Anthropology :Reading review questions

,l'i, rll;.'r-!-: i i:i: :.1:i;:!

CHAPTER

'1 1 Ethnography: ParticipantII Observation and Key- Informant Interviewing

Th" n"*, three chapters outline the primary research methods anthropologists use for prac- tice. It is not my intention to provide a substitute for some fine books (see Agar 1996; Bernard 1995; Pelto and Pelto 1978; Spradley 1979,1980) that are available on anthropo- logical methodology. What I will do is relate how these techniques are used to practical ends. As has been underscored in the chapters on policy analysis, methodological virtuos- ity may be the most important skill to bring to practice. I discuss the applied dimensions of participant observation and key-informant interviewing in this chapter, group interviewing in the next, and some quantitative approaches using questionnaires and indicator analysis in Chapter 13.

Among the social sciences, anthropology has been a leader in the development of qualitative methodologies. Patton, a sociologist, outlines their major benefits for applied work:

Qualitative data consists of detailed descriptions of situations, events, people, interactions, and observed behaviors; direct quotations from people about their experiences, attitudes, beliefs, and thoughts; and excelpts or entire passages from documents, correspondence, records, and case histories. The detailed descriptions, direct quotations, and case documen- tation of qualitative measurement are raw data from the empirical world. The data are col- lected as open-ended narrative without attempting to fit program activities or peoples' experiences into predetermined standardized categories such as the response choices that compromise typical questionnaires or tests. (Patton 1980:22)

In applied situations, anthropologists' assets will continue to be their capacities for naturalistic, qualitative inquiry in local communities and organizations using participant observation and key-informant interviewing. The researcher can provide authenticity and empathy to social science "data" or information. Effective ethnographic portrayals are also very useful for correcting misrepresentations or misperceptions of a people's actions, motives, or needs. Accompanied by sensible recommendations, they can contribute to the design of more effective programming, services, and policies.

,ols* {orrf < cfi vt?J 1'Aom: 1"1, trvln,,#711r*l irlir'i:!Tfi, I l*,. Cc .n t*F r^'r.1" (foofz< €, t+ili; r fioo"n, 2a C 0

t4t

142 PART T H R EE / Methods forAppliedResearch

Ethnography and Participant Observation

An Omnibus Strategy

Ethnographic fieldwork or participant observation is really not a single method or set of pre- cisely defined procedures. Instead, it is an omnibus strategy, an approach that contains a vari- ety of information-gathering techniques that involve various forms of observation-from unobtrusive ones to full-scale participation by a researcher deeply and actively absorbed in local activities. Researchers record small snippets of informal daily conversation that yield valuable insights. They also perform lengthy open-ended key-informant interviews or administer structured interview schedules to representative samples of informants. Anthro- pologists also use formal data-gathering techniques such as mapping households and the local terrain, recording the use of natural resources, and collecting censuses. For long-term research, they collect genealogies and learn the language in order to comprehend important local cultural constructs and categories. Some research could even involve the administration of projective instruments like Thematic Apperception Tests, in which people are asked to describe or interpret the situations portrayed in a set of photographs or drawings. Toward the

end of the fieldwork period, formal questionnaires may be administered. Anthropological fieldwork, then, is not linear. It is sporadic and zigzags in numerous

directions. Because it is flexible, it responds to the natural unfolding of events. A person might begin to follow one research lead or gain pieces of information relevant to some important subtopics, but new' often unanticipated, ones turn the researcher's attention and efforts elsewhere. Later, he returns to previous topics, augmenting or verifying previous bits of information through observation and interviews. Yet the process is not a random one. Derived from a formal research design, an inventory of relevant topics has been established and information is sought on each of them during fieldwork. But some topics may be dis- covered to be irrelevant to the situation, although immersion invariably reveals new and often more pertinent topics. Clearly, this flexibility has an advantage for the examination of issues and problems for practical solutions.

While doing community or organizational research, the ethnographer participates in networks of personal associations and friendships. He becomes involved with one cluster of people, engages in a number of activities and conversations, turns to another group, then may come back to one of the earlier networks. For instance, when doing fieldwork in Inuvik in the Northwest Territories, I would spend time with a number of circles of people-Inuit, Dene, and M6tis-at different periods as the reasons for the interactions waxed and waned.

Anthropologists contend that information retrieved in these ways is equally or even more valid than those data that, on the surface, appear more rigorous, robust, or "scien- 1ifig"-5ush as questionnaires and statistics from the more formalistic social sciences. There is also a certain amount of authenticity to participant observation, because the researcher was actually there.The variety of the interactions, waxing and waning in various completed and emerging episodes, allows for the possibility of cross-checking or validating previous assumptions or theories. Researchers may see hunches or leads negated, con- firmed, or made indeterminate. Accordingly, a good fieldworker tends to develop a highly skeptical attitude about his or her own ideas or theories and checks them out through alter- native techniques.

