The Social Relations of Ethnographic Fieldwork: Access, Ethics and Research Governance

Abstract

The focus of this methodological paper is to discuss the challenges of conducting fieldwork, using reflections from our experiences of accessing a research site for ethnographic data collection. The research project aimed to explore nurses’ social relations in their workplace and the inequities between and within these relations among nurses of diverse social positions. Due to the sensitive nature of this topic, access to the research site posed several challenges and was further complicated by the bureaucratic ethics process that governs clinical sites in Australia. Although this study was considered a low and negligible risk research, negotiating the ethics process was full of hitches and hindrances resulting in the refusal of access. This paper offers ethnographers a reflection on challenges in accessing clinical sites to conduct research and a discussion of strategies that may be useful to navigate and counter these challenges by managing social relations in the field.

Introduction

Research literature predominantly discusses the findings of research, reviews of published articles and books, or develops arguments on a variety of concepts. Papers providing accounts of methodological challenges are comparatively rare. This paper tries to fill this gap in the literature by bringing methodological issues faced by ethnographers to the foreground as social relations of the field. Ethnographers navigate through various organizations and key people in the process of gaining access to the research setting and to participants for data collection. This navigation constitutes fieldwork as a social process and this paper outlines our experiences in dealing with people and interactions during various stages of fieldwork. These stages involved the initial approach to the field, gaining ethics permissions from the Human Research Ethics Committee, obtaining site approval, the sign-off from Research Governance Office, information sessions to potential participants, recruitment of participants, and data collection through observation and interviews. Although this study was categorized as “Low and Negligible Risk” by ethics standards, the negotiation of access and ethics processes challenged that framing. While the focus of this paper is to discuss these challenges in gaining access to the research site and how we managed these social relations, the context of the study is relevant to this discussion. Thus, in this next section, we introduce the study and summarize the methodological approach and findings.

The Study

In line with global migration, nurses in Australian workplaces come from an increasing variety of social and cultural backgrounds with various levels of skills and experience. These differences intersect to construct the social relations of work; hence, a study was designed to explore how these multiplicities interact and relate with each other on a day-to-day basis in the workplace. Since intersectionality offers a lens to see the multiple forms of exclusion resulting from the complex intersections of social, religious, ethnic, racial and gender categories (Crenshaw, 1995), it was a key conceptual framework used in analyzing and interpreting the subtleties of social relations between nurses, including the everyday effects of nurse migration. Trust and reciprocity were found to assist nurses in creating a sense of belonging. Support at work counted in opportunities for professional growth, learning and development within the workplace. These findings describe how support and opportunities were distributed across different categories of nurses working together and how nurses made sense of the social and power relations occurring between themselves and their senior colleagues who distributed opportunities and resources in the workplace (Nepali et al., 2023). These relations were constituted as patronage networks that disadvantaged immigrant and coloured nurses, who are visually different to white skin colour and racialised (Dyer, 1997; Nepali, 2020). Such disproportionate access to resources and opportunities not only rendered the latter clinically and culturally unsafe but also made the workplace uninhabitable for them (Nepali, 2020). Thus, despite the necessity of employing a diverse workforce, immigrant and coloured nurses became numbers, upon which diversity claims were made.

Recommendations from the study focused on understanding the importance of staff development through cultural safety as a key to better supporting culturally diverse nurses by respecting their identities and recognizing their expertise. The support programs should include orientation and training on self-awareness of white privilege and cultural safety. It is important that nurses in senior positions realize how their leadership styles affect relations and learn how their approaches to managing staff may be influenced by whiteness and white privilege. Whiteness is a form of systemic, institutional, and interpersonal logic that excludes, marginalizes, and disadvantages immigrant and coloured nurses (Nepali, 2020). Nurse leaders have the power, opportunity, and obligation to lead the efforts of reducing inequality and introducing practices of cultural safety that counter the structures of whiteness and its accompanying racism. Bedside nurses as direct care providers, can also learn how everyday work life can be affected by power and patronage relations, and how this awareness can be turned into resources for improving conditions of work for all nurses to more equitably “have a good shift.” As nurses are shift workers, “Having a good shift” is conceptualized as a day at work, filled with a sense of accomplishment, satisfaction, and positive social relations with and in collaboration with co-workers and patients on the day.

The workplace in the context of this study is a Neonatal Intensive Care Unit (NICU), where the sickest babies are cared for. It was chosen for its unique workplace characteristics demanding unique challenges in social relations of work because of the high acuity of the work environment, the unique needs of the patient population including the complex care necessitated by their physical size and their early life developmental care needs, and the accurate decision-making required for their labile clinical conditions. The NICU thus provides an interesting context to study the complex social relations among the diverse nursing workforces. Since ethnography enables the examination of everyday social practices in the workplace, the approach was adopted to record such relations, capturing details of how nurses’ multiple relationships construct a workplace and shape the social relations of work that account for care provision.

Nurses’ social and professional diversity relevant to their social relations in the workplace includes age, gender, country of birth, education, work experience and position held in the unit. Their ages ranged from their 20s to 60s, education from bachelor to doctoral degrees, experience from novices to four decades of work, job positions from registered nurses to managers, nurse practitioners, and researchers, and background from local to overseas-born and educated. Out of 76 participants, 25 (33%) nurses were born outside Australia. Their countries of origin were the Philippines, India, the United Kingdom, New Zealand, Malaysia, Zimbabwe, Canada, China, Germany, South Africa, Sweden, and the United States of America. Of these, 7 were “white” and 18 were nurses of colour. Of the 18, 2 were born in Global North and migrated to Australia, 4 were born in Global South and migrated during childhood, and 12 were born and educated in Global South. Of the 51 nurses born in Australia, 2 identified as Aboriginal Australians, 1 was from a country in Africa, and the other was from a country in the Middle East. The two Aboriginal nurses were light-skinned and shared their identity only during the in-depth interviews of this study.

