Introduction

In this chapter, we recount how, as a team, we came to use a narrative methodology and combine it with a longitudinal approach. A key theme in the chapter is our conviction to make the purposes for using a narrative methodology clear and the procedures we employed transparent. We first describe the gradual changes in the research program over 10 years. Our goal is to contextualize our practice of narrative within the growing use of it as a research methodology in the social sciences. Further, we explain why and how we link identity and narrative, since not all who use identity as a conceptual tool use the narrative, nor do all those using the narrative link it to identity. We then explore our relationships with the participants, including efforts to be reciprocal: describing how we endeavored to bring an ethic of care to our work with them (Harrison, MacGibbon, & Morton, 2001). We conclude with what we have learned about the challenges of using a narrative approach, based on our experience of using it longitudinally.

The Team

When we submitted the initial funding proposal for this research in 2005, the two of us had known each other and worked together for more than 10 years. We first worked together at McGill, and when Cheryl moved to Simon Fraser University, we continued our collaboration through a grant about faculty development. At a time when technologies were not as supportive of long-distance collaborations as they are now, we developed a sense of shared values and a good working relationship and established a productive way of working together at a distance. For instance, we had already begun using North American conferences as places to meet face-to-face, spend time together, and plan our work. Without this previous experience, it is unlikely that we would have been able to undertake this research. Among the values we shared were: a commitment to ensure the results of our research were directly relevant to practice, a belief in the value of research collaboration to produce more robust results, and a desire to integrate students into our research teams.

When the research began in 2006, we had two teams, one in Vancouver (Cheryl, Gregory Hum and later still, Esma Emmioglu) and the other in Montreal (Lynn, Marian Jazvac-Martek, Shuhua Chen, and Allison Gonsalves). Shortly afterward, Lynn began to work at Oxford as well, returning to McGill on a regular basis. So, this provided an opportunity to start a third team (Lynn, Nick Hopwood, one or two research assistants, Gill Turner, and later Mahima Mitra) collecting parallel data. However, given the constraints of European Union data security regulations, the data sets were not pooled. We did however arrange some collaborative work through joint but separate analyses. We also added European conferences to our face-to-face meetings, and Greg spent an extended period of time with the Oxford team.

We used Skype on a regular basis both for team meetings and for meetings between the two of us. In team meetings, we caught up on team members’ own experiences; discussed the year’s data collection plan (setting rough dates when different types of data would need to be collected); developed, reviewed, and occasionally modified data collection instruments; established and checked in on protocols for data storage and file names; established agreement on coding definitions and procedures; discussed presentation and publication plans (including copublishing). While these meetings permitted us to establish guidelines and make research progress, we also tried to meet face-to-face once a year at a conference when, besides presenting, the team would spend a day together either at the beginning or the end of the conference.

During the ten-year period, different members of the team completed their master’s and PhDs and a postdoc fellowship, sometimes drawing on the longitudinal data alongside other data they collected and sometimes independently researching topics related to the research. You can find out about some of their work at https://mcgill.ca/doc-work/.

Lastly, while the bulk of our research has been longitudinal, we also conducted a number of smaller studies which involved one-time only data collection. These smaller studies generally emerged out of and were directly related to the findings in the longitudinal study (and were referenced in Part III). When we refer to the studies emerging out of the data, we do not mean that such studies were necessarily based on a theme we saw in the data, though this could be the case, as in the study of becoming a PI (McAlpine, Turner, Saunders, & Wilson, 2016). More often, we realized that we lacked a perspective on an important aspect of early career experience and decided to carry out a separate study. Examples include learning to be a supervisor (Amundsen & McAlpine, 2009), PhD students who had particularly difficult journeys (McAlpine, Paulson, Gonsalves, & Jazvac-Martek, 2012), or early career researcher perceptions and experience of policy (Ashwin, Deem, & McAlpine, 2015). In other words, we tried to be vigilant about what we were not learning from the data as well as what we were learning.

