From shadow to person: Exploring roles in participant observations in an eldercare context

Introduction

I am sitting in the staff room at ‘Stonewood Manor’ ¹ accompanied by care workers and a nurse who is working as a substitute. As we have not met before, the nurse asks who I am and what I do there. I describe my presence, and Julia, one of the staff members, fills in: ‘Yes, she wanted to study staff and they sent her here, to the very worst ones!’ Laughs. ‘We call her the Shadow.’ (Care worker at Stonewood Manor)

As I entered what would be my research field, studying staff competence at two eldercare units in Sweden, a new and unexpected area of interest emerged: the method of participant observations. This issue at first seemed challenging and then demanding. The field notes as material for this study were collected through participant observations at two eldercare units in a mid-sized city in Sweden. The observations involved staff working with the elderly residents, specifically in talking about knowledge and competence in daily work.

Participant observation as a qualitative research method is widely used and is a central method within several disciplines in the social sciences (Fangen, 2005). This method has been described as ‘the most scientific method’ by Gans (1999) because ‘it is the only one that gets close to people’ and ‘it allows researchers to observe what people do, while all the other empirical methods are limited to reporting what people say about what they do’ (p. 540). Fieldwork as the data collection method of ethnography has a long history (van Maanen, 1988). Although ethnographic fieldwork or participant observation is frequently used as a data collection method and much has been written about the outcome of these studies, little is known about the nature of the researcher and the roles a researcher can inhabit in these situations. When writing about roles, the focus is mainly on collection and interpretation of data rather than on the roles of the researcher (Wellin and Fine, 2010). Gold’s (1958) classification of roles in participant observations into four categories is a widely used reference. Although this classification may provide the aspiring researcher with an important starting point when engaging in participant observations, it is not enough when it comes to truly trying to understand the situation for the researcher in the field. Writings about the roles in participant observations are not problematized in a way that actually prepares a researcher and does the method itself justice.

This article explores participant observations as a research method and as a practice, and the situation of my role as a researcher as it developed with the interaction between the participants in the study and me as the observer. The more specific initial question posed is, (How) do roles for the researcher in participant observations change during the course of fieldwork?

Context and theoretical framework

Preparations for my observations consisted in part of reading the work of researchers experienced in ethnographic studies or participant observations: Fangen (2005), Fine (2003), Gans (1999), Gold (1958), Bryman (2001), and Hammersley and Atkinson (2007), among others. These readings did not leave me equipped with all the answers or knowing what to expect when entering the field; rather the opposite was the case. Tinney (2008) describes similar feelings about the difficulty preparing for her observations in eldercare homes in Australia. As Karp and Kendall (2001) point out, the nature of fieldwork is not presented with consensus: Every experience is unique, as each fieldworker brings a history and background into the field.

Fully aware that my presence to some extent would interfere with the staff and also with the residents to some degree, I was anxious about the observations. I primarily relied on Gold’s (1958) classic work on fieldwork roles in observations (though Gold gives credit to Buford Junker with whom Gold worked in a research team (Gold, 1958: 217)). The classification entails four roles in fieldwork. The first role is the complete participant: The observer is in the social setting as a member, the other members do not know the identity of the observer, and observations are made covertly. The second role is the participant as observer: The observer is an active member of the social setting, and the other members are aware of the identity of the observer and his/her presence as a researcher. For this second role, Gold acknowledges the risk of identifying with the participants and ‘going native’—a classic term in ethnography that results in the observer’s identifying with participants and losing the distance needed to observe and analyze. The observer as participant is the third role: The observer is mainly an interviewer and does not engage much in the activities. The fourth and final role is that of the complete observer, who does not interact with the participants. This last role might not be considered preferable in participant observations or ethnography (see Bryman, 2001, for further comments). Because of the widespread use of these roles that Gold described, I chose them as my starting or reference point. In spite of their common use, however, Gold’s roles lack variation and flexibility, as Emerson (1981), for example, has noted. They also neglect situational factors (Adler et al., 1986).

