Shadowing as Qualitative Inquiry: Exploring its Potential and Limitations in Educational Research

Abstract

The term shadowing is ubiquitously used to describe a variety of research approaches, often without explanation of underlying epistemological assumptions. This paper critiques shadowing in educational research, highlighting qualitative shadowing as a novel method for gathering contextually relevant findings. The paper advances shadowing as a unique qualitative approach by exploring its particularities with reference to the literature and the researcher’s experiences of using the method to observe occupational therapists’ work in ten Australian primary schools. Insights related to accessing the field and ongoing ethical considerations, creating and maintaining the distinct researcher-participant relationship, and the emotional and logistical work involved in shadowing are considered. Findings may support future researchers to understand the advantages and challenges of using qualitative shadowing as a method. By granting researchers entry into the unseen and intimate realms of participants’ professional lives, qualitative shadowing offers a unique way of understanding perspectives, yielding rich data to provide a distinct form of knowledge.

Introduction: Shadowing in Educational Research

Shadowing in educational research has roots in Wolcott’s (1973) seminal study of an elementary school principal and has been used to investigate professional learning (Fox & McCormick, 2009), the work of teachers (Ferguson, 2016; Knutas, 2019; Lee & Poon, 2014), and educational leadership (Bøe et al., 2017; Earley, 2012; Hughes et al., 2022). Although shadowing is widely used, the term describes a variety of approaches, often with little elucidation of underlying epistemological assumptions (McDonald, 2005). Some studies define shadowing and position it as a qualitative method (Bøe et al., 2017; Hughes et al., 2022; Lee & Poon, 2014), while others use it without any clarification, neither defining it nor discussing its qualitative or quantitative nature (Brennan & Mac Ruairc, 2017; Drysdale et al., 2009; Fox & McCormick, 2009; Hoppey, 2006; Wan et al., 2020).

Types of Shadowing

McDonald (2005) describes three types of shadowing, all used in education. Despite their similarities, their aims and methodological traditions differ. The first type, used for the purpose of experiential learning, provides shadowers with direct first-hand experience for its own sake. Individuals learn aspects of the role for which they are being prepared or learn about another’s professional role (Vasset & Almås, 2015). This type is common in teacher education, to support pre-service or novice teachers’ development (Wilks & Ross, 2014; Williams & Sembiante, 2022), and career exploration, as it allows individuals to assess role suitability based on firsthand experience (Reese, 2005).

The second type, originating with Mintzberg’s (1973) study of managers, is used to quantify behaviour. As observed phenomena are organised using structured observation protocols, and actions logged against predetermined categories, this type aligns with a positivist methodology. However, this shadowing may also be positioned as qualitative, depending upon the process of category development which can be abductive (Arman et al., 2012). In education, this type of shadowing is used to examine how individuals spend their time on daily tasks (Horng et al., 2010).

The third type is focused on contextualised purpose and meaning (McDonald, 2005) rather than quantifying behaviours. This shadowing is interpretivist, with reality viewed as subjective and constructed. Researchers adopt participants’ viewpoints to understand their personal experience and perspectives, and patterns emerge from data rather than being predetermined. In education this type is used to understand teachers’ work (Ferguson, 2016; Knutas, 2019), and educational leadership (Bøe et al., 2017; Bøe & Hognestad, 2016; Polite et al., 1997).

The methodological ambiguity of shadowing in education is emphasised in Tulowitzki’s (2019) scoping review, which provides insights into the conduct of shadowing, concurrent methods used, and the strengths and its limitations for studying educational leadership. Of the 29 included studies, 12 lacked a definition of shadowing, while four cited McDonald (2005), Mintzberg (1973), or Wolcott (1973) in lieu of a definition. Tulowitzki’s (2019) reports the studies’ aims as explanatory, focusing on principals’ work, their actions and their behaviours. Shadowing parameters differed between studies. The number of observers ranged from one to 40. The number of shadowed participants ranged from one to more than ten. Observation periods varied from a few hours to more than 5 days. Although most studies emphasised predetermined categories, such as time, place and activity, or model of leadership, some exclusively used an inductive approach. Three studies used only shadowing, while the remaining 26 combined shadowing with other methods, primarily interviews.

The literature on shadowing as a method in educational research exhibits a lack of definition and epistemological grounding. This ambiguity, coupled with the diverse ways shadowing is implemented, necessitates a closer examination of its value as a unique qualitative method.