These studies usually involved the team members, but not always. Figure 1 provides a chronological overview of the links between some of the small studies and the larger longitudinal study as well as the integration of the data collected by the Canadian and UK teams. In this diagram you can see that we first began with social scientists focusing largely on Canadian doctoral students, then with social science PhD students, and post-PhD researchers in the UK. The science aspect of our program is of a shorter duration so has fewer independent studies. In the sciences, as with the social sciences, we began in Canada looking at a range of early career researcher roles and then started a related study of PhD students only in the UK.

Fig. 1
figure 1

Chronological representation of linked longitudinal and one-off studies in Canada and the UK

Full size image

The Research Process

One of the things we could not have imagined when we began to visualize this research in the 2005 funding application was not only a deepening understanding of the research approach, but also a growing understanding of ourselves as researchers. The opportunity to engage in the same research for an extended period of time using a little reported approach pushed us to reflect carefully on what we were doing and to explain it to ourselves and others in a concrete, transparent fashion. We believe there was an increasing maturity in our thinking about how we conducted the research, resulting in a clearer and sharpened perspective throughout the research process. In this chapter, we focus on these three aspects: clarifying the methodology, negotiating interest in the research and later access to the research findings, and ethical practices.

Narrative and Longitudinal Research: Clarifying the Methodology

We began this qualitative research program in 2006 as an exploration of the seemingly intractable, well-reported problem of low doctoral completion. We thought to capture day-to-day doctoral experience to see if this might provide more insight than previous research based on large survey studies or one-time interview studies had. The approach we used was weekly activity logs (see Chap. 10 for more details). In the beginning, we did not conceive of the research as longitudinal and only thought of this potential when we began to prepare to interview participants after they had completed a number of weekly logs. So, while there was little in the literature to guide us as to how to undertake longitudinal research, we decided to see if participants were willing to continue and asked them this at the end of what became the first interview. A number commented on their participation in the research as being personally supportive and most were willing to continue into the second year. Further, as we read through the data, we became sensitized to the incredible variation in individual experience despite, for instance, some individuals even being in the same program. While much qualitative analysis synthesizes multiple individual experiences, we wondered if we should maintain a focus on the individual. We decided ‘yes’ but without having thoroughly grappled with the challenges of doing so. At the same time, we still wanted to look across individuals for shared themes.

Naturally, the decision to continue data collection meant we found ourselves tracking people over extended periods of time, which required a major rethinking of how to negotiate our relationship with them. The additional data also required rethinking how to merge and analyze multiple data sources, not to mention how to manage an ever-growing and large dataset. While there was minimal attention in the literature on how to conduct longitudinal research, we drew on the few studies we found to guide us. Probably the most useful, though the research did not focus on higher education, were the reports by Thomson and Holland (2003) and Thomson (2007).Footnote 1

Over time, we developed a robust process that involved repetitions of data collection and analysis. See Fig. 2 which provides an overview of the annual data collection cycle we decided on (more details in Chap. 10), the ways in which we summarized independently the experiences of each participant, and how in analyzing and reporting we kept accounts of individuals separate while also noting similarities and differences between individuals (more details in Chap. 11). Our goal was to preserve a focus on the individual over time, but also look for common patterns or themes across individuals.

Fig. 2
figure 2

Data collection and analysis that maintained a focus on the individual

Full size image

We thought of the methodology broadly as emergent and thematic, finding patterns through successive readings of the multiple texts provided by research participants. We clearly saw in the stories that participants were recounting how they were protagonists trying to take action in relation to emerging events. Partly influenced by earlier work that Lynn had done using narrative as a methodological approach (McAlpine, 2016 Footnote 2), we began to explore the potential of narrative as a useful methodology. We found it provided a clear mechanism for analysis, and met our desire to provide action-based results.

We are not alone, of course, in coming to see the value of narrative as an interpretive approach. It has, in fact, been used in, for instance, sociology, organizational studies, gender studies, and educational research. Narratives involve recounting—accounting for—how individuals make sense of events and actions in their lives with themselves as the agents of their lives. In other words, narratives provide accounts of how the narrator, the protagonist, engages with and responds to experiences in ways that demonstrate efforts to achieve intentions despite difficulties. Narratives make connections between events, represent the passage of time, and show the intentions of individuals (Coulter & Smith, 2009). Telling stories about our lives is a common feature of daily interaction, and whom we tell them to, as well as when and where influences what is told. So a story told to, for instance, an interviewer, is thus historically, socially, and physically located in a particular space-moment potentially linking the past, present, and future—and a different account would emerge if the ‘story’ were told later in the week to a close friend.