C H A PT E R 1 I / ParticipantObservation and Key-Infbrmant Interviewing t43

Time Spent in the Field

Typically, participant observation involves a reasonably long stay in the setting. Many rec-

ommend at least a year, so that researchers can experience the full seasonal cycle to account for any periodic anomalies that might occur in behavior. For instance, if people are season- ally dispersed for a portion of the year rather than living in nucleated settlements, they might exhibit temporary differences in social organization. That is true of many Native groups in the isolated communities of Alaska or northern Canada. If researchers restrict their time to one of the two periods, they may misunderstand the nature of overall social organization.

In applied venues, the time of stay may not always embrace the entire seasonal cycle. The anthropologist may already be tamiliar with the community, culture, or region from pre-

vious fieldwork. He or she may have a relatively defined focus instead of generating a holis- tic view of the community in the classical ethnographic style. The mandate usually requires that the investigation be done fairly quickly. Also, many applied anthropologists now work in their own countries and frequently with more mainstream populations and institutions. If they are already familiar with the more significant ways of doing things-the cultural rules-their "learning curves" may not be as steep. Finally, the applied anthropologist, as consultant, might also regularly visit the field site or sites, monitoring impact, reviewing the progress of programs, or supervising the work of co-researchers, who might be members of the local community. This would be a form of commuting, unlike the classical form of field- work but still involving participant observation.

Yet applied anthropologists may still fulfill normal expectations and spend a long time in the field even if they are already familiar with the people in question. For instance, a per- son might spend as much as two years doing a major social impact assessment, projecting and monitoring the results of a hydroelectric project that dislocates a community. Also, site visits in applied research may be spread out over several years, thus providing the equivalent of the more traditional year in the field.

Stages of Fieldwork

Ethnographic fieldwork or the participant observation process usually includes a number of stages. The first is entering the fieldwork setting. Typically, that could involve a number of things-gaining permission to be there, explaining one's presence, gaining some minimum or essential degree of trust, and attaining early familiarity with the setting. This stage can cre- ate anxious and unnerving moments for even the most seasoned fieldwork veterans. There are the possibilities of being refused admission, of unpleasant or even hostile encounters, a sense of personal vulnerability, all of which may threaten the researcher's self-esteem. It is rarely a smooth transition. Although the most serious diffrculties of the initial stage usually disappear, they may resurface during fieldwork. Fieldwork can be marvelous and exhilarat- ing, and many people get much enjoyment and satisfaction from it, yet most will readily admit that the period of entering the field can contain some anxious moments.

During this phase, a researcher has to establish clearance with the appropriate gate- keepers, officials such as district administrators, village and tribal councils, and executive directors of government and nongovernment organizations. It is usually appropriate to get

l{4 PA RT T H R E E / MethodsforAppliedResearch

the clearance at the highest levels first. Yet even line bureaucrats in the field, far away from head offrces, can continue to be suspicious or exercise local autonomy by not providing the

next level of clearances that a person needs to proceed or get infbrmation. Still, in applied work, you cannot be sure that even ofhcial endorsements or internal clearances, along with an explicit and pragmatic research design, will make it any easier to establish contacts and gain the trust of people.

Field researchers still have to explain why they are there to local people, who may decide not to comply. Sometimes, their attitude toward researchers may be more one of bemused tolerance rather than full acceptance. A number of years ago when we were setting up a field project in rural Saskatchewan, two graduate students were sent to scout the local communities we had chosen and to see if they could find us accommodation. They ran into a local farmer and became engaged in a friendly chat. When he asked why they were there,

they painstakingly described our interests in investigating local farmers'ecological adapta- tions, marketing choices, shifts in strategies among choices of grains and livestock, and how

these and other variables over time would collectively affect the shape and nature of local communities. "So then, you're fresh-air inspectors," he responded with a amused twinkle in his eye. He was expressing the local rural attitude that "surveys," as the locals labeled research, were not considered useful.

Because of experiences like this, I always prepare a one-page summary in lay lan- guage that explains the research project and includes a sketch of all subprojects. It indicates how people can get in touch with me as project director or with any other key participants among the research assistants. I state who has mandated the research and what its practical implications are. Once, while collecting life histories of Southeast Asian refugees, I had such a statement translated into Vietnamese by a native speaker, and then translated back into English by another Vietnamese to check the accuracy of the translation. The objective is to reduce misunderstanding about the project before inaccurate rumors spread. It is also a way of establishing informed consent in a more natural way than by administering release forms.

In the beginning, researchers may experience a phase of cultural shock (see Nolan, 1990, for a succinct overview of this phenomenon). Unfamiliarity with the culture, commu- nity, behavior, and values can promote anxieties, self-doubts, and negative f-eelings about the

value of the work. These may be lessened if the applied anthropologist has a long-term tamil- iarity with the situation. It may be relieved in urban situations in which the investigators commute daily to the setting of investigation and return to the comforts of their home situa- tion. It may be further reduced if the study has been commissioned (possibly by the commu- nity as a whole or a respected agency within i0 for obvious practical and socially relevant reasons. The stress for applied anthropologists might be eased in circumstances in which the project involves some forms of direct collaboration with local people as partners. Yet the pos- sibility that they are dealing with policy issues revolving around scarce resources can add to the pressure on applied anthropologists. Some people will invariably feel threatened or resentful of the fact that the work is being done in the first place. The simple tact of the researcher's presence may be at issue as well.