Qualitative Shadowing as a Distinct Method

Qualitative shadowing focuses on the practice of an individual working within a complex social setting (Gill et al., 2014) to provide the researcher “a rich, dense and comprehensive data set which gives a detailed, first-hand and multidimensional picture of the role, approach, philosophy and tasks of the person being studied” (McDonald, 2005, p. 457). Depending upon the research objectives, the duration of the shadowing may vary from a few hours, to days, months, or even years (Meunier & Vasquez, 2008; Polite et al., 1997; Wolcott, 1973). Czarniawska (2014) describes shadowing as “observation on the move” (p. 43), as participants are followed throughout their work day and from place to place (Sirris et al., 2022). Experiencing the nature and structure of the shadowee’s days leads to an understanding of what it is individuals actually do in the course of their duties, rather than what their role stipulates they do (Quinlan, 2008). As shadowing does not solely rely on individuals accounts of their role, but instead directly observes individuals performing their role, the approach grants access to information that participants may consider trivial or difficult to express. By engaging in ongoing in-situ conversations about social practices, researchers are permitted to understand the intricacies of individuals’ actions and interactions, to reveal the how and why behind their behaviour in specific situations, providing a unique insight.

Since McDonald’s (2005) appeal for research employing and critiquing shadowing, several authors have reflected on and raised awareness of the method, yet scope remains for exploring the approach and its application in different fields (Ferguson, 2016). This paper builds on McDonald’s request, illuminating the idiosyncrasies of qualitative shadowing using examples from the authors’ experience employing the method in education to examine interprofessional collaboration in primary schools. By examining the method through its application, the paper provides insights into the complexities and potential of shadowing as a distinct qualitative research method. The merits are highlighted, and the encountered challenges exposed, to provide direction to others who may be considering using this approach.

Method

Qualitative Shadowing to Investigate Interprofessional Collaboration

Interprofessional collaborative practice between occupational therapists and teachers in mainstream primary schools is an essential factor in supporting all students to access high-quality learning (Tancredi et al., 2019). Qualitative shadowing, with its emphasis on interaction in context, was selected to explore collaborative practice. School-based occupational therapists’ work is often itinerant, with therapists employed across a range of schools, settings, and events. As school-based therapists have multiple responsibilities and roles, shadowing allowed observation of the complexity of these roles, providing insight into the otherwise hidden work of practitioners. Using the method enabled therapists’ real-time explanations of actions and shared reflections on experiences, while the impact of contextual factors was simultaneously considered.

Ethics Approval

The study was approved by the University of Sydney Human Research Ethics Committee (Project Number 2021/278) and the relevant educational governing bodies in the three Australian states in which the study was conducted. Participants provided written informed consent prior to participation. Pseudonyms were assigned to protect participants identities.

Sample

The study was advertised on various social and professional networks. Given participants needed to be comfortable being closely followed, observed, and questioned, recruitment was based on self-nomination. Therapists who met inclusion criteria and expressed an interest in the study were contacted, the study discussed, and consent obtained before field work was scheduled. Eligible therapists were qualified and registered to practise in their state and were working in a mainstream primary school with teachers. There were no restrictions on geographical location. A purposive sample of 10 occupational therapists from 10 different primary schools in New South Wales (2), Queensland (6), and Victoria (2) participated in the study. The small sample size allowed the depth of analysis fundamental to the research (Vasileiou et al., 2018). The sample, which predominantly comprised of middle-aged white women, was representative of occupational therapy as a profession (Australian Government, 2019; Taff & Blash, 2017).

Shadowing duration ranged from one to two school days (between four and 12 hr), depending upon therapists’ employment and services. Occupational therapists were followed everywhere, from the moment they entered the school to the time they departed, in scheduled and impromptu meetings and therapy sessions, in classrooms, staffrooms, and therapy rooms, and through corridors, playgrounds, and canteens. Contextual conversations allowed an understanding of participants’ sense making (Arman et al., 2012), while semi-structured interviews with therapists and their collaborating teachers provided different perspectives.

Insights Gained Through Using the Method in Educational Research

The following sections discuss insights gained through using qualitative shadowing to investigate interprofessional collaborative practice between therapists and teachers in schools. “I” and “my” in these sections refer to the first author.

Access to the Field

Common to qualitative research, issues of admittance to the field are particularly complex in health and paediatrics (Gilliat-Ray, 2011; Quinlan, 2008). Shadowing is not uncommon in education or health contexts, however, qualitative shadowing of health professionals in an educational institution is rare. Shadowing therapists in schools presented unique challenges and opportunities unexplored in previous research.

