Analyzing Qualitative Data

Alan Bryman, Bob Burgess, Alan Bryman, Bob Burgess

• Publisher: Routledge
• Year: 2002
• ISBN:
  9781134927531
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

This major inter-disciplinary collection, edited by two of the best respected figures in the field, provides a superb general introduction to this subject. Chapters include discussions of fieldwork methodology, analyzing discourse, the advantages and pitfalls of team approaches, the uses of computers, and the applications of qualitative data analysis for social policy. Shrewd and insightful, the collection will be required reading for students of the latest thinking on research methods.

Chapter 1
Thinking through fieldwork

Judith Okely

In this chapter, I describe the way in which I interpreted and wrote up my material from an intensive fieldwork study of Gypsies and aspects of government policy. The approach and methods which informed this work were those of social anthropology. I give details of how I recorded my material and, more relevant for this collection, how I made use of it along with the totality of my field experience for the ensuing publications. Since the term ‘qualitative’ has been applied to a range of different methodologies within the social sciences, it is important to outline the distinctive characteristics of social anthropological research. ‘Qualitative’ can refer to research using only a small sample of interviews, whether structured or unstructured. In either case the qualitative material is bounded by the cultural conventions of the interview. Paradoxically, the interview format is associated first and foremost with quantitative surveys whose positivistic conventions have set the agenda (cf. Oakley 1981). ‘Qualitative’ has also come to be used to describe the research which this chapter addresses, namely that which emerges from participant observation.

There are significant contrasts between anthropological and sociological empirical research, with implications for analysis. The two disciplines came from different contexts. Social anthropology was formerly associated with the study of non-western societies, mainly by westerners. Sociology’s empirical work was concerned mainly with western societies of which the sociologist was a member. Unlike the sociologist, the anthropologist could not take much as given, he or she could not isolate one theme extracted from a wider context, since the society as a whole was largely unknown to the researcher, and undocumented. Rigidly formulated questionnaires were inappropriate. These and the interview mode are culture bound. The sociologist could be more presumptuous in knowledge of the wider social context. Whereas Durkheim (1897) could claim to identify and sub-classify suicide in France, Malinowski (1926) had first to discover and then redefine such a practice among the Trobrianders. He had no statistics to play with.

The way in which the anthropologist carries out fieldwork affects the sort of material produced, then analyzed and presented in the final texts. The anthropologist rarely commences research with an hypothesis to test. There are few pre-set, neatly honed questions, although there are multiple questions in the fieldworker’s head. There are theories, themes, ideas and ethnographic details to discover, examine or dismiss. The anthropologist, despite months of literature reviews, possibly years of theoretical and comparative reading, will have to eject hypotheses like so much ballast. The people may not live as recorded, there could be famine, strife or civil war. Rituals may be missionized, nomads dispersed, leaders imprisoned, documentation a distortion or deflection from the outsider’s gaze. The ethnographer must, like a surrealist, be disponible (cf. Breton 1937), and open to objets trouvés, after arriving in the field. This approach inevitably affects the subsequent interpretation and analysis.

Although early field anthropologists made claims for the scientific status of their work, they have been less vulnerable than empirical sociologists to demands for positivist legitimacy in methods. Formerly, it was considered sufficiently impressive that anthropologists actually uprooted themselves and went to live elsewhere for extended periods. Social scientists who stayed at home were not in a position to challenge the techniques of pioneers in the unknown fields of exotica. There has been greater freedom in the analysis of fieldwork research. A great deal is taken on trust about the way material is written up. There can be no easily replicable formulae. The notion of techniques to be applied uniformly across the globe is inappropriate. Granted, social anthropologists of the earlier school have been too cavalier in both preparing students for the field and in conveying advice about how to write up. The lacunae are best filled by detailed autobiographical accounts of fieldwork and the ways in which interpretations are arrived at. These are relatively rare and split off from what are seen as the core concerns of the discipline (Okely 1992). The increasing bureaucratic and pedagogic demand for explicit methods ‘training’ has sometimes meant that social anthropologists have half-heartedly and inappropriately fallen back on textbooks devised for sociologists and others.