Normally, though, this period of culture shock will subside, and over time the researcher's presence will start to seem more acceptable, and the researcher himself will f'eel more comfortable. Usually, investigators will overcome negativity by establishing rapport

C H A P T E R I I / ParticipantObservation and Key-Informantlnterviewing 145

or trust in various ways. Researchers attempt to be on their best behavior. They show respect, courtesy, and attention to the wishes of the community members. Researchers accommodate as much as possible the needs and schedules of the local people, avoiding explicit favoritism or aligning with factions and keeping most opinions to themselves unless they are favorable to the group in question. During this period, the anthropologist will likely have made a number of reliable associations with people and established local friendships that serve for personal support and security as well as valuable sources of information. What is required is very diligent ef'fort in building human relations, effective impression manage- ment, and persuasion, combined with amiability, integrity, and professionalism. Most anthropologists, in spite of residual dilficulties and possibilities for anxiety and conflict, are successful in gaining entrde and litting in. Most find that success, in itself, is quite exhila- rating. Not to minimize the experience, but some broad parallels could be made with the sorts of challenges that children or adolescents face when they transfer schools and have to experience a trial period befbre bcing accepted and with their relief at eventually being accepted.

A stage of rapid learning and cvcn euphoria oflen follows this normalization and build- ing oftrust. The euphoria is based on positive acceptance ofthe researcher and a rapid trans- fer of information during this peritxl. 'fhe researcher rnight now make very strong and positive personal identifications with thc pcoplc bcing studied. At this stage, the researcher may record impressions and "facts" rAther inaccurately in an all-too-human and understand- able desire to please local people. Inf'ormation rnight be a bit one sided, even sugarcoated, or exaggerated. Because of this, policy rnakcrs, burcaucrilts, and others may criticize anthro- pologists for a tendency to overidentity with cornmunity members and adopt their biased perceptions of the "facts" to the detrimcnt of a more "balanced" and accurate view.

This is a controversial issue that applied ilnthropologists need to face. One justification of informant identification is that, given power irnbalances, programs and services are often designed by experts and other outsitlcrs who havc aims and assumptions different from the local people. So, perhaps in a spirit of luir play, it is appropriate to nonjudgmentally under- stand the points of view of community rncrnbcrs or "target populations," especially if there has been a legacy of abuse, misunderstancling ur one-sided. policy formulation. Administra- tors and other power holders might continue to dominate through their power bases, and their biases might be supported by other policy scicntists such as economists, who tend to share their criteria for judgment and making dccisions. 'Iaking a strong advocacy stance may be important for establishing the dialogue that could lead to beneficial change. Nevertheless, anthropologists do have a very strong responsibility to establish and maintain the credibility of their findings. So at this tempting stage of rapid learning, anthropologists should feel obliged to carefully record all their observations and emerging opinions, theories, and hypotheses. These should not be taken at face value but continuously scrutinized for accu- racy and biases.

Bernard ( I 995: 162) suggests that the researcher should take a break at some point and leave the field setting. Then the research could be refocused to more accurately reflect recently discovered realities. New approaches could be designed to follow crucial but unex- pected leads. Yet, given the nonlinear and otten random aspects of participant observation and fieldwork, we can anticipate that more new directions are likely to emerge even after the return to the f,reld. So, following Bernard's lead, there may be rationales for additional breaks

l{6 PA RT T H R E E / MethodstbrAppliedResearch

and reassessments of the directions and findings for the research project. However, with the typical shorter terms of applied projects, there may not be time fbr such breaks. On the other hand, periodic commuting to the research site would also allow some time for assessment and redesign.

The final stages of a fieldwork project would include some form of cleaning-up or con- centrated collection of crucial data, observations, and attempts to confirm or deny hypothe- ses. In applied research, that might involve the use of survey questionnaires to confirrn or disconfirm some of the findings, hunches, and hypotheses that emerged during the course of the work. Some of these could have been derived serendipitously from casual conversations or observations in the course of participant observation, or they might have developed from more sustained and linear themes of investigation in the research.

A crucial dimension of applied research is the moral and logistic obligation to seek feedback from the people whose lives could be most affected by the research findings. At the very least, some key findings and possibly preliminary recommendations should be pre- sented before leaving the field. After such a process, the anthropologist would then exit from the field setting to work on the detailed analyses and prepare the final report. Finally, in applied anthropology, it is typical for the researcher to return to the organization or commu- nity to discuss the report.

Observation and Recording

We have considered the broad stages of ethnography involving participant observation. What are some of the finer points of its operations'/ For one thing, the need for researchers to hone their observational powers cannot be overemphasized. As Spradley (1980: 33-38) points out, in the course of everyday life we normally block out many details of our experience; oth- erwise, there would be an unbearable sensory overload. We focus on those details gerrnane to the action or experience and take for granted most other factors. We are on a sort of auto- matic "cruise control." But, for fieldworkers engaged in participant observation, many over- looked details may be highly crucial. However, the right balance for focusing detailed observations has to emerge through common sense and finding the significance of local behavior or attitudes. It may be pointless to enumerate all the items and colors of partici- pants' daily clothing. On the other hand, it might be significant to note in observations of a meeting that some members of one negotiating party tended to keep their arms folded; some frdgeted; and many rolled their eyes when any members of the other party made points. These are ultimately judgment calls, but during ethnography, it is important to carefully determine what should be described and what can be ignored.

