The ethnographic process

If participant observation is the ethnographer’s principal method, the ethnographic process that surrounds it comprises a number of stages. In the first stage, once the research question has been formulated the relationship between the proposed research setting and its broader context must be established. Some anthropologists have regarded the unique cultural meanings and practices in the research setting as the inhabitants’ whole world, but the majority of contemporary ethnographers would accept that most of today’s small groups are not isolated from broader eco- nomic, political and social contexts (see Ellis, 2016; Horsley, 2015). For instance, if the research question is focused on some aspect of the gender gap in exam success in science subjects, the researcher must develop an initial understanding of the broad social context, in this instance intersectional social relations, and the broad institu- tional context, in this instance the secondary education system. Then the researcher

can make an informed choice of the research setting(s), in this instance a specific school or schools. The researcher would do well to become familiar with the history of the institutional context and the research setting(s) with preliminary reading, document collection and analysis. In a sense, all ethnographies are specific ‘case stud- ies’ of something broader (O’Reilly, 2009). Case studies can be intrinsic (interesting in themselves, e. g. a single institution in government), instrumental (where the case is secondary and used as an example of a broader structural situation) or comparative (where researchers look at differences between groups, which would usually require a multi-site project).

In the next stage, the researcher must make initial decisions about specific meth- ods, all the time weighing up the advantages and disadvantages. For instance, should the ethnography be overt or covert? Covert research has obvious advantages, such as the increased authenticity of the practices that can be observed. However, a major disadvantage is that it must be justified to satisfy ethics committees or institutional review boards because it is a form of deception. In a covert ethnography, it is also impossible to validate the presence of the researcher, who might not have actually been in the research setting and could literally be fabricating data (Herrera, 2003). In his well-known covert study of occupational police culture, Holdaway’s (1984) observation of techniques used to extract confessions of guilt placed him in a diffi- cult ethical situation. This study might not have satisfied today’s ethics committees because of problems associated with deception and the duty to report malpractice. If overt research is chosen, all participants know they are being researched and they are usually provided with written information about the research and consent forms to sign. Participants may act naturally, but it often takes a long time to build trust, there is no guarantee they will do it consistently and the assumption that they are actually being authentic can never be established beyond doubt. Duneier’s (1994) celebrated study of the ‘respectable’ black male culture in Chicago is an example of an overt study in which the researcher’s presence might well have affected behav- iour and disclosure amongst some of the participants. Fielding’s (1981) study of the National Front (NF) presents us with an interesting overt–covert hybrid. The NF members knew they were being researched, but Fielding kept his anti-fascist feelings to himself to allow honest disclosure from his informants. In her study of drug deal- ers, Adler (1985) shifted from covert to overt as trust was gained. However, a contemporary ethics committee would quite possibly point out that these tactics are still forms of deception that, in research fields such as these, might also put the researcher in danger.

The next stage is to negotiate access to the research setting during periods when

its participants are going about their everyday social or occupational lives. The recruitment of appropriate research associates, gatekeepers and key informants is essential. This is sometimes problematic. Unsuitable choices – particularly in such cases where research associates and key informants fail to strike up productive relationships – can hinder the research process, in rare cases to the extent that the research has to be abandoned. Informed consent must be sought from all

participants, informants and interviewees, usually in written form that explains the purpose and methods of the research in brief but clear detail. This might seem rather formal, but many textbooks warn ethnographers to avoid ‘going native’ because, should they do so, they can end up exploiting participants and losing all sense of distance and objectivity.

However, this is hotly disputed, as we will see in the next section. Others question whether over-rapport can be possible, or whether it is simply a complaint lodged by envious ethnographers who failed to develop a rapport (see O’Reilly, 2009). Objectivity can also be threatened by the personal ideology of the researcher, which, if it is allowed to influence data generation and analysis without some degree of reflexive self-questioning, can make the question of formality or informality rather redundant.