Obtaining access to organisations is a taxing aspect of any observational research, and gatekeeping is typical (Chaudhuri, 2017; Kawulich, 2011), particularly where issues of confidentiality are central (Quinlan, 2008). Each of the three states had its own child protection regulations, leading to considerations about the administration required to gain access. In school-based research, access to research participants is arbitrated by educational institutions, and may be denied due to reasons of time pressures, institutional inconvenience, or the reluctance of the institution to expose quasi-private worlds to public scrutiny (Heath et al., 2007). Shadowing added complexity due to its unconventional participant engagement (McDonald, 2005). Agreeing to a thirty-minute interview is different to agreeing to be followed around for days, and several therapists who originally expressed interest subsequently recoiled at the required commitment. Therapists’ managers also expressed concern at the requested duration, advising approval would be more likely if the duration decreased, and the number of shadowed therapists increased. Thus, the original plan to shadow four occupational therapists for a 2-week period was amended to accommodate an increased number of therapists and a reduced time period of one or two working days.

Securing approval by modifying the timeframe and intensity of observation is not uncommon (Bartkowiak-Theron & Sappey, 2012) but it brought both challenges and benefits. Additional therapists resulted in increased access requests and movement between locations, however, it also led to greater variability in observations and a wider range of viewpoints and perspectives. A broader scope of practice, and greater variation in employment, role, approach, and activities were observed. Had fewer therapists been shadowed this variety may not have been evident, leading to different understandings about collaborative practice.

Reducing the observation period prevented the same therapist being shadowed across different schools to observe interinstitutional variation (Trouille & Tavory, 2019). Yet, the shorter time period permitted an intense and focused observation, providing a “snapshot” of practice, potentially reducing the bias associated with prolonged interaction with participants (Bøe et al., 2017). It also entailed forming new relationships daily and navigating one’s position in each new research setting. Shorter windows influenced understanding, as the nuances of complex roles or the long-term dynamics between participants were not observed. Similarly, the activities observed may not be typical. The increased number of therapists and the reduced shadowing duration suited the research objectives; however, depth was sacrificed for coverage, and this is acknowledged as a limitation of the study.

Issues of access extended from institutions to individuals, who retain the decision to allow or refuse entry to various aspects of their world (Quinlan, 2008), preserving control over the data by managing what the researcher observes and their interpretations of what is seen (Hughes et al., 2022). Occupational therapists’ access decisions were based on their thoughtful appraisal of the potential reactions of those with whom they would be interacting and considered the impact on students. Anna demonstrated her prioritisation of student needs over the study’s needs when she remarked that observation was permitted only because the student was “super chilled”. Although each therapist carefully considered my continued access throughout the different activities of their day, I was not excluded from observing any aspect of participants’ daily practice. Unfettered access allowed observation of the complexity of therapists’ work with students with disabilities, as well as the complexity of their collaborative practice with the teachers and staff supporting these students. Like Quinlan’s (2008) nurse practitioners, I believe study participants recognised the value of my seeing their work in the full context of their practice.

The one instance where my presence was questioned occurred during a multi-disciplinary team meeting, when a staff member raised concerns about confidentiality. However, once non-disclosure was assured, I was permitted to stay. I was thankful for this, as observing the therapist negotiate their role and work within the team offered a valuable insight into an aspect of their work that would have been otherwise missed. That the legitimacy of my presence was rarely challenged demonstrates the high esteem in which therapists were held by their colleagues. It appeared I was permitted access to even the most delicate conversations simply by association.

Access issues extended to scheduling, requiring careful deliberation on when and where to shadow. Gill et al. (2014) note that the schedule of shadowing must balance both the shadower’s and the shadowee’s needs. As school-based occupational therapists’ work is itinerant and therapeutic in nature, I left scheduling decisions to them, organising my own life commitments around their timetable. Bøe et al. (2017) note that practicalities of transportation and time must be considered when shadowing and confined their selection of participants to those within a 50 km radius of their home. Since several of therapist participants were in different towns and states to me, some more than 1000 km away, extended periods away from home and family were unavoidable.

Ethical Considerations When Shadowing

Ethical clearance was sought prior to commencement of the study; however, shadowing raises particular concerns about ethical practices in the field (Bartkowiak-Theron & Sappey, 2012; Bøe et al., 2017; Ferguson, 2016). Shadowing inherently involves encountering unforeseen ethical dilemmas, requiring “constant attention and continuous ethical decisions” (Czarniawska, 2007, p. 58). Issues arise regarding consent, not only for the individual shadowee but also for those interacting with participants (Bartkowiak-Theron & Sappey, 2012).