Related to improved powers of observation are good descriptive writing skills. Field- notes are an absolutely essential dimension of participant observation-researchers should not rely on memory. Notes should be written immediately or very shortly after the event before selective memory to distort accurate recall of the observed events. Times and dates should be noted. It might be valuable to record short versions of the events as well as long, more detailed ones. Wide margins can be used for significant keywords representing the top- ics that might have been part of the events being observed. Somewhere in the notes, the researcher might include hypotheses or impressions, bits of dialogue, and direct observa-

b

B !, il

C H A PTE R 1 I / ParticipantObservation andKey-Informantlnterviewing 147

tions of actions, but they should be separated from each other and clearly noted as such. A separate daily journal has often been recommended for ethnographic fieldwork to record daily sequences of events and any difficulties, such as one's personal frustrations or anxieties that might have negatively colored perceptions of what happened.

Fieldnotes and journals become personalized: one knows what works best for oneself. The main objective, which is quite crucial, is to facilitate accurate, highly descriptive record- ing of information for effective retrieval in analysis.

The usual beginning procedure for analysis, as Bemard (1988: 196) has pointed out, consists of much "pawing and shuffling" through notes, searching for relevant information. Sometimes porlions of long and highly detailed notes can be inserted directly into the final report. The same is true for verbatim quotes that add "punch" or impact to the points of view of participants or community members. The shorter version of notes helps to establish sequences of events or provides episodes fiom which generalizations will be drawn in the larger report. The keyword marginal headings allow investigators to locate and extract mate- rial relevant to particular topics easily, so that they can all be in place when needed. Of course, many computer programs allow for arranging materials automatically on the basis of key words.

As Spradley suggests (1980: 77), "grand"- and "mini"-tour descriptions of important recurrent settings relevant to the applied research are useful in field notes. For instance, pro- grams being evaluated or needs being assessed might be located in an agency like a drop-in health clinic, a social service bureau, or an immigrant agency. Details of how such a facility appears to a new client could be valuable. How is a person received? What are the facilities like in terms of space, room, furniture, and other facilities? What is the staff doing? In other words, a grand tour accounts for relevant details about people, space, and facilities in the institutional setting. We can sense if the organization provides a welcoming but professional setting for clients or if it is intimidating and inaccessible. Next, a mini-tour could involve all the relevant ingredients as they might relate to more specific services or interactions con- ducted at the agency. As examples, descriptions of typical one-to-one or family counseling sessions would be valuable if permitted at a family service bureau. Other examples might include prenatal classes or checkups at a health clinic or, in the case of the immigrant agency, settings in which English or employment readiness training is done. I think that the reader probably can anticipate how some of these descriptlons might contain information useful for evaluations as well as providing the necessary background for describing interactions and functions.

Experiences arising from participant observation should tell us something about the main actors and principal groups. When we study organizations, we should describe how their personnel interact with clients and community members. Another appropriate context for ethnographic study is a staff meeting or other kinds of meetings. Our observations should tell who is doing what, with whom, where, and when it is happening. The framework of interaction is provided, but then it is necessary to expand-filling in the rich details describ- ing what is said and what is actuaily done. That is one of the major strengths of participant observation-the events are recorded as they occur. In terms of application, if done effec- tively, participant observation is potentially unbeatable in its contribution to policy solutions. Observations directly relevant to the significant actors within their settings, with participants

1"18 PA R'f T H R E E / MethodstbrAppliedResearch

voicing their opinions of the meanings, values, and levels of satisfaction along with their sug-

gestions for change should carry a lot of weight. Conversely, statistics, questionnaires, the

use of public records, and all of the other major research methods and sources of data cannot

accurately convey what actually happens; they are only indirect measures or projections of that reality. Participant observation brings us much closer to "reality" in all of its nuances. Of course, validity can be undermined by researcher bias, and the presence of the investigator will influence the activities of those she is investigating, but these problems can be corrected by professionalism and experience in heldwork.

Unobtrusive Obseryation to Full Participation

How much do fieldworkers actually participate? To what degree are they detached observers,

and how intrusive are the observations? In some situations, anthropologists simply may not be able to enter directly into the social group or activities in question because they do not have the right qualifications to participate. Race, ethnicity, occupation, in-group affiliation, or definition of membership may preclude acceptance. For instance, an anthropologist would only get in the way of high-steel workers on the job setting. Age or gender could also present impediments in many situations. For their own reasons, the members of the group might resent or view the participation of the researchers as patronizing.