While qualitative methods focus on a topic of exploration rather than a representation of a target population, an ethnographic approach considers the population and its culture within a social setting, where temporal and spatial factors are relevant to the analysis. Hence, 18 months of fieldwork involved 100 hr of intensive observations on 76 nurses working in the unit, covering 24 hr a day and 7 days a week including short casual conversations, as well as 65 in-depth semi-structured interviews ranging from 30 to 90 min with those nurses who agreed to take part in longer interviews. The interviews were audio-recorded and transcribed verbatim. The data were stored on a password-protected computer until their analysis, and then in the university research data storage system. Quirkos (a qualitative data management software) and iMindMap (a theme organizing software) were used to sort the data, which was then interpreted using the theoretical lens of intersectionality, to explore the qualities of social relations between nurses across various social positions. The in-depth interviews took place at the later stage of fieldwork after the observations, addressing the questions derived from those observations. This approach of triangulation not only reduced the possible bias of using a single approach to data collection, but it also helped ensure the trustworthiness of data.

Ethnography

Ethnography examines “social interactions, behaviours, and perceptions that occur within groups, teams, organizations, and communities” (Reeves et al., 2008, p. 512), and therefore, like many other areas and issues of exploration, it provides depth and richness in examining the everyday social practices in the nursing workplace. As it enables researchers to explore how people see themselves, how they regard one another and also how they view the situations they face (Hammersley & Atkinson, 2019), this methodology allowed us to tell stories about the daily lives of nurses working in a NICU, viewing and interpreting their behaviour and practices through a “cultural lens” (Fetterman, 2010). The use of an ethnographic approach thus provides two-fold benefits: it serves as a research methodology exploring the culture of complex workplaces such as the NICU of a hospital; and in the end, becomes the report of the findings, serving as the written account of a research project (Hughes, 1992; Savage, 2000). Ethnographic fieldwork is both the heart and the trademark of ethnographic research, which uses empirical processes in gathering the data required by the researcher. These processes allow observation of the nurses’ actions and interactions, which form objective data; and talking with them about their perceptions and experiences of such interactions and relations that shape subjective data or meanings ascribed by participants (Ryan, 2018; Ryan-Nicholls & Will, 2009). Fieldwork does not occur without access to the field and its people living or working in that field. However, access in ethnography is particularly problematic (Bryman, 2016; Duneier, 2011; Gobo, 2008; Hammersley & Atkinson, 2019; Toffoli & Rudge, 2006) as it requires constant negotiation at different stages of fieldwork. The next section illuminates how access can be problematic and problematized.

The Access and Ethics Process

While access challenges are typical to ethnographic research, the process for this study was especially complex due to the possibility of exposure to sensitive issues such as racism in the context of social relations of work. Requesting permission for the site was only the first step, the more complicated processes involved navigating the bureaucracy of ethics committees to gain approval. The challenges ranged from edits to the forms to a complete refusal of access to the proposed research site. In the next section, we discuss these issues and how we managed to gain site access for data collection.

Initial Approach to Field

In our correspondence to the network of Clinical Nurse Consultants (CNCs) in the NICUs in Australia to find out the feasibility of the study, we received positive responses from a number of NICUs with a proportion of immigrant nurses. Across NICUs who responded, the proportion ranged from a scant number (only three immigrant nurses) to one-third of the total number of nurses working in the units. The NICUs also varied in size from 10 to 40 beds, and intensity (acuity) level from one to six. We chose the one that had more diversity, capacity, and acuity. When we corresponded to the chosen NICU, the Head of the Department responded positively, “We are just the place!” and “It’s clearly an important topic.” At the same time, there was an expressed concern that the study might “fuel racial behaviours.” This concern was addressed through an assurance that the study would explore how social relations are developed among staff within a workplace, and that close consideration of these relations would offer opportunities for constructive changes. Rather than fueling racism, this study could contribute to understanding, and thus support a tailored approach to dismantling racism and supporting improvements in workplace culture.

A coincidental introduction to the CNC from the intended research site eased the situation and led to an arrangement of a preliminary meeting. The Head of the Department had copied their reply to all concerned and left the decision to the unit’s nursing leadership. The meeting at the research site proved to be fruitful as both the CNC and Nurse Manager were supportive of the project. However, like the Head of the Department, they also expressed their concerns about the possibility of finding a serious issue. They were assured that we would maintain the utmost privacy of the unit as well as the participants and comply with the ethical conduct of research. This is similar to what Bryman (2016) suggested about dealing with people’s suspicions and worries while securing access to the field. The meeting concluded with the agreement that we present the study to nurses after gaining ethics approval and that the CNC be the contact person and supervising investigator at the site.