There are other ways of explaining the roles in participant observations; Kostera (2007), for instance, differentiated between participant or non-participant observations. Spradley (1980) lists the degree of involvement in Participant Observations, moving from Complete to Active to Moderate to Passive and finally to Non-Participant. Another relevant term in this context is ‘shadowing,’ as discussed by Czarniawska (2007), Erlingsdóttir (1999), and Kostera (2007), which implies a withdrawn and discreet role.

Adler and Adler (1987) differentiate fieldwork roles from more observational fieldwork. The authors distinguish among the Chicago School, ethnomethodology, and existential sociology when drawing a map of the history of fieldwork. A criticism of the Chicago School is that this view leads to an absolutist approach to fieldwork, emphasizing researcher impartiality too much. This emphasis results from the influences of positivist thinking on objectivity and the need for detachment in science. Ethnomethodology is a broad concept, partly built on the phenomenological philosophy of Husserl and Schutz. It aims at an understanding of the studied field by engaging in the lived order, the life under consideration (Pollner and Emerson, 2010). Pollner and Emerson make a distinction between early and recent ethnomethodology. Where the emphasis on the lived order used to be the ordering or the practices that organize it, recent ethnomethodology focuses more on living the order (p. 124). This distinction is made while comparing ethnomethodology to ethnography, where ethnography is described as keeping more of a distance towards the studied field. Ethnomethodology thus aims at a closer relationship, similar to how Adler and Adler (1987) describe existential sociology. Existential sociology is based on the philosophy of existentialists like Heidegger and Sartre as well as the phenomenology of Husserl and the hermeneutics of Dilthey. This framework involves the researcher’s becoming an insider in fieldwork, thus creating the risk of going native. As an alternative to one of these three views, Adler and Adler suggest and prefer a combination of ethnomethodology and existential sociology, which recognizes the essential need for the fieldworker to interact with participants.

As I completed my observations and continued my analyses, I linked my work to that of the sociologist Collins (2005, 2008), in particular his theory about interaction ritual chains. In discussing interaction ritual chains, Collins presents the following definition: ‘a mechanism of mutually focused emotion and attention producing a momentarily shared reality, which thereby generates solidarity and symbols of group membership’ (Collins, 2005: 7). The focus in Collins’ (2008) perception of the world is the notion of the interaction and group dynamic in society as the departure point. Interactions are the focal point rather than the individuals involved in them. Groups in society are held together not by rational agreements but rather by deep emotional processes creating emotions in the group: Emotions then serve as the glue. Emotions and energy develop through activities consisting of interaction rituals, in a setting of two or more people where everyone shares a focus as well as emotional experiences. The result of the rituals can be symbols, such as words and language used within the group. Symbols also signify the clear boundaries between group members and others not belonging to the group. The interaction ritual chains seem especially appropriate in this context, given the added element of ‘emotions.’

Rituals create symbols in first-order, face-to-face interaction, which constitutes the starting point in an array of further second- and third-order circuits in which symbols can be re-circulated. Once infused with situational emotion, symbols can be circulated through networks of conversation, and internalized as thinking within the individual circuits of the mind. (Collins, 2005: xii)

Materials and methods

My first contact with the first eldercare unit was in 2008; I spent five months going to the unit for observations. The observations at the second unit began in winter 2009 and ended in fall 2010, also lasting five months. I spent two to three days per week at each unit with days and times alternated for better insight into the different tasks scheduled. This scheduling allowed observation of all shifts and all days of the week. On average, I spent about 80 hours per month in the field for a total of around 800 hours. In addition, I am still in contact with both eldercare units for continual information exchange and updates. The collected data consist of notes from these observations as well as formal interviews and informal conversations (small talk) with staff during the observations. Both eldercare units are part of the public care system in the municipality chosen for the project. The field notes along with this text are a part of a larger project, my dissertation.

