Studies of decision-making in real-world settings provide an arena for addressing the question: Why do people do what they do? For cognitively oriented ethnographic studies, a frequent starting assumption is that in recurring decision situations where alternative courses of possible action exist, members of a group come to have shared understandings, a common set of standards concerning how such choices are made (Goodenough, 1963, pp. 265-270; Quinn, 1978; Young & Garro, 1994). Cognitive-ethnographic studies of medical decision-making seek to understand what people do when faced with illness and typically attempt to account for actions taken to deal with illness. To gain insight into the relationship between cultural knowledge and specific treatment actions, careful consideration is given to how people talk about treatment decisions. Such studies are concerned with the nature of cultural knowledge brought to the occurrence of illness, how this knowledge is applied in evaluating illness, and the process whereby decisions about treatment are made. Some, although by no means all decision-making studies are overtly concerned with developing and testing decision models. When attempted, a common strategy is to build a decision model using interview data and other information obtained primarily from one sample, and then to validate the model using decisions made by a second, independent sample. Here, reference to a decision-making approach or perspective serves an inclusive higher-level category, with a "decision model"—the more formal representation of the decision-making process that can be evaluated using actual choices—as a subcategory. Garro (1998a) presented a review and counter-perspective to several broadly based critiques leveled at anthropological studies of care-seeking and decision-making.
Rather than attempt a comprehensive discussion of similarities and differences among relevant studies (e.g., Bauer & Wright, 1996; Hill, 1998; Kayser-Jones, 1995; Mathews, 1982; Mathews & Hill, 1990; Nardi, 1983;
Ryan & Martinez, 1996; Sargent, 1982, 1989; Stoner, 1985; Weller, Ruebush, & Klein, 1997), discussion is limited to two studies—one carried out in the Mexican town of Pichataro (Garro, 1998a, Young & Garro, 1994) and the other in an Anishinaabe (Ojibway) community in Manitoba, Canada (Garro, 1998b). While the two settings differed radically in terms of the political-economic context of health care, both studies were concerned with how families make treatment decisions among different alternatives.
In both studies an objective was to discover if there were generally shared considerations and cultural understandings that helped to explain variability in each community. At each site there was variability in patterns of resort to different treatment alternatives at different times, both within and between households. The comparability of the two studies was facilitated by adopting the same overall research design in both, despite the methodological challenges and substantive issues involved in adapting the general data-collection strategy to such quite different research settings. In addition to participant observation, informal talks, and interviews, there were two main phases. First, a group of individuals within each community participated in a series of structured interviews organized to learn about local medical understandings and how medical treatment decisions were made when a family member was ill. These interviews served to elucidate the shared understandings and constraints on preferred actions relevant to treatment decisions. The term-frame interview in Pichataro discussed in the preceding section was part of this process. In Pichataro, interviews designed to learn about patterns of care-seeking included systematic contrastive questioning about the use of treatment alternatives, presenting hypothetical illness situations designed around contrasts mentioned in earlier interviews and asking individuals what should be done, and recording family-based histories of past illness. In addition, ranking tasks were used to explore assessments of "faith" in different treatment alternatives for a set of illnesses. Pile sorts and ranking tasks obtained judgments of severity for separate sets of illnesses and symptoms. This phase of the research design took a somewhat different form in the two settings, to be expected as the format and content of structured interview methods depend upon prior ethnographic research. In the Anishinaabe community, for example, considerably more attention was given to etiological concerns. However, some approaches, such as the term-frame interview that worked well in Pichataro, did not transfer to the Anishinaabe community (see Garro, 2000b, for details).
For both communities, Garro (1998a) maintained that a decision-making perspective was compatible with how individuals talked about actions taken in response to illness and proved to be a useful means for learning about the process of seeking care. Still, as commented on again below, Garro concluded that the stages of developing and testing a formal decision model were better suited to the Pichataro context than to the Anishinaabe community. For Pichataro, two basic strategies or general principles underpinned the decision model which specified how assessments of a given illness and judgments about appropriate sources of treatment resulted in expected patterns of treatment choice. But the model also detailed the constraints, such as lack of money or transportation, that at times led to less preferred treatment alternatives being selected.