The Process at Human Research Ethics Committee

We initiated the ethics process with an enquiry to the Ethics Manager at the Human Research Ethics Committee (HREC). Reading the guidelines set by the Ethics Manager as well as the procedure mentioned in the Online Ethics Forms and The National Statement on Ethical Conduct in Human Research 2007 (National Health and Medical Research Council, 2018) we determined the category of our study as a low and negligible risk research, as it had minor risks to the participants. Murphy and Dingwall (2007) support this justification that the harm to the participants in ethnography is more indirect than direct, open to interpretation and negligible. We, therefore, completed Low and Negligible Risk and Site-Specific Assessment forms and prepared the Scientific Protocol and Participant Information and Consent Forms. When we submitted the application, the Ethics Officer checked the application as if it were complete and ready for HREC review and allowed us to fix the technical matters such as the use of the Local Health District form templates. The Scientific Advisory Committee reviewed the application and sent us a letter asking for clarification on a few issues raised by the reviewers, which we addressed and submitted.

Planning of the Fieldwork

We had completed the preliminary stages of the ethics process and waited for the outcomes. While waiting for the HREC approval the first and third authors had a meeting with people at the research site to thank them for their assistance through the ethics process and to ensure their continuous support for on-going access (van der Waal, 2009). The CNC shared the HREC letter of ethics approval, which was a surprise for us as access is usually problematic in ethnographic research (Duneier, 2011; Gobo, 2008; Toffoli & Rudge, 2006). The CNC gave us an orientation to the unit which helped us get to know the “lay of the land” (van der Waal, 2009, p. 31). The CNC also introduced us and our project to the nurses, who looked interested. Prospective data collection arrangements and ongoing support for the upcoming fieldwork were agreed upon. As the CNC and the Nurse Manager felt to be sponsors rather than gatekeepers to the project, we did not imagine any other hindrances on the way and took the approval from the Research Governance Office for granted.

The Research Governance Office: A Bureaucratic Hitch

The Research Governance Office is another ethics organization adjacent to HREC with the mandate of site approval for a research project to take place. The screening process through this office proved more challenging than HREC. The e-mail from the Research Governance Officer required us to: clarify the project beyond the research protocol, which included confirming statements on the scientific protocol and Low and Negligible Risk and Site-specific Assessment forms; having the site CNC document their commitment to the project (despite already committing to site Primary Investigator on the HREC); submitting an insurance certificate; becoming accredited by the bridging agency between the university and State Health authorities to conduct research activities within the Local Health District; and attaining departmental sign off from the Executive Nurse Manager, instead of the Head of Department. This last requirement was a clear deviation from the usual procedure and an extra layer of bureaucracy. When we checked with the officer as to why the Executive Nurse Manager was to sign the Site-Specific Assessment instead of the Head of Department from the research site, the Research Governance Officer replied that the policy recently changed. However, this change was not documented in the current policy directive, which was not due for review for 5 years and had not changed since coming into effect 3 years prior. However, there was no choice except to follow the instructions, so we enquired with the recently appointed Executive Nurse Manager about obtaining support.

While waiting to hear from the Executive Nurse Manager, we worked on the issues raised and the documents required by the Research Governance Office. The site CNC replied immediately with their willingness to be the Principal Investigator in the study, which was already attached to the application. The university provided an insurance certificate to cover the research activity, even though the insurance and indemnity were not applicable for low and negligible-risk studies. The requirement to obtain accreditation from the bridging agency between the university and State Health authorities was a process intended to accredit clinical nurse educators and authorization of student supervision. The university refused, arguing that because a doctoral student was a researcher, not an undergraduate nursing student who would require clinical supervision, this accreditation process was not relevant. This requirement by the Research Governance Officer arose from an assumption that nursing education (including doctoral research) required a clinical placement.

The “Worst Case Scenario”

Although the course of action was complex, we were navigating the process and providing the Research Governance Office with the requirements. Thus, it was a surprise when access was denied. The CNC and others onsite were also surprised and “totally in the dark,” so it appeared that the decision was made by the Executive Nurse Manager without consultation.

The rationale given was that the study lacked alignment with current organizational goals. It was difficult to understand how NICU’s goals were different from that of the organization, and why research with the potential to improve workplace relations was not a priority. The possible avenues for of reversing this decision were sought with no success. Communication was made even more difficult by several changes in the organizational roles and personnel.

We did not know the “behind the scenes” politics of the organization. There had been an organizational and managerial restructuring that changed the support for the study (loss of the Head of Department), and the new Executive Nurse Manager who we had not met with initially. These shifts constitute the politics of research access (Browne & McBride, 2015) requiring researchers to develop knowledge about changes in power and hierarchies, which might have brought the Executive Nurse Manager into the position. Despite the shifts in politics, the mission statement of the Local Health District remains the same; it supports research and education to provide quality health services to the people. To ensure quality health service, organizational politics are expected to warrant the research supporting the workers’ social relations.

The Search for an Alternative: A Dilemma

Since access was denied, we had to start again. Seeking an alternate research site was not successful as some of the NICUs that were interested at the beginning had now had other research commitments, some had insufficiently diverse staff, and the others belonged to the same Local Health District where access was refused. With much effort, we came up with the option of using the first author’s workplace. This idea had, however, two opposing effects: one was a solution to the problem and the other was a loss of employment. To research one’s own colleagues while working together would cause a conflict of interest and as Pellatt (2003) argued, was unethical and carried a risk of analytical problems. A choice was thus required: abandon the project or quit employment.