Unobtrusive observations are sometimes built in on the edges of normally inaccessible social scenes such as the galleries of operating rooms and waiting rooms of social or health agencies. Children are a special challenge. Sometimes unobtrusive observations of children in play groups can be done through one-way mirrors or on the edges of playgrounds as long

as the necessary ethical precautions have been taken. Sharing neutral or semi-neutral "offstage" contexts in which the anthropologist can be

accepted could be another way of increasing participation and gathering information. The anthropologist might join steelworkers, surgeons, or off-duty policemen at their favorite bars or restaurants and participate in the general convivialities while hearing about recent exploits and listening in on the "shop talk." Much can be learned in these offstage contexts, so long as "tall tales" are kept in perspective and checked out through other approaches.

Another way to participate is by taking on an auxiliary participant role that is more or less improvised for the situation. Imagine an anthropologist studying the isolation and occu- pational hazards facing long-distance truckers. The anthropologist may not be allowed to actually drive an eighteen-wheeler, but he might accompany the trucker as a companion and share the experiences of the road. The trucker might find the companionship worthwhile and share his perceptions about the profession and the meanings of particular incidents. During the trip, the anthropologist might be able to directly observe some of the sources of risk and stress, but the level of participation would be limited to perhaps helping to change a tire or to load and unload.

There are opportunities for full-fledged or relatively complete participation. For instance, Patton ( 1890: 140-l4l) participated as a student in a Wilderness Training Program class. That participation was one phase of a progrlm evaluation. Michael Agar (1973) in his research at an addiction research center, began as an employee-as a janitor-but later switched to being a full-tledged researcher after the patients became familiar with him.

C H A PT E R I I / ParticipantObservation and Key-Informant Interviewing 149

In normal fieldwork situations, the opportunities to participate in the activities under review tend to wax and wane. In some cases, people might appreciate the anthropologist's assistance in such activities as haying, harvesting, house or bam building, hauling in nets, and looking after children, especially in local situations when labor is intensive and commu- nity members are norrnally expected to volunteer. In studies of agencies and their programs,

researchers might relieve other staff by answering the phone or helping in office clean-ups. Although participating in these activities does not always yield research results, the cama- raderie and appreciation for pitching in often lead to more open and trusting responses from the participants.

Another avenue for participant observation is at normal occasions of informal social- izing. People value their leisure time, and many societies or groupings studied by anthropol- ogists put a special emphasis on the importance of hospitality and visiting--often to a much greater extent than middle-class North Americans do. The informality of interactions can provide the potential for easier participation, and as a guest or fellow socializer, the anthro- pologist gains information and insight. Even in urban middle-class society, much business is really conducted through informal interaction.

Unlike some of the other techniques described in this book, it is impossible to teach participant observation by outlining precise steps. It has to be experienced directly and requires full immersion. Of course there are intermediate ways of giving a taste of the expe- rience-through ethnographic field schools and applied internships. Neophytes immerse themselves in smaller doses and do not have full responsibilities for any failures of the research. Students about to do fieldwork can gain a certain amount of insight by reading the literature on fieldwork (some of which is listed at the end of the chapter). Particularly useful, from a morale viewpoint, are stories describing difficulties and stresses, ethical dilemmas, and general hardships experienced by veteran researchers. When novices later experience difficulties, they will realize that they are not unique and that the challenges are rarely the result of some personal or moral deficiency on the part of the researcher.

Key-Informant Interviews

A mainstay of anthropological work is the key-inforrnant interview. It can be used within the general framework of participant observation or alone. For example, situations of more rapid research, as with needs assessment or program evaluation done over several weeks, might make heavy use of some well-placed key-informant interviews,

Informants

James Spradley (1979:45-54) lists criteria for a good informant. First, the informant should have been thoroughly "enculturated' into the area of activity or knowledge being studied. Second, he or she should be currently involved with it. Third, the topics and cultural scenes being investigated should be basically unfamiliar to the investigator. Fourth, the informant should be able to devote enough time to the interviews to give satisfactory information. Fifth, the informant should be nonanalytical. Presumably, the job of analysis is meant to be done

150 PA RT T H R E E / Methods forApplied Research

by the anthropologist, and any analytical comments provided by the informant would bias the analysis of the investigator.

These criteria are sensible in many respects for certain types of academic research, but I think they are a bit rigid for an applied anthropologist. First of all, consider direct and thorough enculturation in the domains under question. This implies that the informant should have experience of what he or she is talking about, but some informants may be keen observers rather than active participants. When I was doing fieldwork on issues of sociocul- tural change in the north, I interviewed people who had participated in the earlier trapping period and had lived a semi-nomadic life attached to the land. They were genuine partici- pants and had been thoroughly enculturated into the life that they were talking about. On the other hand, I also interviewed several Catholic priests who had lived in the region for about three decades. They would accompany Dene people to trap lines, meet with them at their ren- dezvous, and deliver mass to them in the bush. As seasoned observers, they provided excel- lent commentaries on the seasonal cycles that augmented and corroborated details given by participants who had been directly enculturated into the domain.