The historically divisive association of sociology with western societies and anthropology with non-western societies is no longer appropriate. Each discipline has strayed into the other’s territory. While retaining its traditional methods, social anthropology can be used in the study of any group or society. I have, for example, applied it in the study of transport, the elderly and planning in East Anglia (Okely 1991), in addition to the study of Gypsies and government policy in England. Others have adopted an anthropological approach in Britain for scrutinizing the police (Young 1991) and views of death (Hockey 1990). These studies show that the kind of qualitative material which anthropology’s methods and theories generate is different from other disciplines within the social sciences.

Each discipline retains its different historical approaches to methodology. Within sociology there appears to be a widespread association of participant observation with the theoretical perspective of symbolic interactionism (Hammersley and Atkinson 1983, Silverman 1985).¹ By contrast, the research material gathered by anthropologists can be placed in as many of the theoretical perspectives as there exist in both the social sciences and, if relevant, the humanities.

It is the custom for the anthropologist to be both fieldworker and analyst-author. Division of research labour into discrete tasks, or between individuals, is at a minimum. The anthropologist fieldworker records, interprets and writes up his or her own material. For the anthropologist, the stages of knowledge as the research progresses are not sectioned between persons. So there is no need to formulate mechanical procedures and managerial-type instructions to ensure uniformity of perspective along some chain of command. The anthropologist does not have to check and double-check whether numerous assistants and interviewers have understood or even faked the collection of data. He or she has instead to look to his or her specific relationship with the people who are the subject of study. The anthropologist becomes the collector and a walking archive, with ever unfolding resources for interpretation. By contrast, a social scientist in a prestigious research centre asserted that in order to follow the correct social science procedure and to attain ‘objectivity’, ideally someone other than I, the fieldworker, should write up the final report with the aid of my field notes. The fact that I completed the task myself was seen uneasily as a form of intellectual cheating rather than a scientific necessity and standard anthropological practice. Such a division between collection and analysis might be possible in a research tradition where the researcher delegates the former to a reserve army of interviewers with pre-ordained questionnaire and clone-like application.² The pre-selected choice of answers gives material which can be mechanically classified as part of the analysis.

Agar, the anthropologist, has offered an alternative descriptive term for research and field work which is not hypothesis bound. A somewhat mechanistic metaphor, which doubtless allays the worries of those wanting proof of ‘tools’ of research, is what he names the ‘funnel approach’ (1980:13). From the outset of field work, the anthropologist adopts an open-ended approach to the full range of information and to all manner of people. This is the essence of the holistic approach. The material and ethnographic concerns are not cut to size at the start. The people who are the subject of study are themselves free to volunteer their concerns in their own voice and context. All this has implications for the kind of material and field notes which the anthropologist is faced with when it comes to writing up.

Both during the fieldwork and after, themes gradually emerge. Patterns and priorities impose themselves upon the ethnographer. Voices and ideas are neither muffled nor dismissed. To the professional positivist this seems like chaos. The voices and material lead the researcher in unpredictable, uncontrollable directions. This is indeed not a controlled experiment. The fieldworker cannot separate the act of gathering material from that of its continuing interpretation. Ideas and hunches emerge during the encounter and are explored or eventually discarded as fieldwork progresses. Writing up involves a similar experience. The ensuing analysis is creative, demanding and all consuming. It cannot be fully comprehended at the early writing-up stages by someone other than the fieldworker.