Informed consent is a known ethical challenge in shadowing (McDonald & Simpson, 2014). This study required obtaining consent from multiple stakeholders prior to conducting fieldwork. Securing consent from the occupational therapist participants I planned to shadow was a straightforward process. Securing consent from the principal of each school and from each teacher who worked with the therapist was more challenging. Similarly, obtaining consent from students’ guardians and assent from students, when necessary, proved more difficult. I relied heavily on the assistance of the therapists in this endeavour. While recognising the importance of informed consent, obtaining consent from so many different individuals was a significant logistical consideration.

Issues of informed consent encompassed individuals with whom the therapists interacted, but who were not considered research participants. Per ethics approval these individuals were excused from providing written consent; however, they still had to be provided sufficient information to agree to being observed. To address this I employed what Johnson (2014) calls “proportionate reason”, such that I explained who I was, the purpose of my presence, and obtained verbal consent prior to observations. It is testament to these individuals’ generosity that no-one objected to being observed.

Shadowing at schools meant that I observed students in a non-teaching capacity. On occasion it was necessary for the school community to provide negative consent for this. In these instances, the principal notified the community of my presence and requested carers to contact the school should they wish me to refrain from observing their child(ren). Again, no-one withheld their agreement. Although students were often inadvertently observed while shadowing therapists in classrooms, they were not considered participants, therefore did not need to consent to me being in their class. It is impossible to determine the impact my presence in their classroom had on their behaviours or actions. Therapists and teachers stated that my presence had no influence, as students were used to different adults coming in and out of their class.

Therapist consent was not considered a carte blanche, access all areas pass, and consent was verbally re-negotiated on an ongoing basis as situations changed throughout the day (Czarniawska, 2007). When a therapist moved from a classroom observation to a teacher meeting, I verified that they were happy for me to continue to follow them. My experiences echoed those of Shaw (2008), who noted that participants being shadowed run the risk of involuntarily disclosing information as they may erroneously assume that they are not being recorded in informal settings, such as the lunch room. I faced such experiences occasionally, being exposed to sensitive information and situations where it was unclear whether the revealed information was “off the record”. In such cases I explicitly sought consent to clarify whether information was to be kept private.

Anonymity is another ethical concern in the literature, primarily related to the small sample sizes characteristic of qualitative research. To some extent, shadowing is an invasion of the shadowee’s privacy (Arman et al., 2012). Although preserving participants’ anonymity was of concern to me, participants appeared indifferent to this. While shadowing several therapists within the same department they talked openly with each other about me and the study, making anonymity a challenge. Unlike Johnson’s (2014) study, no therapist expressed concern about being compared to their colleagues, and none inquired about fellow participants. This may have been due to robust team bonds, and possibly played in my favour, as those whom I had finished shadowing communicated their approval of me to the rest of the team, evidenced by therapists’ commenting that their colleagues had enjoyed the company and experience of being shadowed.

Building the Relationship

It is recommended that researchers establish rapport with participants (Glesne, 1989), as the researcher-participant relationship impacts the quality and quantity of data collected (Patton, 2015). The research relationship may also influence the validity and trustworthiness of the study (Patton, 2015). The literature stresses the importance of the research relationship, noting it develops as the researcher is immersed in the lives of research participants (Ferguson, 2016; Gill et al., 2014; Quinlan, 2008). Given the limited shadowing duration, time was a potential challenge to establishing the rapport and trust necessary for therapists to openly share their experiences. Nonetheless, it seemed the therapists and I quickly built rapport, which may be due to our professional histories. We shared a health background, experience of interprofessional collaborative practice, and good interpersonal skills. Similarly, we were familiar with shadowing-type relationships as students or supervisors during practicum. That the majority of therapists I shadowed were of the same gender, ethnicity, and education level, and of similar age, may have also contributed to the development of rapport (Russell, 2005).

Reflexivity and the “peculiar duo”

The unique relationship between shadower and shadowee is what Czarniawska (2007) labels a “peculiar duo”. In shadowing, conventional roles are transcended, as both parties must figure each other out while simultaneously making sense of what is going on together. Thus, a “conspiratorial bond” (Vasquez et al., 2012) is formed, as each party breaks out of their familiar world to form a temporary yet insightful “bubble” within the flow of activity. Inside this space new perspectives emerge, as both gain an insider’s view of each other’s world. Therapists’ informal contextual conversations, their answering of questions and exploring enquiries, provided me the opportunity to clarify understandings or perspectives (Quinlan, 2008), and facilitated therapists self-reflection on their actions (Bøe et al., 2017). These occasionally provoked different perspectives, as therapists noted “I hadn’t thought of it that way”. Similarly, listening to therapists’ reflections shifted my own thinking. As knowledge and understanding are co-created, shadowing is a collaborative methodology (Gill, 2011; Quinlan, 2008).