The second major phase involved collecting actual illness case histories and treatment decisions from a separate, randomly selected group of families in each community through regular household visits over an approximately half-year period. These case studies provided a basis for assessing the relationship between what people say they do and what they actually do. For both sites, the case studies provide independent confirmation for the depiction of the decision-making process from the first phase of the project. This correspondence can be more clearly seen in the Pichataro study where the decision model successfully predicted a high proportion of both initial and subsequent treatment choices, approximately 90% of all treatment actions (80% if initial home treatments, which can be seen as a "largely routine initial response," are excluded from the calculations). Further, what the data indicated was that it was not the case that community members generally felt that biomedical treatment was incompatible with their own understandings about illness—such instances do exist but they are relatively uncommon—nor did they believe (for the vast majority of cases) that biomedical treatment was less likely to result in a cure than the other alternatives. Instead, the observed patterning in treatment actions can be accounted for by the relative inaccessibility of physician services, including the high costs of obtaining such care and transportation difficulties. A subsequent comparative study examining use of physician services in Pichataro as well as another nearby, culturally similar community with much better access to physician services tested the prediction that better access would result in much higher rates of physician utilization (Young & Garro, 1982). And this is what was found. The study was also designed to examine a plausible alternative explanation for the observed differences in patterns of resort, namely that the second village had a much stronger biomedical orientation with regard to illness understandings. Support for this hypothesis was not found, however, as no significant differences in illness understandings for the two communities were discovered through an analysis comparing responses to two different structured interview methods (a term-frame interview and a triads comparison task). It was not a difference in cultural knowledge, including the adoption of biomedical understandings, that accounted for the increased rates of physician utilization in the second community, but rather a reduction in the financial and structural constraints impeding care-seeking from physicians.
Still, even for studies like the one in Pichataro where a decision model is seen to provide a reasonably good guide to an understanding of treatment actions and the culturally based rationality that underlies them, insufficient consideration has been given to the jointly cultural, social, and cognitive constructive processes through which meaning is conferred upon perceived afflictions (Garro, 1998a). To move in this direction would involve attending to how signs of trouble come to be "framed" in the context of everyday life—by individuals and in social interactions—as certain types of illness amenable to certain kinds of treatment. This framing of a problem is integral to the decision-making process but is often treated as given in decision modeling studies (cf. Mathews, 1987). Multiple possibilities may be entertained as the framing process unfolds in time, drawing upon cultural and personal knowledge, while often taking shape within social interactions and in relation to ongoing "social activity" (cf. Frake, 1961).
The import of this framing process in studies of illness treatment decision-making relates to the claim that developing and testing a decision model is better suited to accounting for treatment actions taken in the Mexican community than in the Anishinaabe community (Garro, 1998b). It is the framed situation, not the process of speculating or converging on a particular framing, that decision models are set up to handle. This precondition of a framed illness was met in the Pichataro study, where diagnoses and other efforts characterizing illness provide a guide to the presumed cause and suitable forms of care for an illness at a given point in time. In contrast, a greater degree of interpretive openness and ambiguity often accompany illness in the Anishinaabe community. There was an openness to framing the "same" condition in quite variable ways—an individual may juggle alternative possible framings with divergent implications for care-seeking, and family members may have different interpretations of what is going on. In addition, there was a variable potentiality for illness and its treatment to be understood within a cultural framework of Anishinaabe sickness a framework in which the underlying assumptions are distinctive from those implicit in studies of decision modeling (but which do not extend to studies of decision- making more generally). Decision models entail tacit assumptions: the problem prompting care-seeking is localized to an individual; illness episodes are discrete and bounded in time; there is a decision-making entity (person or group of persons) whose decisions are to be modeled. But these tacit cultural assumptions underlying the decision modeling approach cannot accommodate all aspects of the version of reality that often guides decisions in the Anishinaabe community. These tacit assumptions will be met in many settings; the Pichataro study is one example. In the Anishinaabe community, however, these assumptions did not apply to an appre-ciable proportion of cases involving consultations with Anishinaabe healers. While research in both sites affirmed the broad applicability of a cognitive-ethnographic decision-making perspective to the understanding of actions taken in response to illness, the possibility remains that a decision-making approach may not be as productive in other cultural settings or that it may not even be applicable.
Was this article helpful?