With this as background, we now turn to the different ways in which social scientists conceive of the narrative as a methodological approach. This is because those who use narrative do not necessarily have the same practices we do in terms of how data are conceived, collected, analyzed, and reported. It is generally agreed there are three stances in how a narrative methodology is understood and used (Elliott, 2005; Pinnegar & Daynes, 2007). The first reflects a sociocultural perspective and focuses on how broad cultural narratives influence individual experience, for example, Ylijoki (2001). Another starts with a naturalist perspective and seeks rich descriptions of the influence of significant personal issues on life decisions and actions, for instance, Cumming (2009). The third takes a literary perspective and focuses the analytic lens on the discourse that individuals use to describe their experiences; this stance, seen in Hopwood and Paulson (2012), is often blended with one of the other two. The naturalist stance best represents the one underlying our research.

Of course, just as with any research design, a narrative lens can serve as the primary methodology underpinning the design. It can also serve as one approach combined with others, as in mixed method designs.Footnote 3 In our case, the narrative served as our primary lens and was integrated throughout the design. In Chap. 10, we describe the methods of data collection we used and why we used them. Chapter 11 takes up our processes of data analysis and reporting. In short, participant narratives compose our dataset. We engage in narrative analysis using a naturalist approach followed by thematic analysis across cases. We then use both narrative cameos and cross-case themes in our reporting.

Conceptually Situating Narrative with Identity Development

As we continued to engage with a narrative methodology, we were led to think more deeply about how our epistemological stance and our interest in identity and development aligned with our approach to the narrative.

We describe our stance as that of critical realism (Archer, 2003), the essence of which is that individuals are agents as they live in, respond to and create cultural, social, and physical spaces, and engage in social interactions and activities. And, their decisions and actions are situated within their own physical beings incorporating, for instance, gender, age, illness, disability—what Billett (2009) refers to as the ‘brute.’ In other words, a collection of interacting elements offer affordances and constraints for individual thought and action as well as developing a sense of identity.

We believe a naturalist narrative methodological stance aligns well with this epistemology since we were interested in documenting and understanding the distinctiveness of each individual’s trajectory. In other words, each participant narrative represents an oral or textual snapshot on the identity under construction. When individuals tell a story about their lives, whether to themselves or others, that story has the power to influence what they see themselves becoming and how others see them. Each narrative provides the teller with a robust way of locating himself or herself, the agent, in the story along with feelings, motivation, and values. Further, the telling of the story offers the potential to reflect on past experiences for learning. In other words, in our work, we make little distinction between thinking, learning, and the formation of identity (Billett & Somerville, 2004).

Thus, we see ourselves tracking individuals’ identity development, an identity that incorporates the permanence of an individual’s perception of a unique identity combined with a sense of personal change through time (Riessman, 2008). The accounts or narratives that individuals provide to us at different points in time serve as representations of that developing identity. In examining the multiple accounts, we are seeking to understand how individuals experience life and through their actions, conceive their degrees of freedom, and exercise agency in ways that include efforts to avoid, challenge, or resist perceived practices and policies.

In summary, using narrative to understand identity construction focuses attention on the individual rather than the group. Further, narrative addresses what is often overlooked in other methodological approaches to early career researcher experience—the individual’s sense of agency and intention. Linking identity construction to longitudinal narrative research provides a robust basis for using participant data to understand identity as constructed through time.