Reflexivity is about the researcher acknowledging their role in the research process, being aware of how prior experiences, assumptions and beliefs influence the process and the outcome of the study, and understanding how their background influences their interpretations (Watt, 2007). The shadowing literature suggests beginning shadowing in organisational contexts and with professionals with which researcher is familiar (Arman et al., 2012; Bøe et al., 2017). My first-hand knowledge of the discourse, procedures, and practices of the education and health systems provided me a particular way of understanding collaboration and the challenges to collaborative practice. While my history afforded me valuable insight and understanding of the context, it also imparted degrees of professional socialisation, engendering assumptions and ways of thinking that potentially influenced my expectations and interpretations of the data. Thus, while useful, tacit knowledge may have prevented me considering perspectives that someone naïve to the context might see (Quinlan, 2008).

The intention of a reflexive stance is to illuminate the researcher’s perspective and raise awareness of how this influences the research. To understand my perspectives prior to going into the field I engaged in reflexivity. I viewed teachers as overworked, with little knowledge of occupational therapy’s role in supporting inclusion, and with perhaps a hesitance towards having therapists in their classroom. I considered occupational therapists as desiring more involvement in the classroom with the ability to practise a more universal level of support. I also perceived them to have insufficient knowledge of school structures, curriculum, or how to work with large groups of students. Practising reflexivity to identify potential biases and understand how they shaped my perspectives was important to ensure I privileged participants’ opinions, experiences, and desires over my own.

Participants’ perspectives largely aligned with my own, however, sometimes they conflicted. Several teachers enthused about the support provided by an in-class therapist and expressed a longing for more of this type of support and co-planned and co-taught activities. Similarly, therapists enthused about times when they were welcomed into the classroom and delighted in co-planning universal services with the teacher. Practicing reflexivity ensured that the participants’ perspectives were the focus of field-notes, analysis and reporting.

Whilst favouring therapists’ perspectives, I was careful not to “take their side”, a potential hazard when shadowing as a result of the moment by moment sharing of experience. McDonald (2005) and Quinlan (2008) caution against the indiscriminate acceptance of a particular view, and encourage researchers to retain a critical distance. However, Gilliat-Ray (2011) interrogates the need for critical distance and the consequences of failing to maintain it, instead advocating a “sympathetic proximity”. Finding a balance between closeness and detachment, between “sympathetic proximity” and “critical distance”, requires self-conscious reflexivity. I examined my own positionality and priorities during the process to strive for such a balance (Maier & Monahan, 2009).

Insiders, Outsiders, and Blending in

Although helpful, my dual biography complicated my positioning within the study. As a nurse and teacher, possessing a knowledge of the inclusive education community and its members, I was an “insider” (Hellawell (2006). Like Gilliat-Ray (2011), who notes that ethnographic researchers can become “honorary members” of a particular group for a time, I was frequently introduced as an “honorary OT [Occupational Therapist]” by the occupational therapists. Conversely, as a researcher not currently employed within the education system or the schools I visited, I was also an “outsider”. Neither position is superior, both have advantages and disadvantages (Hammersley, 1993); yet, at the beginning of fieldwork I struggled to reconcile my “insider” professional identity, which participants knew and respected, with my “outsider” identity as a researcher.

Hellawell (2006) contends there is not one single insider-outsider continuum in fieldwork, but a series of parallel continuums. On some dimension the researcher has elements of “insiderness”, and on other dimensions elements of “outsiderness”, therefore, the researcher can simultaneously be both “insider” and “outsider”, or what Milligan (2016) calls an “inbetweener”. Arthur (2010, cited by Milligan, 2016) agrees, arguing against the “insider”-“outsider” fixed dichotomous entity, suggesting that a researcher’s identity and their status can shift dependent on the situation, and the social, political, and cultural values of a given context or moment. As the fieldwork progressed, I adopted this position of “inbetweener” or “knowledgeable outsider”, embracing the closeness and detachment that such a hybrid position allowed (Hughes et al., 2022).

The idea of negotiating proximity and distance is pertinent to the concept of “blending in” during shadowing. Avoiding drawing attention to oneself or to the act of shadowing is crucial (Czarniawska, 2018), and my “insider” role and similar demographic profile facilitated my blending into the research context (Bøe et al., 2017; Knutas, 2019). Quinlan (2008) terms this blending in, the situation of being present but unremarkable, “conspicuous invisibility”. While some authors suggest the researcher should be an unnoticed observer, the inherent intrusive and obtrusive-ness of shadowing challenges this perspective (Meunier & Vasquez, 2008). Like other observation techniques, shadowing goes beyond mere presence, and to be “conspicuously invisible” may require active engagement in the activities of the person being shadowed (Tulowitzki, 2019).