A regional ethics committee approved this study, and at both units, all staff members agreed through their managers to my presence, having been informed about my presence and the aim of my study before my arrival. Both managers informed me that everyone had approved my being there and that they were looking forward to the project, so if any of the staff members felt discomfort or objected to the presence of a researcher, that information was not passed on to me. At both eldercare units, I sent out two information letters before my arrival, one addressed to staff and one to the residents and/or their family members. In both letters, I enclosed my contact information so that I could address any questions or thoughts. In the letter to the residents and/or family members, I also specified a time when I would be available at the eldercare unit to talk to them and answer their questions. At both eldercare units, no one approached me at the specified place and time. The residents and their family members whom I talked with during the observations about the purpose of my stay with them expressed an interest and a positive attitude about the study. Nevertheless, I do not fully know what staff members thought about my presence before and during my time as a participant observer. I admire their openness throughout the project while subjecting themselves to what I can only imagine must have felt like an intrusion in their work.

Gold’s ‘observer as participant’ role has been questioned as an ethnographic approach for the potential flaw of keeping too much distance from the participants in the setting (Bryman, 2001) but initially seemed most appropriate for this project. Because the purpose was to acquire as much information and as many direct quotes as possible during the observations and to capture situations into rich field notes, simultaneously engaging in daily activities seemed like a difficult prospect. I focused on ensuring that I could take notes during the field studies in order to register quotes from participants. On the other hand, I was clearly noticeable in the setting: Anyone who enters as a new member is noticed and recognized as a stranger and then after some time becomes a member. Having some experience in the eldercare sector (as an assistant nurse and later as a manager), I expected to be accepted eventually and then not noticed much during my stay. Further, I expected to be invited into the interaction in various situations of the daily life at the eldercare units, typically by the residents. Because they would have seen it as strange or impolite of me to decline, I decided to accept and participate at times. My strategy during the observations, therefore, was not to actively include myself but to participate or respond to questions when staff or residents invited me to do so. My intention initially was thus to stay in the periphery as much as possible to study and take careful notes.

After completing the observations, my analysis initially entailed reading and transcribing the field notes. The interaction ritual chains (Collins, 2005) became my tool for analysis of the field notes. The identified symbols in the groups of staff are illustrated by examples that I translated from the field notes. These examples were further discussed with another dual-language speaker. Because this article aims at exploring participant observations as a research method and the role of the researcher, the chosen examples reflect this aim. They also relate directly to the question I posed: (How) do roles for the researcher in participant observations change during fieldwork?

Member checking (Sandelowski, 2008) or member validation (Sherman Heyl, 2001) as a way of ensuring validity in qualitative research is a debated issue (see also Foote Whyte, 1993). In this study, I see member checking not as a separate activity but rather as a part of the observation process. I would ask follow-up questions of the staff members to clarify things they had said or situations I had observed. The interpretations are, however, mine. I recognize the complexity of presenting situations in a way that everyone in the setting would agree with. Consequently, one cannot strive to give a true picture but rather can give only one version of these situations.

Setting: Daily life at Bayside Park and Stonewood Manor

The first unit, Bayside Park, is a unit for people with dementia. Ten older people live in their own apartments in the unit, connected by a kitchen and common area. Bayside Park is a locked unit, and my physical access to it depended on the staff. Although the apartments were kept unlocked, staff entered them only when the resident was present. During my observations of staff in their work, I would accompany them into the apartments. The second unit, Stonewood Manor, ¹ also consists of separate apartments for older people but is not restricted to people with dementia. It is a bigger unit than Bayside Park, with apartments connected to the elderly residence. My access to Stonewood Manor also depended on staff; I would follow them to different apartments and had to ask them to unlock the staff toilet, as needed. At both Bayside Park and Stonewood Manor, I had no place to withdraw during my time in the field.