Making the methodological decisions we did reminded us of the challenges to be mindful of in taking a naturalist narrative stance with a focus on individual identity development. Our focus is on the close-to-home experience rather than the abstract and relatively decontextualized larger economic and public structures (Billett & Somerville, 2004). Thus, given that each narrative is told in a particular time and space, much is left out (and we cannot know what that is), but this makes it difficult to move from the micro level to the structural level (Walker, 2001). Further, Taylor (2008) cautions that those providing narratives seek sense-making about their lives rather than a sense of indeterminacy or complexity. In other words, a narrative approach enables us to understand through individuals’ stories their experiences of the physical realities, such as illness, age, as well as social realities, for example, who the participants work with, who they live with. However, these same stories may not provide insight into the broader social structures or realities in which individuals are embedded, for instance, the job market or the actual possibility of being awarded a grant. We recognize this as a serious concern, and in Chap. 4 described an approach to future data collection and analysis that begins to address this issue. In our research since we came to this broader view in the past few years, the best we have been able to do is draw on national quantitative figures about academic and non-academic positions as well as public records that identify the larger structural context that forms the reality that our participants faced.

Negotiating Interest in and Access to the Research

We shift now to engaging others in the research, addressing in particular the process of recruiting participants and the feeding back of results to different stakeholders.

Recruiting

In recruiting research participants, our overall message regardless of who we were communicating with (we sometimes had to approach intermediaries) was to highlight the potential value of the research findings for future doctoral students and other early career researchers. With possible participants, we also noted the potential personal value of participating. As the research progressed and participants confirmed this personal value, we spoke of this effect with more authority. We also used different strategies to invite participation in relation to the institutional role we were seeking to recruit and the policies concerning research ethics in each institution. Consequently, when recruiting newly hired research-teaching academics, we tended to approach deans of faculties and ask if they would be prepared to distribute an email that we had written inviting participation and promising confidentiality. With doctoral students, we approached chairs or heads of departments making a similar request. If there was a strong student organization, we asked officers to distribute information through their mail list. Lastly, with post-PhD researchers, we approached the institutional postdoc organization first since departments did not always have complete lists of those in this role.Footnote 4

We have often been asked how we were able to recruit so many participants. We can only speculate, but we are quite sure that the topic itself was attractive, that is, struck a personal chord with many. It was also likely attractive to a particular subset, those who were interested in their own development and learning. At the same time, we had one instance in which we were not at all successful in recruitment using the strategies described above. We had intended in the first grant to collect data from PhD social scientists and humanists (English and History) and we were singularly unsuccessful in recruiting humanists—though we were able to recruit a small number of research-teaching humanists, who helped us find some PhD students for a one-off study.

As reported earlier, we had not intended a longitudinal study. But as year one ended, we became intrigued by the potential to follow people over time. This, of course, depended on whether participants were willing to continue to participate. So, at the end of the interview related to the completed weekly activity logs, we included a question regarding interest in continuing, and the study grew from there.

Reporting Back

We had originally intended to report the pedagogical findings back to faculties and departments and this was, in some respects, not straightforward. First of all, the longitudinal approach meant it was some time before we began to have useful results (and to publish). As a result, the institutional leaders who had originally been approached about the research were not necessarily still in the same leadership roles. We found it easiest to report back to units responsible for doctoral education in each institution, and provided briefings or short reports to Deans of Graduate Education, for example. We were also able in two instances to be involved in the creation of institutional supervision websites, each of which provided a venue for sharing our research findings (as well as other resources) in ways that could be useful to doctoral students and those who worked with them. As the first grant ended and the work became more widely known, we focused more on reporting to external groups. So, we accepted any university invitations we received to give workshops, both related to policy and practice. We also negotiated a contract which resulted in a book that highlighted the pedagogical implications of our research for those supporting doctoral students (McAlpine & Amundsen, 2011). Finally, we approached interinstitutional organizations which had an interest in doctoral education and in one case were able to post the policy and pedagogical implications of our research on the organization’s website.

These dissemination efforts continued as the second grant got under way. We continued to accept invitations to hold workshops and were invited to submit a proposal for a book specifically written for early career researchers, so they could learn from the experiences of our research participants (McAlpine & Amundsen, 2016). We negotiated a companion book for those interested in understanding how we carried out the research (this book). We have also more recently begun to create noninstitutional online resources (still under construction), which will provide information to those considering a range of roles whether in or out of the academy.