Similarly, we might question the requirement that the informant must be currently engaged in the domain under investigation. As retired town-dwellers, the middle-aged and elderly Inuit and Dene people whom I interviewed were no longer occupied in the trapping and subsistence life-styles. Although their memories may have faded a bit, surely the value of such peoples' information should be incontestable.

Similarly, I do not think that we should avoid informants who give us analytical insights because, as key informants, their expertise might include the most astute commen- taries about the topics in question. In other words, I do not think that it is the role of an infor- mant to merely provide descriptive information on the topic. In my previously mentioned research, one of the Catholic priests provided an analysis of a topic that he was actually aware of as a result of reading anthropological literature. It had to do with the concept of rec- iprocity, normally associated with hunting and gathering peoples. In some previous ethno- graphic works, there had been descriptions of how moose hunters would butcher meat and give prize cuts to the needy, passing the hide on to a widow. They supposedly saved the least valuable cuts for themselves and their families. This was meant to display the value of shar- ing, a subversion of one's own desires and self-interest in order to fulfill the group's needs. There was an unstated expectation that one might be in a similar situation of misfortune in the future. The priest pointed out that although this was a culturally expected practice, most hunters would strategically provide the best cuts of meat to those considered the best hunters while still exercising generosity, at least on the surface level. This was done to ensure that such proficient hunters would, in turn, remember the givers when they had success. The priest's analytical observation was made in a detached manner and in no way judged the behavior.

Similarly, during that same research period, town-dwelling Inuit and Dene offered me analyses of some of the difficulties in their less-than-a-decade-long experience with semi- urbanized town life. They would make direct comparisons of their current lives with life- styles in the past and with life-styles of the dominant or temporary white middle-class residents. Among the more interesting analyses provided by one Dene informant was a com- parison of child-rearing strategies. Both Inuit and Dene styles of child rearing were nondi- rective, and children had autonomy in their actions-staying up late under the midnight sun,

C H A P T E R I I / ParticipantObservation and Key-lnformant Interviewing 151

wandering off, starting small fues, and so forth. As one person put it: the environment was the main disciplinarian for children, and they learned from such experiences. However, there were more dangers in the town, and such experiences did not fit with the realities of life there. More scheduling and adult supervision was the style of the white "southerners," and these expectations were imposed during school and work. I think that such observations were very useful and helped to draw out some significant conffast. It would have been very foolish to have ignored them. In fact, it is probably the case that most anthropologists owe much of their middle-range analyses to their informants. If it were not for the need to main- tain informant anonymity for their own protection, it might be far more appropriate to list them as coauthors of many of our monographs.

Although Spradley's criteria for selecting informants may be a bit stringent, they prob- ably can be explained by his main emphasis-that of uncovering the ingredients of unfamil- iar cultural domains and the cognitive and linguistic logic underlying them. It does indeed make sense to locate those who are thoroughly enculturated and actively engaged in the top- ics under question. The criteria might be appropriate for segments of many applied projects, such as coming to understand local knowledge of resource use and conservation practices. Yet, for many applied projects, more flexibility in selecting informants makes sense.

How do we find key informants? Some of them might obviously be prominent local figures or gatekeepers in the community who are familiar with organizations and topics under investigation. These could include officials such as chiefs, tribal councilors, directors of government bureaus or nonprofit agencies, religious or healing leaders. They could also be outsiders who have had long-term, sympathetic liaisons with the peoples in question. Other informants are experts in the areas of health, education, conflict resolution, agricul- ture, and so forth. Interviewing front-line workers who deliver services under investigation makes good sense. Such people can be interviewed for their particular areas of expertise or knowledge and can provide opinions and information on broader areas of concern. Their responses should always be considered with some understanding of the biases and agendas coming from any vested interests. The nature of the research project will dictate how much we should rely on their opinions. If the study is very broad and covers all aspects of the com- munity, it makes sense to have them participate in interviews. Given their knowledge and possible insight, they may be able to direct us to other leads and subdomains of study not pre- viously considered. Also, if the study deals with the interactions and difficulties of service providers and clients in programming, part of the equation would be to interview those pro- viding the services as well as those receiving them.

For most applied anthropological studies, there is no substitute for directly collecting information and opinions at the grass-roots level. In studies fbcusing on the needs of street children or to reshape local health facilities so that they are more rooted in the community, it makes much more sense to go quickly and directly to those whose opinions matter the most. Relying too heavily on prominent citizens or gatekeepers as key informants will distort the data that a researcher eventually needs.

Although categories of informants may be relatively easy to identify, how are specific ones identified? They may emerge more or less automatically as the result of knowledge gained during participant observation. They frequently emerge as a result of a "snowballing" effect, whereby informants identify other likely informants. They can sometimes be selected on the basis of random sampling although that tends to be unusual.