The study was deemed a priority over employment. At this new research site, the subject and aims of the study did not need to shift, as can happen with ethnographic research such as Plankey-Videla’s (2012) study of a garment factory in Mexico, where the factory was closed 8 months after her ethics approval. The NICU had a similar high acuity, capacity, and diversity of nursing staff, which fit the design of the study. The Head of the Department at the new research site suggested that the Participant Information and Consent Forms be only for nurses as the main participants, in view of the impracticability of consenting the secondary participants: the doctors, allied health professionals and the parents of babies, who nurses interact within the process of care provision and are observed alongside nurses. It made our work lighter and the recruitment process easier as we could obtain verbal consent from the secondary participants at the time of observation. However, this prospective ease was threatened by confusion among the new Research Governance Office, the site CNC who supported the transfer of previous HREC approval directly to the new HREC, and us. Confusions were worked through, and overall, access to the new site was easier. The Research Governance Officer worked according to the written guidelines. The leadership at the site and the nurses were engaged and supportive of the research. A collegial relationship previously built supported the collaborative relationship in the field (van der Waal, 2009).

In the Field and With Participants

While access to the research field may be easy, hard or even impossible, continuation of access for the whole data generation period can also be demanding. Hammersley and Atkinson (2019) and Neyland (2008) refer to this process as field relations, which requires establishing and maintaining relationships with the people residing or working in that field. It is the art of an ethnographer’s work and can entail physical and emotional labour (Carmel, 2011). From our experience of this project and the ethnographic accounts of various fieldworkers, we could not be assured that it was free of risk. Anything could happen at any time or in the end like Toffoli and Rudge’s (2006) case, where the researcher collected data working in the field for a year and intended to end the fieldwork with an interview with the nurse executive. The response to this interview request surprisingly led to the suspension of the study and restrictions on the publication of study results because of imputed reputational damage from the findings of the research. Thus, initial access does not guarantee ongoing access (Hammersley & Atkinson, 2019), or completion of a study (Toffoli & Rudge, 2006), and the acceptance from the research site and ethical approval from the HREC do not warrant the site-specific approval and the commencement of the study. This requires careful maintenance through social relations in the field. This section details how we experienced the process of dealing with key people and the research participants in the field.

Entering the Field: Dealing With Key People

Doing fieldwork in ethnography meant a challenge dealing with the key people as gatekeepers to enter the research setting and the participants’ everyday lives. Keeping in mind the strategies recommended by Bryman (2016), we first addressed the questions and concerns raised by key people in the unit including Neonatal Consultants, senior nurses, researchers, and other healthcare professionals in the initial and subsequent meetings. They were concerned about the safety of the patients, who were not the focus of the study. They also enquired how we would control bias in the study, measure the interactions and relations and maintain validity and reliability. We explained the nature of a qualitative study, particularly ethnography, which is guided by the principles of trustworthiness and practises triangulation for the accuracy and integrity of data (Hammersley & Atkinson, 2019). The most interesting was the request not to video-record the observations and interviews, which was never proposed as a method of data collection for this study. The reason for this request was an earlier experience of being video recorded by a researcher, which had been uncomfortable. An explanation of ethnography, with reassurance, was all that was needed, as they were new to social research.

Second, the first author as a past employee knew the layout of the unit and who to approach when required. However, when requested help to organize information sessions for nurses, the senior nurses expressed promise to help to the researcher in front of the senior management of the NICU but in practice were dismissive and did not help in the process. This treatment shifted the researcher from insider status as a neonatal nurse to an unwanted outsider. Third, we always maintained a non-judgmental approach when speaking to key people. Fourth, we kept the information that participants gave us private, did not discuss anything related to the study with anyone on the unit although at times the key people asked to provide a briefing of the findings, and changed the supervisory panel for the study to protect the confidentiality of the participants.

Fifth, the researcher refused the advice to wear a high-visibility jacket as this would alarm the participants of her presence, which would increase the risk of the Hawthorne effect on their behaviour affecting the accuracy and integrity of data. The Hawthorne effect relates to behavioural modification of the participants when they know they are being observed and poses challenges in gaining valid and credible data (Oswald et al., 2014). The researcher also chose not to wear a nurse’s uniform, as she had resigned from her clinical role and a uniform could cause role confusion, and she could end up being asked to assist nurses as Fine (2008) experienced in his study of restaurant work, where he ended up helping out in the kitchen. Last, we prepared ourselves for situational changes, as experienced by Whyte (1955), who asked a man and two women in a pub if he could join them; the man stared at him and offered to throw him down the stairs. Whyte could not study the secrets of this trio. Consent in ethnography is thus precarious.

Consenting and Recruitment

In the information sessions explanations were provided about the study, methods of data collection, rights of participants, privacy and confidentiality matters, possibilities of harm, and who to contact if they had questions or concerns. These constituted fully informed consent, which led to the recruitment of participants in the study. The nurses reading the participant information and signing the consent form allowed the researcher to enter their social worlds, their everyday lives of caregiving and interactions. However, although the researcher was familiar with the setting, participants, and daily activities, collegial relations developed with the nurses shifted to a researcher–participant relationship, which carried suspicion for some nurses. As a result, the process of recruitment was sometimes fast and easy, and other times challenging and time-consuming. When the consent did not happen spontaneously, we changed our strategy to be proactive. The first author started “hanging around” (Browne & McBride, 2015; Bryman, 2016), making herself available for questions and reminding nurses through her presence. She also followed up with nurses individually, which resulted in an increase in signed consents. A few required four or five attempts of follow-up, and some were recognized as non-consenting. While the trusting relations as immediate past colleagues worked to facilitate some nurses’ spontaneous consents, she continued social relations with those, who declined their participation. Such a non-judgmental approach helped attract the participants as some initially were wary but later felt safe to participate.