As a form of direct observation shadowing falls somewhere between participant and non-participant observation. Sirris et al. (2022) note that, compared to the more stationary nature of participant observation, shadowing is mobile, and that the focus is on the shadowee rather than the setting. Czarniawska (2018) locates shadowing as non-participatory, conceding that while all types of direct observation involve some kind of participation, shadowers tend not to proactively engage in the work, therefore do not perform concurrent action and observation. McDonald and Simpson (2014) concur, noting shadowing should not be considered participant observation, as it is unlikely that those conducting the shadowing possess the necessary skills required for the job and therefore participation is trivial or external.

Czarniawska (2007) argues against classifying field techniques noting when in the field if someone requests help you do not decline by saying “I’m not doing participant observation” (pp. 54–55). Thus, I found myself helping out on several occasions. Although some of the tasks I assisted with were trivial, carrying equipment or handing out resources, I also participated in more significant activities, such as supporting students in therapy groups and assisting in interventions. My experience of shadowing could therefore be considered participatory, as I possessed many of the requisite skills for the job. Mostly, however, I was what Wolcott (1973) calls a “participant as observer”, participating by my presence in the school as a teacher researcher, but permitted to retain the role of observer rather than being expected to perform as others do.

That my presence produced “observer effects” is undeniable. The Hawthorne effect is unavoidable in observational studies (Quinlan, 2008), and my presence inevitably changed the therapists’ behaviour in some way, if only by encouraging them to reflect on their actions through commentary. Nonetheless, within the busy school day, therapists had duties to perform that could not be suspended simply because I was present (Knutas, 2019), and therapists continued their scheduled meetings and work. McKechnie (2000) offers techniques for evaluating observer effects, including asking participants about the normalcy of their day and monitoring how often they inquire about the observer’s activities. During my observations no-one queried my journal entries, and when questioned, it was evident that, aside from minor inconveniences, my presence did not unduly influence therapists’ behaviour, inferring that participants continued to act predominantly automatically and habitually in my presence.

Contrary to Gill et al. (2014), who found that participants requested feedback on their performance, no therapist solicited my feedback. While fear of judgment is a consideration for shadowing studies (Bøe et al., 2017), participants appeared unconcerned that I might criticise their practice. Perhaps this was because they knew I was not an occupational therapist and that evaluating their practice was not my aim. Although most suggested they were initially nervous about my presence, all conceded that upon meeting and conversing any anxiety quickly dispelled. Teachers, however, regularly explained or justified their teaching practice to me, even though I clarified that evaluating teaching practice was not my objective. Unlike Gill et al. (2014), who found they were perceived in the role of the person they were shadowing, teachers did not position me as a therapist. Perhaps my presence was construed as an instructional round. As they knew I was also a teacher, their justification may have been a professional endeavour. Or, as I had no extended interaction with any specific teacher, it may simply have been that they did not have enough time to become accustomed to my presence (McDonald, 2005). The students whose classrooms I visited accepted my presence without question, assuming I was there to help, and approaching to ask questions or request assistance. In these situations, I followed the therapist’s lead. If therapists were observing, I declined to assist. If they were helping, I helped too.

Data Collection During Shadowing

Data gathering in shadowing is shaped by participants’ movements through time and space, as well as their professional role, and the researcher cannot control how or what type of data is collected (Bøe et al., 2017). Most therapists I shadowed were itinerant and serviced multiple schools. Although some had a base room at the school, the nature of their work meant they moved amongst and between meeting rooms, therapy rooms, staffrooms, and classrooms, with me following behind. The challenge of recording detailed fieldnotes while on the move is a known obstacle (Czarniawska, 2018), as is the time-consuming process of data collection and the volume of data generated (Ferguson, 2016; Quinlan, 2008). Different solutions have been proposed to improve efficiency, such as the use of video or audiotape to record (Bøe et al., 2017), however, this technology was not practical in the study context, and a notebook and pen sufficed. In retrospect I recommend note taking technology, such as smart pen and tablet, to ease the burden of data recording and transcription.

The process of note-taking demands focus. Detailed note-taking for hours is impossible (Ferguson, 2016; Knutas, 2019), and like others I initially vacillated between recording everything but worrying I missed the bigger picture, and producing summary notes with too little detail (Ferguson, 2016; Gill et al., 2014). As fieldwork progressed my recording skills improved and I began to fluently produce comprehensive data (Arman et al., 2012). As each day of shadowing is unique there is no perfect way to capture the field (Gill et al., 2014). I found using an observation protocol based on the collaboration literature to be helpful in ensuring I gathered the necessary data to answer my research questions. The use of predetermined categories is common in quantitative shadowing studies, however, the structured observation tool I employed was not a log of actions against time, rather a way to capture information about collaborative practice in the context. This helped delimit my “gaze” and prevent overwhelm (Arman et al., 2012).