Long-term participant experience helps to make sense of even the most detached survey data. Leach, in a critique (1967) of an extensive survey of landownership in Ceylon (Sarkar and Tambiah 1957), was able to draw on his fieldwork in just one village to counter some interpretations of the statistics for 57 villages. The survey had concluded that 335 households were landless. However, from his detailed first-hand observation of inheritance practice in the region, Leach was able to point out that over time, a considerable number of the younger informants would inherit land. The same applied to many sharecroppers who were in fact heirs of the owners. He also suggested that some of the interpretation of the apparently unproblematic survey data was convincing only because the main researchers, already familiar with the region, arrived inadvertently ‘at their conclusions by intuitive methods…. The numerical apparatus in which these conclusions are embedded seems to me to be very largely a complicated piece of self deception’ (1967:76).

After fieldwork, the material found in notebooks, in transcripts and even in contemporary written sources is only a guide and trigger. The anthropologist-writer draws also on the totality of the experience, parts of which may not, cannot, be cerebrally written down at the time. It is recorded in memory, body and all the senses. Ideas and themes have worked through the whole being throughout the experience of fieldwork. They have gestated in dreams and the subconscious in both sleep and in waking hours, away from the field, at the anthropologist’s desk, in libraries and in dialogue with the people on return visits. Photographs point to details hitherto unnoticed by the fieldworker in the midst of the action. They may also revive hidden memories. The anthropologist may notice ethnographic detail which photographers do not perceive. The photograph on the cover of my 1983 book shows the interior of a Gypsy trailer-caravan. The professional who provided it failed to notice both before and after the event that the kitchen area had no sink. This is a crucial clue to Gypsies’ pollution taboos. Other sources have also to be carefully scrutinized. Snatches of music may conjure up images and forgotten or half-submerged insights. The act of interpretation and writing from past fieldwork may be as evocative and sensory as Proust’s description of the tasting of the madeleine cake in A la recherche du temps perdu (1954).

The understanding and ways of making sense of the material and of writing cannot be routinized and streamlined as instructions for methodology textbooks. Nor can it be fully assessed at this stage by a non-participant. Instead, to admit to the vastness, unpredictability and creative turbulence in which the ethnographic writer is immersed can be a reassurance that positivism is no guide. The methods in which many social scientists have been instructed have been an intellectual carapace. The puzzled novice researcher may be contaminated by positivistic notions of ‘contamination’, ‘detachment’, ‘prediction’, ‘operationalization’ or ‘typicality’. Since ethnographic openness or disponibilité have defied hypotheses, the material cannot be subjected to strict formulae. The problem is how to convince researchers from other traditions or those who are schooled in positivistic formulae.

Years after my intensive fieldwork on Gypsies, and after follow-up research, there are still reverberations, there are still things to write about or to reinterpret, especially in the light of shifts in the theoretical and substantive concerns of social anthropology. It is only now that I feel free to detail the way in which the material was written up for books published in 1975 and 1983. The reasons are both political and intellectual. The first book, Gypsies and Government Policy in England (Adams et al. 1975) was policy-oriented. It was believed by the research centre that overseered it that policymakers would respond most effectively to quantitative data (Okely 1987). On the other hand, they had chosen to employ an anthropologist because this discipline was traditionally associated with the study of the exotic and non-literate ‘other’. Thus there was a built-in contradiction which the centre had not properly confronted, if extensive quantitative data was to be the ideal. The details of the struggle for qualitative methods could not be easily published (Okely 1987). Although an anthropological methodology was eventually accepted, its description in any public sphere risked political controversy. A detailed account of participant observation among both local officials and a vulnerable ethnic minority, and the seemingly clinical way in which field notes were dissected, risked being misread and misrepresented. Simultaneously, an exploration of how cumulative field experience leads to imaginative and intuitive interpretation would have been unsettling or irrelevant to the bureaucrats and politicians whom the research was intended to influence. Their received notions of research were and are still drenched in scientistic clichés.