My examples from the field

Because I spent hours and long days in the two staff groups, I soon became a recognized visitor and later just ‘someone’ or ‘even one of them.’ My presence included me in the groups and affected staff as well as the interaction in the groups. Coming in as the observer, I was initially excluded in the interaction rituals in both groups. Although I had some experience working in the field of eldercare, each place was new and each group of staff had their own rituals and common language. It did not take long, though, before the characteristic joking was also directed towards me. Lennox Terrion and Ashforth (2002) have analyzed the effect of humor in shaping a developing group, and Witkin (1999) has described the function of humor in social work as crucial. After spending some time doing observations, I realized that I was no longer thinking about how to react. Instead, I was focusing on the interactions in the group and reacting instinctively in each situation; instinctively, or with a ‘gut feeling,’ as Lipson (1994) discussed and emphasized.

The flexibility I initially chose for my roles for the participant observations both enabled and averted interactions between the staff and myself in the observations. The interaction first and foremost seemed to relate to the attitude and positioning of the staff members, in addition to the attitude of the residents, as the examples will illustrate. Some of the staff members would interact with me and ask me questions as well as share their experiences from work. They approached me and let me partake in their everyday life at the units, seeming to appreciate meeting me as someone showing interest in their work and their situation. Other staff members kept a distance throughout the observations and did not initiate any kind of conversation with me. A few of those staff members even seemed uncomfortable with my presence, asking me what I was writing in my note pad. Moreover, the duality of the flexibility I chose is illustrated on the one hand by the rewarding conversations the staff initiated, in terms of reflections shared, and on the other hand by the sense that some staff found the observations restricting; one staff member said that she found it difficult to explain to a resident who I was. The more I read the field notes and reflected on situations that occurred during the observations, the more I realized how complex these situations were.

Examples from my field notes ² illustrate symbols in the groups as well as my involvement during the participant observations and fall into four categories (further illustrated in Table 1 below). A common denominator among the examples is that the care workers or the residents actively included me in the interaction. I did not initiate or lead the occurrences; the participant in the setting did. Although I did have some agency while in the field, I did not often actively seek to use it. In the analysis of the data, when focusing on interaction and my own inclusion in the field, I identified two symbols in the interaction between staff members as well as between staff members and residents (in line with Collins, 2005). These symbols are humor and frustration, manifested separately or intertwined together. My field notes are filled with jokes and laughter but also with signs of frustration over working conditions for care workers or living conditions for residents. This section and the four examples conclude with Table 1, in which the symbols are clarified.

OPEN IN VIEWER

Table 1.

Interaction symbols and initiation of various observer roles.

Interaction symbols	Occurrence	Outcome
Example 1 Humor (and frustration) as a common symbol between staff and residents	‘Are you writing a dissertation at 7 a.m.?’	The resident addresses me in the conversation. I am included in the interaction between the care worker and the older person.
Example 2 Frustration as a common symbol, in this instance displayed by the resident	‘This is what happens at eldercare units.’	I am included in the resident’s attempt to raise his voice; the interaction is between the resident and me. I become a tool for the resident.
Example 3 Frustration as a common symbol, in this instance displayed by the care worker	‘Did I say she had a stroke?!’	Staff use my presence, and I am included in the outcome of a conversation. I become a tool for the care worker.
Example 4 Humor (and frustration) as a common symbol	‘You can work a little, too, can’t you?’	I am included, by staff, in eldercare work.

Reflections

Gold (1958) describes the four different roles in fieldwork as ‘social interaction devices’ (p. 217). This description emphasizes the importance he assigns the researcher doing fieldwork, where the researcher is the main conduit through which important data are collected. In the description of the interaction devices, a parallel can be drawn to a mechanical device. This view of the fieldworker’s role as something static implies an ability to distance oneself from impressions and to stick to a chosen role at all times. This option seems neither possible nor desirable.

The roles of the researcher in participant observations not only change once or twice during the course of fieldwork but also change continuously. Consequently, fieldworkers are required to shift roles in different situations when observing a variety of people and settings. Moreover, the fieldworker must be aware of his/her influence in these shifts. When Gold (1958) talks about the need for a fieldworker to select and play a role, the underlying assumption is that merely one role is sufficient. In this study, my experience was that this was not the case. The ambition to understand people and a setting means a need to be open and flexible in the interaction. The researcher alone does not merely choose or shift the role or roles but the participants do, as well, as developed in the interaction with them. This need for multiple roles and their change is thus not simply the result of my positioning as the researcher. The roles are mainly a response to the participant interaction in the study, in this case the care workers and the residents at the two eldercare units.