The stamina and concentration required to record for hours, and for several days in a row, means that shadowing is mentally and physically draining. Like other authors, I spent evenings typing up field notes, recollecting, recording, and reflecting on the day’s data, meaning days were long (Ferguson, 2016; Gilliat-Ray, 2011). Some researchers speak of using “down time” in the field for such purposes (Gill et al., 2014), however, as the shadowed therapists had few breaks in their schedules I did not experience this respite. Sustained attention is required to ensure the quality of observations and consideration should be given to the physical and mental demands of shadowing when selecting it as an approach. Similarly deliberation regarding data management and analysis are important given the amount of empirical material generated (Arman et al., 2012).

The Emotional Work of Shadowing

Having participated in shadowing as experiential learning in my professional life I was familiar with the emotional labour it involved. Emotional reflexivity is a prerequisite for high quality ethnographic research (Gilbert, 2000). Emotional reflexivity is not only about the way one reflexively monitors and orders one’s emotions, but how emotion informs reflexivity (Burkitt, 2012). The emotional labour, required in gaining, securing, and maintaining access to the field, and in building and sustaining a relationship, is under-explained in the literature (Bergman Blix & Wettergren, 2015).

Intense emotional experiences can result in revelatory moments (Gilliat-Ray, 2011). Unlike Watts’ (2008) work with dying patients the shadowing I undertook was not in a sensitive context. Nonetheless, revelatory moments were generated when observing therapists’ work in supporting students with disabilities and/or at risk of exclusion. Watching therapists demonstrate sensitivity, respect, and compassion in addressing the diverse and complex needs of students, teachers, and families revealed their prioritisation of others’ needs above their own and the demands of the system. A conversation with Anna, a private practice therapist, is an example that provided me insight into therapists’ values, motivation, and philosophy of care. While discussing her regular participation in school meetings, Anna reported that she attended them to ensure she met the needs of her clients but received no financial reimbursement. Upon further enquiry she wryly said, “The system runs on the fumes of my goodwill”. Anna’s actions, the going above and beyond for no financial reward, demonstrated the embodiment of service, and her dedication and commitment.

Shadowing permitted me to observe the incredible emotional work of the therapists, which is core to their practice and vital to their ability to perform their duties, but which generally remained unseen, undocumented, and inadequately described in role descriptions. Work like assuming responsibility for building the relationships fundamental to collaborative practice. As Kirsty said, “It’s my job to build the trust in order to collaborate … I’m here to support the teacher, not the other way around. I’m very passionate that that’s part of our job, to build the collaboration”.

Being brought into the confidence of the therapists, hearing their (sometimes distressing) stories of discrimination against students with disabilities, challenged me to embrace the emotional and affective experience of fieldwork and consider how I placed myself as an “active agent” in the production of knowledge (Watts, 2008). Like other ethnographers I brought my life experiences, and issues linked to that history, to the research. Some observed situations made me ruminate on my own teaching practice and face perceived failures, my unsuccessful attempts to include students with disabilities in an education system that does not provide adequate support. As I listened to discussions about the supports required to include a particular student with disabilities in the mainstream, I reflected on my previous belief that including such a child would be too difficult. The student had several physical, intellectual and psychosocial disabilities, which, prior to this experience I would have considered too complex for a mainstream classroom, supposing the student would be better served in a special education environment. Speaking with the therapist challenged my assumption and confronted my prejudice regarding inclusion. The therapist emphasised that she was committed to doing everything possible to ensure the student had every opportunity to experience success in the mainstream environment where she was placed. For every challenge that was raised, the therapist and teacher worked to find a feasible solution, and when confronted with the failure of previously suggested solutions, the therapist did not lose heart nor portion blame, she simply stated “Ok, let’s try something else. What about….”. Had I only interviewed this therapist, I would have missed these powerful insights.

As researchers, we are professionals, yet we can be personally affected by the difficult situations we encounter. Similar to other shadowing authors (Gill et al., 2014), I occasionally left the field uncertain, wondering what the future held for the students I observed or whose stories I had learned. Watts (2008) notes that researchers working with populations at increased risk of harm may experience emotional distress, but that the experience itself does not inevitably cause distress. She cautions that researchers must be aware of the potential for distress and ensure adequate supports are in place to manage it. Managing emotion is personal. My biography, a middle-aged mother of four with a history in nursing, where management of emotional demands is a necessity for delivering successful healthcare, may have provided me with a degree of resilience or emotional protection. I pondered how I might have coped without such a history. Attentiveness to this aspect of shadowing, considering the way researchers, particularly young or novice ones, might be supported to cope with the emotional side effects of shadowing is important (Gill et al., 2014).