Fiona, one of the nurses working in the NICU for example, took a long time to decide. She sounded interested at the beginning, but later changed her mind. As she would be a good source of data for her experience and background, we did not want to lose her and found another opportunity to talk about the study. The following fieldnote excerpt describes this interaction.

She was caring for a baby in bed 7 when I approached her. We chatted about our whereabouts as it was a long gap of our seeing each other. I then asked her intention to participate in the study. ‘No, I was going to but [looks down] actually I have so many things to work out, and I can’t do them at once,’ she showed her problems. I described the process of observation and tried to assure that she would not be disturbed and asked to do anything special that would be time-consuming or mental labour. ‘I know but. . . I am not good at doing many things,’ she was not convinced. I explained further that the only task for her was to read the information sheet and fill in her details if she decided to participate unless she agreed to a longer interview later and that I would not disturb her work and she would not notice my presence to watch her. ‘I will have a look. . . I have a look’; her face brightened up a bit. I was extra cautious not to pressure her, and so assured her that the decision was totally up to her and that even if she did not consent, our relationship would remain the same. (Fieldnote 28, p. 1)

At that stage, Fiona was indecisive, which led the researcher to a more personal level in dealing with her. Sitting in a quiet corner of the staff cafeteria and chatting in an informal manner she asked for the consent form and signed it.

Persistent follow-up of potential participants is not coercion but rather is a standard ethnographic practice. The researcher needs to “hang around” and explain the study to people to demystify the process and their participation. Some people need more time, explanation, and encouragement to think and decide. Obtaining the consent of the senior nurses was a real challenge. As they were the key people facilitating the study, they were expected to provide their consent. However, despite individual follow-up and requests, they did not sign their consent, which made it difficult to observe the interactions and relations of staff with them, and the events and ceremonies in the unit that their social relations involved. They were reminded in writing:

I know you are busy so might have taken time to respond to my request letter that I had left for you for your consent to my study. I also understand that it is possible to forget in a busy time, so I am writing to you as a reminder.

Obtaining consent from the secondary participants was straightforward. The researcher briefly explained the study when they felt to be part of the scene. An informed option that they could deny or withdraw their inclusion was offered. However, most of the time, they would come into the scene suddenly and it was impossible to explain that they were being observed for the study. As the instant consent would interfere with the sequence of events and interactions, we adopted the strategy of informing them afterwards. When they happened to come in before the observation started, their permission was obtained beforehand. Fortunately, both approaches elicited positive outcomes. We did not have to discontinue observations, exclude anyone during an observation or discard any field notes. We discussed with the Head of Department, who, not only agreed on the fact that it was an appropriate approach but also pointed out that it would decrease the Hawthorne effect.

Observing Nurses’ Interactions and Relations

As nurses were the focus of the study, our observations and conversations surrounded their interactions and social relations, both within the nursing group and with other healthcare team members. We focused on the scenes of teamwork involving procedures, routines, and emergencies to begin with; events, celebrations and meetings held in the unit including social outings in the middle; and individual nurses and their periphery of interactions and relations in the later part of observations. The difficulty in the early stage was deciding whether to observe consented nurses working with those who were yet to consent. Approaching the as-of-yet unconsented nurses for their decision at that moment would result in a loss of data, as the events and interactions would pass by. The recordings of these observations, that is, field notes, excluded the responses made directly by the unconsented nurses but included the ones made about them by consented nurses. On a few occasions, we had to quit and go back to another time of the day. In the middle stage, we struggled with gaining the consent of the senior nurses. At the later stage, as we changed the focus of observation and the number of consented nurses increased, it was easier to follow the chain of interactions around each consented nurse.

The researcher’s position was close to the third category of participant observation as an observer, participating less in the daily routines and activities (Bryman, 2016; Fetterman, 2010) of the nurses being studied. However, at times, we helped with small tasks, such as checking alarms and reaching for the equipment on the other side of the cot (bed of the baby patient) and chatted as an acquaintance (especially in isolation rooms). We adopted both active and passive roles in the field. Especially during the initial days of fieldwork, the participants fell silent when they saw us in front of them, which we responded to by either picking up what they were talking about or starting a new conversation, based on the atmosphere. This would trigger their conversation, and as soon as they took over, we would stop talking and start listening, shifting our active role into a passive one—an observer as intended. This approach not only helped relax the participants and reduce the Hawthorne effect but also facilitated the gathering of valuable data. The researcher’s twofold native status as a neonatal nurse and a past employee provided us with professional knowledge, NICU-specific language, and familiarity with the research setting and participants. However, there was a potential risk of bias and conflict of interest, which we minimized by being extra cautious and having the researcher resign from the NICU prior to commencing the study.

We experienced both positive and negative reactions from the participants, which were well expected in ethnographic fieldwork. In an explanation of the study, Zara said, “Such a worthy project, wish I could be the star of it.” Camilla made a joking satire at the sight of the researcher around the office area, “Are you spying [on] us?” Heather, in an early morning in the NICU, expressed her distress, “Go away, we are busy!” While being observed at 01:00, Jeana, an experienced bedside nurse, expressed her surprise, “What are you doing here in the middle of [the] night?” Some looked at the researcher with sympathy, “Oh gosh! I wouldn’t do it at all!” It was indeed a big commitment to stay awake and visit a hospital unit at night to do the “ghost work.” Some other nurses expressed their discomfort with the observations and required explanations and reassurance. Camisa said one evening, “I am not in the mood today, don’t observe me.” Charlie required assurance that her part-time status at work would not be a problem regarding her participation in the study and that the findings associated with the observations and conversations with her would be entirely confidential. Participants put the researcher in all sorts of positions. The comments, jokes, concerns, and rejections were accepted and accommodated in the ethnographer’s account.