Working with Participants: Negotiating an Ethic of Care

As researchers, we are the ones who seek out relationships with participants since we wish to learn from their experiences. We are therefore given a certain power and expertise by participants, in that ‘we listen people into speech’ (Josselson, 2007, p. 547) or into writing. In other words we, as researchers, are highly engaged partners in the narratives-under-construction that we collect (Sfard & Prusak, 2005). Nevertheless, we cannot know the unstated expectations of the interviewee nor how the relationship will develop.

We have come to realize that performing longitudinal narrative research heightens awareness of ethical decision-making and practices in relation to those who provide us with insight into their experiences. Over time, we develop a privileged intimate knowledge of an individual—as one participant said ‘you know my life.’

Taking a narrative stance as we view it means endeavoring to document and understand an individual’s experiences from his or her own perspective. Further, we recognize that the nature of the relationships we have with participants influences the stories they tell (Juzwik, 2006). More importantly, as Josselson (2007) has noted, narrative research is ‘fraught with [the] dilemmas of choice that attend all ethics in all relationships’ (p. 537); so we face dilemmas which have no ‘right’ answer. The following scenarios provide just a sampling of the kinds of ethical issues that emerged during our research.

  1. 1.

    A participant in his logs reported serious challenges over several logs, particularly institutional pressure to finish, alongside family and financial concerns; he expressed feelings of depression.

  2. 2.

    A participant endeavored to engage us in a closer relationship than that of research participant.

  3. 3.

    We wanted to create a caring long-term relationship with each participant.

  4. 4.

    We wanted to ensure personal data did not reveal participant identities in publications.

  5. 5.

    A participant contacted us sometime after he had finished participating saying he was concerned that his identity might be revealed and wanted the data destroyed.

  6. 6.

    We wanted to create a more reciprocal relationship between us and the participants.

In the rest of this part, we describe the approach we took to working with participants and in the process describe how we addressed these ethical scenarios.

Globally, formal ethical procedures have increasingly come under institutional oversight and control, so we do not explore this aspect of ethical practice, rather just note that in the different studies we consistently sought and received institutional approval, gathered and safely stored written consent, and established secure data storage.Footnote 5 Instead, we focus here on what we found to be particularly profound in our learning from doing longitudinal narrative research, the nature of the relationships in which we engaged over many years. We have come to conceptualize the stance we endeavored to have with participants as an ethic of care.

Gilligan’s seminal work (1982) brought to the fore a set of principles distinct from the predominant Western logic of justice for decision-making and action. Based on her research comparing the responses of boys and girls to a moral dilemma, she proposed what she called an ethic of care as a viable principle for decision-making and action. Such an ethic is based on connectivity or relationships through time as the principle underlying action.Footnote 6 Like Gilligan, Tronto (1995) has argued the need to move from an Aristotelian and Kantian view of rational ethical decision-making to seeing people as enmeshed in relationships of care. While describing the need for care as universal, Gilligan noted that this need and any response to it must be understood and practiced in culturally specific ways.

While earlier conceptualizations of care and caring focused on the individual, Tronto (1995) has also argued for care to be understood not solely as a private or parochial principle and undertaking, but rather as something that can be taken up by groups and organizations. However, if caring involves more than one person (as in our team-based approach), caring may be more challenging; still, it can perhaps facilitate a greater range of possible responses (Tronto, 1995). We found this to be the case. Lastly, the standard of care and equity of care rest on judgments that assess needs in personal, social, and political contexts (Kardon, 2005). He argues the quality of these judgments be assessed by considering whether they are similar to the diligence and the best judgment as practiced by reputable professionals in similar situations. Working in a team helped us to work toward such reasoned judgments as we sought to honor and protect the participants while maintaining standards of responsible scholarship.

By seeking to practice an ethic of care, our intent was not to emotionally empathize since this could make us unaware of other perspectives and privilege relationships with some participants over others. Rather, we sought cognitive empathy, identifying and understanding others’ emotions and perspectives and, in this way, ensuring that we were mindful of their well-being, while maintaining a scholarly stance. What did this mean in practice? Scenarios 1 and 2 provide examples of how we endeavored to use cognitive empathy to support well-being.

