Ethnographic methods: specific research problems

1. Ethnographic research is used to define a research problem

Certain well-established research topics attract the researcher because of their extensive bodies of associated literature, which make it possible to formulate rea­sonable working hypotheses that can then be tested using focused data collection tools. Other topics, by contrast, are more amorphous and need to be studied on the ground, as it were, before suitable hypotheses can be devised. It is for these latter topics that ethnographic methods are particularly well suited.

For example, in the Trinidad project, the Indian indenture in various parts of the old British Empire had been extensively studied by historians, economists, political scientists, sociologists, and social psychologists, as well as cultural anthropologists. Especially with regard to the West Indies, however, there had been at the time I began my research a tendency to focus on the most isolated and culturally traditional Indian communities. But Trinidad, with a modern industrial sector linked to the global petrochemical economy, provided many potential oppor­tunities for Indians to break out of their traditional isolation. And indeed many of them had done so. Younger people were taking jobs in the non-agricultural sector, getting higher education, and living in homes outside the rural villages. But from what I had heard before doing field research of my own, I knew that the sense of Indian community identity remained very strong. What was going on in this transitional society? In what ways did Indians themselves understand the dynamic of living modern lives and yet defining themselves in terms of cultural tradition?

In the deinstitutionalization project, people who are mentally challenged have obvious difficulties when it comes to negotiating the complexities of everyday life. Such scholarly literature as existed at the beginning of the deinstitutional­ization process suggested that those who made it outside the hospital were those who either established or otherwise fell into the care of agencies providing comprehensive ‘case management’ services or of compassionate individual bene­factors. There seemed to be no choice: either give up the promised freedoms of deinstitutionalization in order to secure the protection of a benevolent other, or fail to cope and become a hopeless, homeless vagrant. But did people in this situation really see things in such either/or terms, or had they found other ways in which they could cope?

In both these research projects, the main question asked by the researcher was: ‘What does it feel like to be [a modern-day Trinidad Indian; deinstitutionalized adult with mental challenges]?’ This is obviously a less clear-cut question than one that could be answered with demographic statistics (‘How many people were brought to Trinidad during the indenture?’ ‘What percentage of the modern pop­ulation of Trinidad is Indian, and where on the island do they live?’) or epidemi­ological data (‘How many people are diagnosed with severe mental illness?’ ‘What are the main behavioral symptoms associated with mental retardation?’) Answering it required the researcher to participate in the lived experience of the people under study, and not simply to observe it from a detached position.

2. Ethnographic research is used to define a problem that cannot immediately be expressed in ‘If X, then Y terms and that seems to result in behaviors that would not have been predicted by the existing literature

Standard quantitative research is predicated on the assumption that problems can best be studied if they can be stated in terms of a predictable relationship: dependent variables (factors that change) when an independent variable (a factor that seems to be a predisposing condition) is present. But sometimes, real-life problems are difficult to fit into such a testable format, at least at first.

For example, there seemed to be an unusually high rate of alcoholism among the Trinidad Indians, a fact noted with some surprise in the literature. The tradi­tional religions of the indentured Indians (Hinduism and Islam), as well as the version of Christianity offered to them by missionaries during the colonial period, were strongly anti-drink. Why, then had the Indians – who professed to be so traditional in their cultural affiliations – become problem drinkers? There were possibly historical factors: some historians suggested that the indentured planta­tion laborers were paid in rum – a main product of the sugar estates in that era. There were also possible explanations of a psychological nature: a disenfran­chised minority tends to turn to self-destructive behaviors when its culture is threatened. There were certainly economic factors at play: poor people seek the solace of oblivion in drink or drugs in order to forget the hopelessness of their condition. But the Trinidad Indians were not disenfranchised in the same way that Native Americans, for example, had been – their alienation from the political process had for a long time been a matter of their own choice, not the result of overt discrimination. And their poverty, while marked in contrast to conditions in the First World, was not significantly worse than anyone else’s in the West Indies. It was clear that the only way to sort out the apparent contradiction of Indian alco­holism was to observe it in action and to reconstruct the history of the Indians’ association with alcohol as they themselves understood it.

In a similar way, the adaptation of people with mental challenges (particularly those with mental retardation) to life outside the hospital has been shadowed by sexual misadventure. People with mental retardation have traditionally been viewed as naive innocents who, lacking ordinary self-control mechanisms, explode into sexual depravity at the least provocation. As such, the traditional response of caregivers has been to withhold sexual information – sexuality train­ing has rarely been part of the ‘habilitation’ plans along with such matters as how to make change, tell time, or read a bus schedule. But far from keeping the peo­ple in a state of innocence, such ignorance only leads to confusion, sometimes with disastrous consequences. So are people with mental retardation condemned to live as asexual beings (although physical castration or enforced sterilization is no longer a legally approved option)? Is there a way to integrate sexuality into the coping strategies of deinstitutionalized adults? Again, the answers could only come through experiencing life as the people themselves see it, not by formulat­ing judgments based on value-neutral clinical data.

3. Ethnographic research is used to identify participants in a social setting

Even when researchers set out to study a community that is considered to be well known and understood, they must realize that the dynamic of change leads to the inclusion of heretofore unacknowledged participants in the network of social interaction.