Short Conversations With Participants During Observations

Unlike a complete observer, this study used an engaged observation, that is, observer-as-participant. To remain connected with the participants and their perspectives, we asked questions about the events, interactions, and relations while observing the daily activities and interactions of nurses. This approach ensured clarification of matters during observation or for the participant’s thoughts on a particular issue. These questions and answers formed short conversations as part of observations and acted as supplementary information to the research data. These conversations often took place at the bedside given the availability of privacy as it was easier to link the contexts and was convenient for nurses. However, considering the inconvenience for them to be observed and asked questions at the same time, the researcher did not make such conversations initially. Instead, she informally chatted with them and showed them records of their actions and interactions to make them comfortable and develop rapport. While such informal discussions and sharing the field notes offered opportunities to build transparency and check for the accuracy of the observations, the lack of clarification resulted in incomplete data, so the strategy shifted, and the researcher began asking questions about events and dialogues that took place during observation sessions. When the nurses were busy, or there was no privacy, the researcher jotted down the questions and waited for an opportunity to discuss them later. As it was hard for them to leave the bed space in such an acute care environment, they were taken out for very limited occasions for this purpose.

It was necessary to win their trust, as there had been a previous study where nurses had been video-recorded during their interactions and this study had caused considerable conflict in the unit. The responses included Judy’s question, “Will you video record us?,” Camilla’s inquiry, “Will you record our conversations?,” Barbie’s concern, “Will you come between us when we talk?,” Cheryl’s worries, “Can we continue our work when you are observing us?” This required reassurance that no digital media would be used while observing them unless they agreed to record in-depth interviews, and our promise to respect their space by positioning the observer wherever they would feel comfortable and not to interrupt their routines and emergencies.

Writing Fieldnotes

Fieldnotes are the records of observations of nurses’ interactions and relations of work as well as the short conversations with them during and after those observations. Writing field notes commenced at the start of the access and ethics process as these are central to the ethnographic practice. The events, activities, and social processes were recorded in a notebook. During the initial observation sessions, however, precaution was taken not to record in front of the nurses in view of their concerns about being recorded. Instead, time out was taken after each segment of observation to record the interactions that were seen and heard in the clinical setting. As the fieldwork advanced and the nurses trusted the researcher’s activities, they noticed the notebook less, which eased the recording of their social relations of work in the scene. In the unit meetings and training sessions, however, notebooks were not an issue, and they could be recorded instantly. Often, we changed the people and scene on return from the recording, so nurses would not notice the researcher’s absences. But when continued with the same group, they would inquire about the researcher’s whereabouts. This was an easier approach than Spradley’s (1970), who struggled to write his observations of drunk nomads; he would go to the toilet after each conversation to record it—having time out in that way made his subjects wonder if he had an incontinence problem.

Interviewing Nurses

In-depth interviews are “oral accounts” of the participants’ perspectives about the workplace and their social relations with colleagues. The interviews functioned as a cross-check of, and complement to, the observational data, as they were participants’ subjective responses and reduced observer bias. Ethnographers such as Hammersley and Atkinson (2019), refer to this as triangulation—a way of validating data. The interviews confirm what participants thought was happening, rather than reasons any observer could ascribe to activities.

During interviews, some nurses felt embarrassed and apologized for not answering some questions. As embarrassment and guilt are the possible harms caused by ethnographic research (Murphy & Dingwall, 2007), we reassured them with empathy and benevolence. Being empathetic in this way opened up nurses’ emotions and experiences truly during interviews. Kitty, for example, burst into tears while talking about her colleague, who had left due to what she described as “unbearable treatment from management.” Crissy, who appeared to be happy at work and reported having fun on observation, revealed her painful experience of biased treatment in the workplace during her interview.

Having time out for in-depth interviews in such an intensive care environment was challenging for nurses. However, we received cooperation from some nurses who oversaw shifts. Judy, for example, was extremely helpful—she looked for people who could relieve nurses for interviews. One Saturday afternoon, there were two nurses in the same section to be interviewed, and Judy offered a support person to relieve them. Hindrances were also common. For instance, Debbie asked to wait for the end of her shift for an interview on a Sunday evening, but at the end of the shift she disappeared. On repeated attempts, she engaged in an interview, which was interrupted by senior people requiring a shift in location multiple times. Some participants required multiple attempts, thus handling relationships positively is important in the field.

Some nurses readily agreed to interviews, while others found it problematic. Rosy, a Clinical Resource Nurse, checked with the nurses at the bedside if they needed anything (e.g., help with procedures such as intravenous canulation and re-strapping of ventilation tube), and quickly followed the researcher. In contrast, Yara told the researcher, “If I’m in charge I can’t do it because it’s stressful.” On a previous occasion, she had said, “I’m not in the mood, I have so much going on.” Generally, the nurses born and educated locally appeared to be able to decide to attend an interview during the shift if they wanted. Some independently arranged with colleagues to relieve themselves. Conversely, the immigrant and coloured nurses were hesitant to leave the ward unless they were told to by their seniors. Even when they were allowed, they had to feel comfortable to do that. A representative case is discussed in the following fieldnote excerpt.