Adler et al. (1986) state that Gold’s traditional roles of fieldwork neglect to recognize the situational factors that ‘help shape the nature of fieldwork roles’ (p. 372). The combination of ethnomethodology and existential sociology as the background or inspiration for the participant observations, as suggested by Adler and Adler, can encourage or enable the interaction desired for the observations. These interactions provide rewarding insight into the setting studied.

Peter Kloos (1969) presents another dimension to this discussion and speaks to the dilemmas in which fieldworkers can find themselves. Kloos’ definition of fieldwork is ‘that research situation in the social sciences in which the investigator goes to live in close contact with the group he wishes to study—makes the investigator both an observer and a participant inhabitant of the community’ (1969: 509). When discussing role conflicts in social fieldwork, Kloos refers to three reference groups evident to a fieldworker: the group studied, his own society, and the scientific community (1969: 509). Kloos argues that in the field, conflicts may arise among these reference groups at various times, creating a conflict for the fieldworker and causing emotional strain. While avoiding these conflicts altogether is not possible, a fieldworker should recognize them and the strain they cause.

The two groups of staff both displayed an array of symbols that constitute boundaries between them and others, in line with Collins’ (2005) theory on interaction ritual chains. As the observer, I was initially excluded from the interaction rituals and the group solidarity as accentuated by mutual symbols. The common language was unfamiliar to me, as were the jokes between the staff members, and it was evident that I was an outsider observing groups closely. However, this period of time was short. It did not take long before I started to recognize the familiar patterns of behavior and interaction towards each other and also towards me. Jokes and laughter can be a rite of passage; they were symbols for both groups, and I was included in internal jokes. Furthermore, the jokes could be interpreted as a way of dealing with the discomfort some felt about my presence in their safe environment. The fact that I was included in the interaction with the care workers might not have surprised me, but the short time that passed before I was included did surprise me. Initially, the frequency of my being included in the group interaction chains caught me off guard. Before long, I was ‘the Shadow,’ as a joke between the care workers and myself. The response from me in these situations in which I was asked or told to participate by either staff or residents was not always straightforward and easily decided. In the words of Tinney (2008), ‘I was occasionally faced with requests to help either residents or staff, which I found difficult to refuse but difficult to fulfill’ (p. 222). This quote illustrates just how complex the situations in participant observations are. Decisions about the deepening of involvement must be made throughout the process of participant observations; it is not possible to decide beforehand on a strict role that will work in every situation and every interaction with participants in the study.

Concluding remarks

This article explores participant observations as a research method and as a practice. The initial question was, (How) do roles for the researcher in participant observations change during the course of fieldwork? The issue holds no simple answer. The question rather contains more questions within itself: What is the environment observed, who is observed, and, finally, who is the observer? There is no final answer to the role of the observer but rather a number of answers and several roles. Linking Collins’ (2005) theory on interaction ritual chains to the examples from the field provides an understanding of the importance and necessity of interaction between the participant observer and the group being studied. Furthermore, acknowledging the symbols used in a setting also gives an understanding of how this interaction is developed and how much it varies. The participation takes different shapes and forms in different situations with different participants.

I argue that participant observations change in the interaction between the researcher and the participants and that therefore referring observations to a number of roles is restrictive and limiting. Participant observations are a process and should be recognized and respected as such: a process requiring the full devotion and sensitivity of the researcher.

I introduced this text with a quotation from my field notes, and it will end with the same quote:

I am sitting in the staff room at ‘Stonewood Manor’ ¹ accompanied by care workers and a nurse who is working as a substitute. As we have not met before, the nurse asks who I am and what I do there. I describe my presence, and Julia, one of the staff members, fills in: ‘Yes, she wanted to study staff and they sent her here, to the very worst ones!’ Laughs. ‘We call her the Shadow.’ (Care worker at Stonewood Manor)