Scenario 1: We had agreed as a team procedure that whenever a weekly log was received, it was immediately read and some email response, as simple as thanks, was made. The reason for this is that it would be easy to just store the log and only look at it before the interview. But the log, completed away from the interviewer, is different from the interview where the researcher is as much a participant as the interviewee so is fully aware of the participant’s concerns. Further, since individuals were sometimes revealing challenging issues in the logs, the team agreed that if any one of them noted ongoing reports of challenges that seemed to be leading to a strong negative emotional response, the issue was to be raised in the team. In this instance, one of the teams reported that the participant who was trying to complete his PhD at a distance, and had family and funding issues, had found himself caught up in institutional regulations regarding completion. He reported this constellation of factors as extremely troubling. So, we discussed as a team what we might do. We could not intercede in his relationship with the university in any way. Instead, we agreed to generate a list of all the institutional resources in the participant’s university that he could draw on and included these in a carefully drafted email. Our hope was that our effort to demonstrate care and the concrete nature of the resources on offer would help him to develop more resilience as well as the ‘head-space’ to finish. He responded very positively to the email, sought some help, and completed his degree.

Scenario 2: While we wanted to maintain a caring relationship, we were clear that this needed to be within the boundaries of the research project. In one instance, a participant, a post-PhD researcher, was having difficulties finding a position. He was no longer providing data in a consistent fashion, but started emailing about the difficulties of his situation, asking for our opinions. As Josselson (2007) has noted, participants may seek a continuation of the relationship beyond the requirements of the research due to the attention given to the participant by the researcher. What we did was to politely but clearly restate the role we played as researchers. Note the difference between this and the previous scenario, where the student did not seek help but we developed concern based on the data.

Scenario 3: This presents another aspect of ethical and caring research practices. It is linked to the ways in which we tried to ensure an ongoing caring relationship, within the parameters of the research program, between each participant and members of the research team. To begin with, we undertook to match the role research participants had to play with a team member in a similar role. As best as we could, we matched doctoral student participant with doctoral student team member, post-PhD researcher participant with post-PhD research team member, and so on. Further, the hope was that the same relationship would be maintained throughout the research. Of course, team members completed their degrees and left. In such cases, we would plan ahead, considering who might take over the responsibility for the participant relationship. This team member, after consulting with the team member who was leaving, read through all the previous data from the participant and also listened to the audio recordings of the interviews to get a sense of the ways in which the participant communicated. Then, through email, the present team member contact introduced the new contact to the participant, and when possible there was an email exchange among the three.

Scenario 4: This relates to a further aspect of ethical and caring practices, namely, confidentiality and anonymity. Early on, we began developing protocols to preserve these. Anonymity involves removing or obscuring the names of participants and research sites in order to avoid revealing information that might lead to participants being identified. However, anonymity does not ensure confidentiality, the management of private information, that if revealed could create prejudice against participants (Tilley & Woodthorpe, 2011). The reason anonymizing does not of itself ensure confidentiality is that, for instance, an anonymized account might be recognizable to someone who knows the participant and knows of his or her involvement in the research. A longitudinal research design makes this even more possible.

While we had attended to these features of ethical and caring research from the beginning, the issues became more complex as we followed individuals for an extended period of time, and we saw the greater potential for harm. At the beginning, all participants chose pseudonyms and the file names given to the data they provided always used these pseudonyms. (We kept a separate file of matching names and pseudonyms.) We also only used pseudonyms in any group discussion of the data. As well, participants were offered the opportunity to review and delete anything of concern in their interviews (there was little use of this opportunity). Data were secured behind an institutional firewall which was also password protected.

However, as a team, we decided not to delete information such as location from the stored data since individuals moved and we needed to keep track of where they were and what their specialism was. However, as transcripts were read for the first time, personal information was highlighted in yellow as a reminder to us of the need to remove or change this information in any public reporting. This is the crux of Scenario 3: ensuring we maintain our scholarly integrity as regards making sense of the data while at the same time ensuring we do not reveal individuals’ identities. We developed quite specific protocols around this in our reporting, for instance, not referring to son or daughter but to child or children, not wife or husband but partner, and region or size of cities rather than names of cities.

