquing qualitative sampling decisions is presented in Box 12.1.

Box 12.1 Guidelines for Critiquing Qualitative Sampling Plans

1. Was the setting appropriate for addressing the research question, and was it adequately described?

2. What type of sampling strategy was used?

3. Were the eligibility criteria for the study specified? How were participants recruited into the study?

4. Given the information needs of the study—and, if applicable, its qualitative tradition—was the sampling approach effective?

5. Was the sample size adequate and appropriate? Did the researcher indicate that saturation had been achieved? Do the findings suggest a richly textured and comprehensive set of data without any apparent “holes” or thin areas?

6. Were key characteristics of the sample described (e.g., age, gender)? Was a rich description of participants and context provided, allowing for an assessment of the transferability of the findings?

TIP The issue of transferability within the context of broader models of generalizability is discussed in the Supplement to this chapter on the book’s website.

Only gold members can continue reading. Log In or Register to continue