A Companion to Medical Anthropology. Группа авторов

in Tallahassee, FL. The goal of this study was to identify how the experience of racism and other social stressors shapes the risk of high blood pressure among African Americans. The exploratory phase included a round of semistructured interviews with a projected sample size of 48. We arrived at this sample size by identifying four individual attributes related to the experience of racism among African Americans: gender, age, skin tone, and socioeconomic status (SES). We then set quotas for all possible combinations of these attributes, treating them simplistically as dichotomous variables. The resulting sample size allows for comparisons between groups of 8–12 informants with different attributes relevant to experiences of racism. This work led to a new measure of vicarious racism, which we subsequently found to be associated with blood pressure through different biological pathways than for one’s own, direct experience of discrimination (Quinlan et al. 2016).

Table 4.2 Stratified quota sampling design for semistructured interviews on experiences of racism among African Americans in Tallahassee, FL

	Age 25–34	Age 35–54	Age 55–65
	Dark	Light	Dark	Light	Dark	Light	Total
Men
Low SES	2	2	2	2	2	2	12
High SES	2	2	2	2	2	2	12
Women
Low SES	2	2	2	2	2	2	12
High SES	2	2	2	2	2	2	12
Total	8	8	8	8	8	8	48

Guest and colleagues (2006) stimulated an explosion of research on the meaning and measurement of saturation. In subsequent work, they extended the question to focus groups (Guest et al. 2017; see also Hennink et al. 2019) and recently proposed a simple measure for reporting on saturation (Guest et al. 2020). Others have produced their own estimates of minimum sample size (Francis et al. 2010), examined different conceptual dimensions of saturation (Hennink et al. 2016; Weller et al. 2018), developed statistical approaches for estimating saturation (Galvin 2015), and questioned whether saturation is the right criterion at all (Braun and Clarke 2021; Leese et al. 2021; Sebele-Mpofu 2020). Work in this area is certain to continue in coming years.

      DATA COLLECTION AND ANALYSIS

The breadth of medical anthropology makes it impractical to discuss the full range of relevant methods of data collection and analysis. Instead, Table 4.3 summarizes the most commonly used methods of data collection and points to key literature for detail about specific techniques. The table briefly describes each method and the purposes for which it is appropriate. The column labeled E––––––C suggests the types of research questions along the exploratory–confirmatory continuum for which the method is useful. This suggestion is only a rough guide; some methods (e.g., spatial analysis, social network analysis) can be used with varying levels of structure and adapted to a wide range of exploratory or confirmatory questions.

Table 4.3 Commonly used methods of data collection in medical anthropology

Method	Purposes	Procedures	E–––––C	Further Reading
Participant observation	Understand everyday life from insider’s perspective; generate hypotheses; enhance quality of data collection and analysis	Engage in day-to-day life; make systematic record of informal interactions, observations, and conversations	|––––––	(Bernard 2018:Ch. 12–13; DeWalt and DeWalt 2002; Spradley 1980)
Systematic observation	Learn what people do, not just what they say; understand behavior in relation to cultural context	Define variables to be measured; develop sampling strategy; develop systematic rules about what to observe; record behavior	––––––|	(Bernard 2018:Ch. 14; Hames and Paolisso 2015; Borgerhoff Mulder and Caro 1985; Gross 1984)
Interviewing
Unstructured	Understand lived experience from informants’ perspective; build rapport; identify salient issues; discover appropriate language	Develop general interview questions; plan probes to explore for detail; listen actively; keep interview on topic, but encourage informant to lead	–|–––––	(Bernard 2018:Ch. 8–10; Spradley 1979; Gorden 1992; Schensul and LeCompte 2013; Weller 2015)
Semistructured	Balance flexibility and structure; enhance comparisons across informants; best for one-time interviews; develop preliminary hypotheses	Develop interview guide; ensure all topics are covered, but allow flexibility in order and pace of the interview	–––|–––
Structured	Test relationships among items in a cultural domain; test distribution of cultural knowledge; test relations between variables	Develop interview schedule or other stimuli; expose informants to the stimuli that are as identical as possible (e.g., cultural domain analysis, surveys)	––––––|
Elicitation methods
Free lists	Identify contents and boundaries of a Romney 1988) cultural domain; learn which concepts and