When we decided at the end of year one to continue the research, we reviewed the consent form to ensure we would still be compliant and sought verbal consent from participants to continue. At this point, individuals had a real understanding of what the research actually involved and were in a better position to decide to continue or opt out. Interestingly, the most loss in terms of participation occurred during the first 15 months. Those who continued that long tended to remain until the study ended.

Of course, consent is always provisional, and even now, individuals could come to us and decide to no longer participate, just as we experienced in Scenario 5. The participant, who had experienced some difficulties institutionally during his participation, was no longer providing data. However, he contacted us concerned that it might conceivably be possible for someone to recognize him so asked us to destroy his data. We did so immediately, but also pointed out to him that while his data would not be drawn on in the future, we could not remove anonymized reference to him from the material that had already been published. Josselson (2007) has noted that concern about identification can be particularly acute in small communities and, in some respects, academic specialisms are often quite small. We need to be prepared to rescind any material that participants may request even now that the research program has ended.

Scenario 6 is related to our desire to create some reciprocity in the relationship, to give back something of value to the participants since they were so generous of their time and personal experience. There was little discussion of this in the literature, so we brainstormed as a team and generated two strategies to create more reciprocity. The first was to send occasional emails (two to three times a year) in which we summarized the practical relevance of a research article (authored by other researchers) that was related to the participant’s role, and also provided a link to the article in case the individual wished to read it. A post-PhD researcher would receive a summary of a different article from that of a doctoral student. We also annually sent a report about our research in which we described how we had been able to use the information they gave us, for instance, in workshops or on supervision websites, and a few links to publications emerging from the research. We do not expect that everyone read these (though we had unsolicited reports that some did), but we felt that we created some reciprocity in the relationship and were also transparent about the uses to which we put their information.

Challenges of Using a Narrative Longitudinal Team Approach

We are mindful that since we used an inductive approach, the methodology that we used and the data that emerged strongly influenced the way we conceive of identity development. For instance, our choice of a naturalist narrative approach meant we were primed to understand individual experience. This approach also meant that the data collection process we designed was perceived as personally meaningful and useful by research participants. The longitudinal aspect gave us the freedom to collect specific information at different points to flesh out our interpretations and made it easy to take a learning and change perspective.

Further, working in teams meant that we could track a relatively large number of people as well as match by role the participant and the researcher. Further, the geographical distribution of the team meant we could start collecting data in different places and establish face-to-face relationships with participants before they moved on to other roles, for example, doctoral student relocating to take a post-PhD researcher position. At the same time, having a relatively large number of researchers in the teams meant we had to work harder to develop and maintain a sense of shared commitment and common procedures, given the geographical distribution.

Choosing to do longitudinal research also brought with it the sense that we were making it up as we went. For instance, it was unclear how many participants would remain in the long term, how data collection would go and what might change over time. For instance, we had to find ways to collect data consistently and virtually as individuals became distributed around the globe.

Such an approach demands of the researcher a constant vigilance as to the emerging data, the standing of the participants, the management of ever larger data sets organized in such a way as to ensure data security, participant confidentiality, and ease of access. A positive effect of this uncertainty was that it opened us up to being exploratory and creative, an effect we explore in greater detail in the next two chapters.

Given our desire to report back to different stakeholders in meaningful ways, we viewed the research in some respects as developmental and action-oriented. So, we were pleased to find the kinds of reports we could generate from our design were identified as particularly meaningful by not just the research participants, but also other early career researchers and readers of our papers and reports. More specifically, we have been able to use the narratives of people’s identity-trajectories and what we learned about their lives in workshops for a range of stakeholders, as well as in online resources. We invested heavily in sustaining an ethic of care in the design of our research, for instance, by ensuring benefits for the participants and other stakeholders. We hope we have succeeded to some extent, but recognize there are always issues of power and representation of others that remain in any social science research.