Once access is successfully negotiated, the researcher can enter the field to conduct the fieldwork. Fieldwork entails gathering data that are sufficiently rich, nuanced and comprehensive to produce a detailed ‘thick description’ of everyday meanings and practices. During this process, it is important that the researcher produce good, clear field notes. These are usually written records or recordings ready for transcrip- tion. Throughout the process, the researcher must be as rigorous and careful as possible, and in particular must be prepared to focus on the fine details of important events that occur as powerful examples of meaning and practice. It is also essential that the researcher carve out time to reflect on fieldwork through the process, because feelings, insights and ideas that crop up in the moment can be quite easily forgotten if too much time is allowed to elapse. Field notes can be contextualized and enhanced by collecting relevant newspaper cuttings, news items from TV, radio or the internet, reports, memos, etc. that are contemporary with the researcher’s time in the field.

In the field, significant events can come and go very quickly; therefore some ethno-

graphers find it helpful to write field notes in stages of progressive complexity that allow for expanding reflection, starting with head notes for quickness and moving through scratch notes to full notes for further reflection before the event becomes hazy in the memory and initial impressions are lost. Transcribing recorded data in the moment is also important. However, some find transcription to be very time-consum- ing and, out of practical necessity, wait until the fieldwork is finished, but doing it this way the recorded data can go ‘cold’ and difficult to integrate with insights that crop up during the fieldwork. During his covert fieldwork with bouncers, Winlow (2001) found a useful compromise by writing up his field notes the following day when the context and insights were still fresh in his mind. Reflections on initial analysis and

theory can unfold from observation and interview data during the fieldwork, but the researcher’s political inclinations, cultural prejudices, theoretical preferences and personal interpretations are present in the ethnographic process from the very begin- ning, so we could settle on the principle that in the field it is on balance better to focus a little more on data collection. Except in cases where a specific hypothesis is to be tested, ethnographic knowledge production tends to be inductive, and this applies to analysis and theory too. Even the research design, although clear in its purpose and broad methodology at the beginning, tends to be developed inductively because the methods will almost invariably need to be modified as problems crop up during the fieldwork.

The next stage is analysis of the data. As we have seen, even though fieldwork should be inclined towards data collection, it helps if the researcher can conduct some initial analysis during fieldwork rather than leave it all to the end. This helps the research to be a reflexive rather than linear process. Ongoing analysis can help the researcher to come up with modified and/or additional questions, and to develop clearer initial understandings of significant events and cultural practices, which will lead to the collection of richer data and the construction of more sophisticated analysis and theory. PhD students, who tend to gather data and analyse them them- selves, are at an advantage in this reflexive process. At the heart of analysis is the technique of coding, which requires the researcher to discern patterns and changes in significant events and everyday meanings and practices in the context of the research setting. Therefore, to achieve this, coding involves the close examination of minute details in the data. Technology can help in the process of coding, but some argue that it encourages a lazy, mechanistic approach in which details vital to the eventual analysis and theory can be missed; therefore it is perhaps best used as an aid rather than a replacement (O’Reilly, 2009).

Coding is often done in two stages: first, open coding, which involves looking at

the whole data set for broad emergent themes; and, second, focused coding, where specific emergent themes are thought to be especially significant and require inter- pretation. As vital as this process is, and as rewarding as the results can be, it is also where the researcher’s values and preconceived assumptions and prejudices can assert themselves, and therefore where confirmation bias can become established at the empirical heart of the research. Researchers should be aware that when their research is published it receives institutional validation, and thus it needs to be con- ducted reflexively and with great honesty. Ideally, all qualitative researchers should engage in an internal dialectic throughout the research process, weighing up different points of view and evaluating possibilities of interpretation and analysis as the even- tual grounded theory is constructed. However, the ideal is rarely achieved, which suggests that researchers should at least compromise by being open about the pre- ferred position to which their assumptions belong. When the data, analysis and initial theorization seem to be in good shape, the researcher can begin writing up, which allows for the further development of analysis and theory (see Hall, Chapter 7, this volume).