152 PA RT THREE / MethodsforAppliedResearch

The Interview

In the actual process of interviewing, one of the key things to remember is that, although it

might resemble a natural dialogue, the interview should be largely a one-way conversation,

in which the informant does most of the talking. As Bernard (1995: 212) notes: "The rule is:

Get an informant onto a topic of interest and get out of the way. Let the informant provide

information that he or she thinks is important." Similarly, it is essential to record the responses in the local idiom as much as possible. This will provide clues to significant local

categories and meanings that are valuable for themselves but also useful for later stages such

as phrasings in the design of questionnaires. Taperecorders are highly desirable but optional,

depending on the feelings of either the informant or even the researcher. Yet note-taking is

mandatory. I feel that it is a good idea to show the list of questions in advance to put the per-

son at ease and also to mentally prepare him or her for the questions. I also stress to infor-

mants that it is not helpful to merely answer questions in ways that might be presumed as pleasing to me. I tell them that truth and accuracy are more important. To overcome possible

diffrdence or reluctance to be interviewed, I spend some time reminding the interviewees that they are authentic experts in the topics that I am investigating. Otherwise, I would not be

interviewing them.

Questions

It is a good idea to initiate the interview with questions that are immediate, focus on present

circumstances, and not too controversial . Present-tense questions ate meant to warm the

interviewee up to the topic, because recall is better about things that happened recently. For

example, " Please tell me about what has happened so far in your day, in coping with life as

a visually impaired person?" The interviewer might want to ask what Spradley (1979:

86-88) has called " grand-tour" euestions-those that cover the scope of the domain in ques-

tion. For example, "Tell me about a typical day in the life of a blind person." In most cases,

the investigator should have a set of questions fiom which to draw upon in the course of the

interview-ask alternative questions if one line of questioning is not working out well. For a

successful interview, the interviewee should have confidence in the abilities and purposes of

the interviewer. A number of years ago, I took a patient-interviewing course fiom a psychiatrist, when

we were sent around a hospital to interview patients on some very sensitive topics. His point

of view was that people will readily answer almost any question if they sense that you have a good reason for asking it, especially if it is clearly fbrmulated, understandable, and you ask

it with confidence. I am not entirely sure this always holds true in nonclinical or field situa-

tions. We must not forget that in this case the context was a teaching hospital in which patients were more inclined to cooperate because they believed that the interviewing might

relate to their evenrual treatment. In more natural settings, the informant might not be so willing to comply with the researcher. Yet, if the project is applied and the researcher has gained all the legitimate clearances and established trust, the informant could see value in

answering most questions. So, a professional questioning style is an important ingredient'

However, in many cultural situations, direct questioning and sometimes even the key-

informant interview are not appropriate. For example, among Native or Indian people, in

C H A P T E R I 1 / ParticipantObservation and Key-Informantlnterviewing 1s3

many situations direct questioning is considered impolite. A researcher is better off seeking what is needed through fragments of conversations or through indirect questioning. Or, if the investigator is lucky and people have decided that he is trustworthy, they will tell him what they think is important to know. It is essential that the interviewer really listen at that point.

Some suggest that there should not be a list of preestablished questions in key- informant interviews, because the object is to find out how the informants perceive every- thing that is relevant to them, especially when we start out in a barely understood domain. I believe that an interviewer should, in most cases, come well prepared with a set of a few good questions. If only one of them seryes to keep the interview going, so much the better. If the informant has something completely new to talk about, obviously the interviewer should improvise on that opportunity. But to come unprepared to an interview is unprofessional and sacrifices credibility. Clearly though, the effective interviewer is f'lexible, alert to nuances, and ready to shift directions.

Returning to the actual questioning, we should remember that, although resembling natural conversations, interviews should essentially be one way. The stage is given to the informant. The interviewer does have to interact in ways beyond the mere asking of ques- tions. What should happen when there is a long pause in the narrative? There is a big temp- tation, especially for those with middle-class backgrounds, to fill the silence with comments of their own. But informants may be collecting their own thoughts. It is important not to interfere with that process. After awhile, though, the appropriate response might be to show affirmation for what the person has been saying by reiterating some points already men- tioned. Then the appropriate thing is to ask if the informant would like to say anything else about the topic. If not, then it would be appropriate to move on to another question. On the other hand, this may be the appropriate time for a probe-a further elaboration on a new topic or subtopic raised by the informant during the interview. An example from a sfudy eval- uating an English-as-a Second-Language Class for immigrants is: "You said that you learned from your instructor and the class a lot more than just about speaking English and North American customs. Can you tell me more about what you meant by that?" The capacity to probe further is a very essential part of the interviewer's skills. Much more will be discussed in the course of an interview than what was anticipated. It is also important to give some feedback during the interview. It would be very disconcerting to face a blank or even smiling face that does not respond by giving ffirmation. The affrrmation, incidentally, does not mean direct agreement. It means indicating that you respect the person's opinions and knowledge and that you understand him although it is not generally appropriate to give your own opinions on the subjects.