For example, the overseas Indian community was thought to revolve around the family, which, in traditional Indian culture, was a ‘joint’ organization (i.e. composed of a group of brothers, their wives and children sharing a common household with their father, the family patriarch). The joint family did not, as it happened, survive the indenture period. The family is still in fact central to Indian social organization in Trinidad, but the identification of who is and is not considered ‘family’, and the relationships among those members, are not what they once were. Descriptive ethnography in which the contemporary family organization was ‘mapped’ in detail could help clarify this situation.

The situation of deinstitutionalized adults with mental retardation was also often rendered in terms of expectations and stereotypes – often depicted as dependent clients and powerful service providers and/or caregivers. This rela­tionship is true up to a point. But for adults with mental challenges who also live in non-institutional communities, there are other elements in the social network to consider. What other people play important roles in the lives of those with mental challenges? What is the nature of their interaction? Again, detailed ethno­graphic description helped sort things out.

4. Ethnographic research is used to document a process

Unlike a clear-cut statistical relationship, a process is composed of numerous and ever-shifting elements. Much of life as it is really lived (as opposed to the way it can be controlled in clinical or laboratory research settings) is a matter of dynamic process.

For example, at the time of my initial field study, the main way in which the alcoholism of the Trinidad Indians was dealt with involved membership in Alcoholics Anonymous. AA has long been a reasonably successful method for helping alcoholics cope with their disease, but it was developed in the United States and grew out of a strongly Christian world-view. Why was it working among Hindu and Muslim Indians in the very different social world of Trinidad? An ethnographic study of AA in Trinidad was needed in order to document the process of recovery; how, in effect, did the Trinidad Indians take the standard ele­ments of AA and shape them to fit their own culture and the particularities of their own situation?

The adaptation of deinstitutionalized adults to the community is clearly more than a matter of signing official release papers and sending people on their way. By following some people as they moved from custodial care to independent liv­ing, it became clear that adaptation is a complex process and one that is managed with varying degrees of success. The ability of people to hook up with formal agency supports (e.g. medical, educational, vocational, transportation, residential services) was always mediated by their ability to find informal support systems composed in various ways of peers, neighbors, family, and friends.

5. Ethnographic research is used to design setting-appropriate measures

Ethnographers are not at all opposed to the use of quantitative measures, but they do insist that such measures grow out of the local experience. While measures thus modified are often based on recognized, reputable standardized tests (so that they are more useful for purposes of comparison), it is important that they be sensitive to local conditions. In some cases, such sensitivity is a matter of modi­fying content (e.g. some topics such as sexual behavior are freely discussed in some cultures but are taboo in others). In other cases, it may require translating the measurement instrument into language that can be understood by study participants. (Sometimes an actual other language is involved if research is being conducted in a non-English-speaking place. Or it may mean translating concepts from complex academic jargon into terms commonly used by non-scientists.) In still other cases, modification may require accommodations in the way in which the measurement instrument is administered. (For example, in some cultures, a male researcher would not be allowed to interview a female research participant, espe­cially about personal matters, without the presence of some sort of chaperone.)

Often in quantitative research, standardized instruments are administered at the beginning of a project as they provide a lot of precise and objective data that can be used to refine working hypotheses. But in ethnographic research, the adminis­tration of such instruments is best reserved to a later part of the research process so as to give the researcher some time to learn enough about the people and their community to present the measurement instrument in ways that are considered both reasonable and acceptable.

In both the Trinidad and the deinstitutionalization research, I made use of stan­dardized measures. In the former, I used the Health Opinion Survey (HOS), designed by medical researchers to measure levels of perceived psychosocial stress in a community. The HOS was originally used to test the correlation between stress and psychiatric disorder. I used it to see if there was a link between stress and alcoholism. The main modification was administrative. I had learned through my participant observation in the community that Indians considered alcoholism to be a social disease rather than an individual failing because they were most concerned with its negative impact on family and community rela­tions. As such, they preferred to talk about their personal problems in groups, rather than in individual encounters. So I administered the HOS at AA meetings or social gatherings where the respondents felt free to discuss their responses with one another before marking them on the paper. This departure from accepted pro­cedure certainly compromised the comparative value of the resulting data, but it made for unexpectedly rich results; a perspective on what people perceived to be stressful that emerged from this group discussion was far more important in this community-oriented society than the ‘pure’ responses of many individual respon­dents in a clinical setting ever could have been.

Once I had detected the concern with sexuality among deinstitutionalized peo­ple with mental challenges, I wanted to survey my population to see how much they actually knew about sex. Working with a colleague who was a psychothera­pist, I came up with a diagnostic checklist that assessed both objective sexual information (e.g. anatomical details) and subjective attitudes about sexuality and relationships. Since the caregivers were in most cases very uncomfortable with discussions of this topic, it would have been disastrous to have barged in with a ready-made measurement instrument. Taking the time to develop one that reflected what I had already learned from interacting with the people (and that also relied on the trust I had established with the participants) meant that the ulti­mate results were meaningful to the particular people in the group I was study­ing. Like the Trinidad Indians, the deinstitutionalized adults found it very helpful to discuss their responses with one another; it was very important for them to have something that had the character of an ordinary conversation, rather than yet another clinical ‘test’ that put them as individuals on the spot.

Source: Angrosino Michael (2008), Doing Ethnographic and Observational Research, SAGE Publications Ltd; 1st edition.