There are also intellectual reasons for the absence of earlier research descriptions. Few anthropologists by the late 1970s had devoted much space to a detailed description of fieldwork practice. Resisting pressure to put a description in an appendix, I included a chapter on my fieldwork (Okely 1983). This deliberately excluded references to note taking. An autobiographical account was not as free as I had wished. Subsequently, social anthropology has shown new interest in examining the construction of ethnographic texts. Key monographs have been scrutinized for the literary devices used in the presentation of non-fictional material (Clifford and Marcus 1986, Geertz 1988).

It is standard practice for an anthropologist to live alongside a group of people for at least a year. I lived in a trailer-caravan on Gypsy encampments and went out to work with them: calling for scrap metal and joining a potato-picking gang. Research entailed periods amounting to about two years, including return visits, as well as participant observation among government officials and others in regular contact with Gypsies. Nothing approximating to an interview was used in the research for which I was responsible. Accordingly, the material which informs my writing is very different from that gleaned from one-to-one interviews with individuals divorced from daily practice and context. When I conducted research among government officials and non-Gypsies or ‘Gorgios’, as Gypsies call them, in the area, free-flowing conversations and dialogue occurred as I accompanied them about their business and at their leisure. All together, the number of people, whether Gypsy or Gorgio, whom I encountered and from whom I gained information amounted to several hundred, but my approach still places me behind what is considered to be the qualitative divide. The number of ‘informants’ is not an adequate guide to the distinction between very different research approaches.

Some of the themes and subsequent chapters in my early publication on Gypsies were explicitly affected by the demands and brief of the policy-oriented centre. The aim was to examine the Gypsies’ position and preferences in the light of recent legislation which presumed long-term sedentarization and assimilation of this ethnic group. The needs of this travelling people and their conflict with the dominant society were unavoidably a key focus in the sort of questions addressed by the ethnographer at all stages of the research. Thus the political context and funding proposal influenced the way fieldwork was conducted, the themes selected in analysis and the projected readership (Adams et al. 1975). Even without the policy subtext, however, it still would have been impossible to write of a ‘self-contained community’ (Okely 1983). The non-Gypsy or Gorgio made an appearance every day on the Gypsy camps in body or spirit. Thus the Gorgio appeared in field notes and published text.

With minimal success, I had combed the anthropological monographs and other literature for guidance and for reassurance in the face of increasing scepticism among my employers about non-questionnaire research. Then a chance meeting with the Africanist anthropologist, Malcolm Mcleod, afforded me the best and only detailed methodological advice I was to find at the outset of fieldwork. From his experience, he suggested, ‘write down everything you hear, smell and see; even the colour of the carpets…. Ideally you...

Table of contents

• COVER PAGE
• TITLE PAGE
• COPYRIGHT PAGE
• ILLUSTRATIONS
• CONTRIBUTORS
• PREFACE
• DEVELOPMENTS IN QUALITATIVE DATA ANALYSIS: AN INTRODUCTION
• CHAPTER 1 THINKING THROUGH FIELDWORK
• CHAPTER 2 FROM FIELD NOTES TO DISSERTATION: ANALYZING THE STEPFAMILY
• CHAPTER 3 ANALYZING DISCOURSE
• CHAPTER 4 ‘SECOND-HAND ETHNOGRAPHY’ SOME PROBLEMS IN ANALYZING A FEMINIST PROJECT
• CHAPTER 5 LINKING QUALITATIVE AND QUANTITATIVE DATA ANALYSIS
• CHAPTER 6 ANALYZING TOGETHER: RECOLLECTIONS OF A TEAM APPROACH
• CHAPTER 7 FOUR STUDIES FROM ONE OR ONE STUDY FROM FOUR? MULTI-SITE CASE STUDY RESEARCH
• CHAPTER 8 FROM FILING CABINET TO COMPUTER
• CHAPTER 9 QUALITATIVE DATA ANALYSIS FOR APPLIED POLICY RESEARCH
• CHAPTER 10 PATTERNS OF CRISIS BEHAVIOUR: A QUALITATIVE INQUIRY
• CHAPTER 11 REFLECTIONS ON QUALITATIVE DATA ANALYSIS