Patton (1980: 207-21l) suggests a number of types of question areas that could be covered. First, there are experience questio,ns that ask about what a person does or has done. Second, there are questions that try to elicit cognitive and interpretive processes----or opinion/value questions. T'hird, feeling questior?,r are designed to comprehend the emotional responses ofpeople to their experiences and thoughts. Patton suggests that feelings and opin- ions are different. A person might express the feeling, "that is all that can be done" (in refer- ence to some action meant to ameliorate a situation), but in reality that would be an opinion. Instead, the actual feelings about the incapacity to do much about something might range among emotions-sadness, anger, despair, and so forth. Fourth, knowledge questions are asked to find out what the informant knows about the facts involved in the situation-such

l.q.f l'A R 1 'l' ll R l: Ii / Merhods for Applied Research

as who is eligible for particular programs, the characteristics of their fellow clients, and so f<tnh' Fifih. sensom questions try to tap into what is potentially touched, tasted, smelled but abovc all seen or heard. In eliciting ex-convicts' impressions of their jail terms, an inter- viewer could ask, "If I entered your cell, what would I see and what would I hear." Sixth, hackgruund or demographic questions are always useful, in fact they may be essential to the whole of the study. What are the ages, genders, education, residence, and other pertinent cheracteristics ofpeople participating in the study. Instead ofasking such questions directly during an interview, I usually give the participant a profile form, which I ask him or her to fill out. Alternatively, we might fill it out together, if the interviewee has difhculty with reading or writing.

Patton points out that questions can be asked within time frames-present, past, and future. Queries could be made about peoples'actions in the past and present and about what they might be doing in the future (work, residential moves, and so on). Regarding the sequencing of queries, Patton encourages beginning with questions that are not too contro- versial as well as those which can locate the informant in the present. Both types encourage the person to speak descriptively. After a good basis of experience and activities has been established, the interview could move on to interpretations, opinions, and feelings. Doing this is easier, because, in effect, the informant will have mentally and verbally relived the experience, establishing a context, and the chances ofaccuracy increase. Similarly, questions about skills and knowledge are best asked after contexts have been established. Rapport or trust has to be confirmed before asking opinion and knowledge questions that can be seen as threatening. Inquiries about the future would best be asked after the contexts for the present and the past have been established, because questions involving future speculations tend to generate less reliable answers.

The actual phrasing of questions is highly important. Some general tips are relatively standard. They include:

r Avoid dichotomous questions that can be answered by a simple yes or no. Example: 'Are you satisfied with the trappers'income supplement program?"you would want the respondent to provide some depth of response, giving details about experiences, opinions, feelings, and so forth.

r Make sure that questions are truly open ended, because you should be interested in the elicitation of the categories, opinions, feelings, etc., from the respondent rather than measuring them along some preconceived continuum.

Example: Instead of asking "How satisfied are you with the trappers' income sup- plement program?" it would be better to ask: "what do you think of the trappers' income supplement program?" or "Tell me about your experiences with the trap- pers' income program" (assuming that it has been established that the informant has had some experience with it).

I Avoid questions that combine too many ideas, leading to confusions as to which parts to answer.

Example: "Tell me what you think of the government's programs for health, edu- cation, and employment training." It would be more appropriate to ask questions that examine each area separately and even better to focus on aspects of each.

c H A PT E R 1 I / Participantobservation and Key-Informantlnterviewing 155

Summary

Qualitative research, as exemplified by ethnographic fieldwork, constitutes the best set of skills that anthropologists can bring to applied research. Participant observation is important because it is authoritative. It means that the researcher has "been there" and has had the opportunity to divest him or herself of possible preconceived biases that often naively gov- ern many other types of research. Categories, behavior, and important ingredients can be studied within their own contexts of meaning. The researcher can observe what actually hap- pens. Especially if done over a reasonable length of time, the method entails checks and bal- ances ensuring validity and the accuracy of information. Most frequently during participant observation, the researcher will have made contact with informants who can provide detailed insider's information and opinions about the topics. The use of key-informant interviewing is essential because the researcher cannot observe everything and does not know what the issues or circumstances under question feel like to the persons involved.

RECOMMENDED READINGS Agar, Michael 1996 The Prcfessional Stranger, Second Edition. New york: Academic press. 1986 Speaking of Ethnograpft_y. Beverly Hills, CA: Sage. 1982 Toward an Ethnographic Language. American Anthroporogist, vol. 84: l7g-79s. Bemard, H. Russell 1995 Research Methods in Anthropology: Qualitative and Quantitative Approaches, Second Edition.

Walnut Creek. CA: Altamira.

Edgerton. R.. and L. L. Langness 1974 Methods and Styles in the Study of Culture. San Francisco: Chandlar and Sharp. Fetterman, David M. 1989 Ethnography, Step by Srep. Newbury park, CA: Sage. Nolan, Riall W. 1990 Culture Shock and Cross-Cultural Adaptation. Or, I Was O.K. until I Got Here. practicing Anthro-

po lo gy, YoL l2(4): 2, 20.

Patton, Michael Quinn 1980 Qualitative Evaluation Methods. Beverly Hills, CA: Sage. Pelto, Pertti, and Gretchen Pelto 1978 Anthropological Research: The Structure of Enquiry. New York: Cambridge University press.

Spradley, James P. 1979 Tlrc Ethnographic Interview. New york: Holt, Rinehart and winston. 1980 Participant Observation. New York: Holt, Rinehart and Winston. Wolcott, Harry F. 1995 The Art of Fieldwork. Walnut Creek, CA: Altamira.