Myrtle is a young nurse, born (to immigrant parents) and educated in Australia. She was assisting with a re-strapping at Bed 7 when I approached her at 20:00. She was aware of my presence as we had scheduled an interview for tonight but didn’t look at me until I particularly spoke to her. She said, ‘I’m support [person], have two more “re-straps” to do and then will see how it goes.’ Kizzy, who was the nurse to that baby and was doing the re-strapping, said, ‘I will help with one.’ I hoped the re-strapping would take a maximum of half an hour and so waited at the nurse’s station. However, even after two hours, she didn’t seem to be ready for the interview. She did both re-strappings despite her colleague’s offer to help her. After the job was done, she went to each bay and asked if they needed her to do anything for them. Then she came to the nurses’ desk, picked a piece of paper, walked around to ask everybody for coffee, collected orders and went to the shop downstairs. She came up with a tray of coffee servings, distributed what she bought and chatted with colleagues while drinking coffee. She then signalled me. Thus, she spent almost 4 hours before she came out to do a 72 min talk and it was 00:55 by the time we finished. (Fieldnote 47, p. 1)

She wanted to make sure that she finished everything she could before taking time for an interview. She attended to everyone so that no one would question her time out. This might indicate a lack of confidence about her ownership of, and belonging to, the workplace or her relationships with her colleagues. Australian white nurses were confident and able to leave their work for their participation in a research study, while the immigrant and coloured nurses were hesitant to leave the bedside, signalled how social relations operate in the NICU.

Exiting the Field

Exiting the field meant leaving the research participants’ worlds to make their stories ours and share them with the outside world through an ethnographic report. We reflected on the fieldwork and how nurses reacted to our presence in their world as an insider in one moment and as a “spying agent” in another, and how their differences intersected to accomplish the care of sick newborns and make their workplace as vivid as a showcase. We also recalled how many of them were concerned about the privacy and confidentiality of their responses, and the possible effect on their work relations. The underpinning ideas of ethnography are worthwhile. A well-known organizational ethnographer, Van Maanen (2015), acknowledged the potential for learning and wondering by just hanging out with people. How do people work together and to what do they respond? We learned that those who were observed put the observer where they were comfortable. In sound ethnographic practice, the participants position the researcher.

The change in relations between the participants and the researcher was remarkable. Some lost their trust in the transition of position from a colleague to a researcher and were suspicious of the new role, but, in the end, they enclosed her in their worlds and forgot about noticing. Some even regarded the new role as that of an advocate and put all their troubles and concerns on a shoulder to lean on. Inevitably, some nurses casually scrutinized each other’s whereabouts, which resulted in valuable information for the research data.

The research project was well accepted in the research setting, and people were ready for it to be there. Having said that, we waited a considerable time to win their trust, which lengthened the time in the field. But it was worth it. We realized that the participants were comfortable, at last, when they were ready for us to ask questions about their work, and they even gave their time for in-depth interviews about the value of this work, for them, and for others. We gave a presentation at the end of the fieldwork, as the unit was curious about the process. This presentation not only updated the unit on the project, but also fostered an understanding of what ethnography involves and how it accommodates the different traits, cultures, and views of people in an organization. We felt pleased that those who practised trials in clinical areas and had a lot of questions about the ethnographic approach came to understand how this type of qualitative research works. It was an opportunity to thank all the nurses who participated in the study and the people in the unit who facilitated this study. This is all about the social relations of the fieldwork: troublesome as well as pleasant, difficult as well as adventurous.

Discussion

Since gaining access to the field was interlinked with the ethics process, it was a matter of careful coordination, that is, managing social relations in the field. A major component of dealing with organizational ethics bodies, the key people as sponsors or gatekeepers and the research participants constructed social relations of the field (for the researcher). The change in collegial relation to the researcher-participant role caused surprise and suspicion among some nurses. Some even stopped doing what they were doing and changed their behaviours. To minimize this kind of Hawthorne effect we waited until they were used to our presence as researchers. When comfortable, they did not notice the researcher watching them. Lessons learned in the process include knowledge of ethnographic processes, refined interpersonal or social relationships, patience with the participants and faith in our ethics and integrity. These factors shape the strengths of ethnographic fieldwork.

The ethnography with extended observation and interview, the researcher’s insider status and an immigrant background augmented a feeling of disquiet, confirmed by reflexivity. Having to research one’s own workplace had a few issues. One of them was the insider status—although resigned from the workplace before the commencement of the study, the researcher was a neonatal nurse and former employee. Being an insider meant she helped nurses at times and started conversations when they were mute in her presence. However, as Rudge (1995) argued that it was crossing the boundaries of the researcher and was ethically and methodologically problematic, she minimized her involvement in such activities. Since she was open to challenges and questioning the researcher’s positions, the insider perspective instead helped. As Bourdieu (1996) noted, the rapport and lessened power relations between the nurse participants and the researcher meant they confided some things that had previously been silenced or never been able to be mentioned in previous discussions and shared their profound experiences of the complexities of working in the NICU.