Less grave, but no less significant, is the emotional toll related to the intrusiveness of the method. It is the researcher’s burden to follow participants throughout the day, but it is the participants’ burden to be followed (Gill, 2011). Being with someone every minute of the day, constantly being “on” and “on show” and negotiating physical demands, such as food and bathroom breaks, has an emotional effect, and may be psychologically uncomfortable for both parties (Czarniawska, 2007). As shadowed participants are the focus of the research, they are vulnerable to being exposed in the data and findings, and the results may carry positive or negative consequences for them (Bartkowiak-Theron & Sappey, 2012). Like Ferguson (2016) I remained acutely aware of this and of my responsibility to participants. I worked hard to ensure their comfort with the shadowing experience and kept in mind the possible consequences of their exposure through the data.

More than other methods of qualitative research, shadowing opens participants lives to scrutiny, and thinking about findings and how participants might be presented can produce feelings of betrayal as the private is exposed to the public (Gill et al., 2014). In my shadowing relationships I made my objectives clear. I clarified my purpose in investigating how therapists and teachers work together, how they communicate and enact collaborative practice. To invite participants’ reflection on their practice and create a member check I also interviewed the therapists, who approved the verbatim transcripts. To provide a measure of agency in the research process I took the opportunity wherever possible to co-reflect about my observations with the therapists and was careful that “off the record” occurrences were kept so. Nonetheless, when findings potentially demonstrated therapists in a negative light, I felt disloyal, even when these findings were explained by factors outside therapists’ control. For example, considering therapists’ practice using the study’s conceptual framework of collaboration revealed a gap between therapists’ perceived level of collaboration and their observed collaborative practice. This could be explained by system barriers over which therapists had no control, however, reporting that therapists were not as collaborative, after observing first-hand how hard they strived to be, felt underhand.

Some authors write of the emotion on leaving the field after shadowing, particularly if the experience has been for an extended period of time (Gill et al., 2014). Although I enjoyed being in the field, appreciated the shared understanding and connection between myself and the therapists, and valued the participants’ time and commitment, I did not experience the parting as particularly difficult. The short duration likely cushioned the emotional impact of departure, and perhaps if I had shadowed for longer, I may have felt differently. Or possibly my professional history of frequently entering and leaving work sites provided a buffer. Or maybe, as Gill et al. (2014) state, leaving is ongoing. As my work in this area continues, I maintain a connection to the field.

Concluding Comments: Shadowing as a Unique Qualitative Method

This paper responds to McDonald’s (2005) appeal for research critiquing qualitative shadowing as a unique method by examining its use in education. To develop the method researchers must be explicit in their epistemological and methodological assumptions for choosing qualitative shadowing as a distinct approach, and describe the parameters employed, while positioning it against similar methods (Tulowitzki, 2019).

Qualitative shadowing is a valuable method that allows researchers access to the invisible and intimate spaces of participants’ work lives. The immersive nature of shadowing permits capture of the nuances of real-world interactions in context, providing rich data to understand participant perspectives. However, shadowing is not without challenges. As this study’s account demonstrates, accessing the field can be complicated and convoluted, requiring multiple approvals and consents, which must be frequently re-negotiated. Ongoing ethical challenges must be addressed, including securing informed consent, balancing observation with privacy, ensuring anonymity, and avoiding researcher bias. The unique researcher-participant relationship distinguishes shadowing from other observation methods, offering a different type of knowledge in educational research. As the field is co-created, a distinct partnership between the shadower and shadowee must be cultivated, and reflexivity on the positionality and identity of the researcher is essential. Data collection challenges are common and are related to the difficulties of capturing appropriate detail amidst diverse activities and the large volume of data produced. Navigating large datasets and data richness can lead to complexities in analysis. Finally, as shadowing is embodied, relational, and emotional work, it can be exhausting for both the researcher and participants. Adequate supports are required to manage the emotional toll.

This paper adds to the literature by providing insights into the distinctiveness of qualitative shadowing as a unique method. It highlights the particularities, as gleaned from the literature, situating these idiosyncrasies within my own experiences of shadowing occupational therapists in Australian primary schools. The paper promotes the method within the field of education, arguing that it offers an approach to the field that other qualitative methods do not. Insights provided can support future shadowing researchers and inform future studies.