Another issue was the risk of bias in observation due to the offer from the researcher’s workplace to become the research setting for this study. This possibility of bias made us aware if we were missing or overlooking any events that occurred during our observation. This awareness, according to Berger (2015), was an agency of self-supervision and reflexivity that maintained rigour. At times, we also faced the dilemma of whether to record and use specific data. The research ethics, study aims, and research questions led us to decide what was essential to the research study. Since we did not see any significant events during observations, which was unlikely, we modified our observation strategies from the scenes and events to the individual nurse and their periphery of interactions. We also extended the fieldwork beyond year-round and repeated interviews with some participants, as suggested by Berger (2015) to ensure rigour through reflexivity. Incidents only emerged when we started interviewing, which not only gave us meaningful data but also strengthened the researcher’s confidence in recording and using the data in the study. Some of the white Australian nurses’ experiences of “patronage relations” also stood out as significant and the underside of white privilege (Nepali, 2020). What nurses talked about endorsed what we saw happening in the observations of nurses at work. Nurses created a different place for the researcher—as an insider but as observant of what mattered at work for them. Observations of the work could then be found to highlight what mattered in what they were saying about the work and their social relations.

There were a few limitations of this ethnographic study, especially the process, which can become its strengths, lessons to be learned and recommendations. The NICU as a single-centred research site could pose questions about its ability to represent other workplaces around Australia and the generalizability of results in an international context. However, workplace issues are broad and studied in many locations worldwide, and since immigrant nurses are employed in various Global North countries, many researchers have explored their experiences using a variety of approaches. For instance, Alexis and Shillingford (2015) demonstrated that immigrant nurses, no matter how experienced they might be, were viewed as novices in their British workplace, where they could not use their skills and lacked autonomy in practice. Moreover, the unit of this study had a large number of nurses working together (n = 88), and this study had a high rate of participation (n = 76). The experiences shared by such a number of participants through observation and interviews ascertain the depth and richness of the issues researched (Hammersley & Atkinson, 2019).

The fieldwork went for an extended period of observations and interviews that included waiting and revisiting nurses at “ghost-times” of the night to capture the one-off events and interactions. This intensity and extensivity produced a vast amount of data that took longer and was harder to manage. This has the potential to register the method as time- and labor-intensive (Hammersley & Atkinson, 2019) and discourage novice researchers from adopting ethnography. However, these temporal and work demands brought powerful stories to the study findings. Similarly, the unstructured nature of data collection, hitches and hindrances in accessing the research setting, and the researcher’s relationships with the participants as an insider made the research practice challenging. We believe, however, that the knowledge of the ethnographic process, strong interest in the methodology, and willingness to face such glitches enable the researcher to develop rich and meaningful data (Atkinson et al., 2007; Hammersley & Atkinson, 2019), which contributes significantly to knowledge construction with a wide variety of issues that are made visible from those data.

Although referring to the unit, hospital, and state policies or disclosing the identity of key people could add more sense to the interpretation of findings, the ethical responsibilities of maintaining privacy and confidentiality did not allow us to do that in view of protecting the identities and the location of the research setting. The national principles of responsible research practice persuaded us to believe that with the power of knowledge, self-reflection, and commitment to moving beyond dominant structural injustices, the organization, and the people in it, are motivated for reform—for their own benefit and the others involved. We also believe, individuals and institutions would thrive when they find the paths to improve through an understanding of their shared challenges and openness to critiques. Allowing social research in their organization could mean their receptivity to the results and willingness to adopt the recommendations. Importantly, what this ethnographic study reported is valuable to nurses’ work-life in this NICU, in other Australian health-care workplaces and in broader contexts, where nurse migration is an issue and more important than methodological glitches.

Conclusion

This paper highlights how access to the field and ethical considerations in ethnographic research are not a one-time process, but instead are continuously negotiated through managing social relations of the field. We described how an initial smooth process of attaining HREC approval turned unexpectedly into a complex challenge within layers of clinical bureaucracy, and how the taken-for-granted access became problematic in an instant. This blocking, despite the departmental support for and significance of the study, was not a situation we initially thought was possible—particularly as it related to how to integrate a significant proportion of immigrant and coloured nurses into a unit that required such nurses to provide care to neonatal patients and their families in a multicultural society. However, the sensitivity of an exploration of inequalities and injustice within the nursing workplace was at the same time threatening to the workplace, and concerns about potential claims of racism posed hindrances in the processes of access and ethics. The processes were challenging and the relations with key people and participants were tricky and taxing. Strategies for access and ethical processes to ensure a rigorous research process involved how we approached the field and entered the everyday worlds of nurses, as well as how we established rapport, maintained those relations and handled the problems that arose in social relations in the field. The strategies discussed in this paper can offer support to other ethnographic researchers who aim to understand and address inequities and racism within the social relations of workplaces.

Author Biographies

Shobha Nepali, RN, PhD is a Registered Nurse at Cumberland Hospital, Mental Health Service, Sydney, New South Wales, Australia.

Rochelle Einboden, RN, PhD is an Associate Professor at the University of Ottawa, School of Nursing, and Research Chair at Children’s Hospital of Eastern Ontario (CHEO) & CHEO Research Institute, Ottawa, Ontario, Canada; Adjunct Associate Professor at Western Sydney University, School of Nursing and Midwifery, Penrith, Australia; and Honorary Lecturer at The University of Sydney, Susan Wakil School of Nursing and Midwifery, Camperdown, Australia.

Trudy Rudge, RN, PhD is an Honorary Professor at The University of Sydney, Susan Wakil School of Nursing and Midwifery, Camperdown, Australia.