• No results found

CHAPTER 3: A CONSTRUCTIVIST GROUNDED-THEORY APPROACH TO

3.5 Research Techniques Used in this Research

3.5.6 Early Theorizing and Theoretical Sampling

Formal theory development is typically done during the final stages of CGT research, but theorizing about phenomena in the data is helpful throughout the research process (Charmaz, 2014; Bryant, 2017). A tool often employed in theorizing is

“theoretical sampling,” which Charmaz (2014) describes as “going back into the

empirical world” to collect new data to clarify properties of developing theories (p. 192). Typically theoretical sampling is done near the completion of a grounded-theory project, as a way of “saturating” categories that have emerged (Bryant, 2017, p. 251). However, in some grounded-theory research a type of theoretical sampling can be conducted within the data-collection period—particularly when time allows for analysis of early data to influence collection of later data (Urquhart, 2012).

Due to the affordances of time and data quality, I was able to follow this iterative approach. The scheduling of data collection into two periods separated by nearly four months required me to delay formal focused coding until all data had been collected and

initially coded. However, this delay also gave me an opportunity to explore in detail the 23 transcripts I had collected at Sites 1 and 2. Comparing data with data, writing memos and discussing my observations with my supervisors and colleagues, I identified salient phenomena within these initial data that led to development of initial theories and categories—ideas that I extended through a form of theoretical sampling in my data collection at Sites 3 and 4.

An example of this theoretical sampling is in my evolving conception of

individual identity within IBC organizational contexts. At the outset of this research I was directly focusing on lecturers’ behavioral intentions, asking for example how they felt about representing their universities. In my research at Sites 1 and 2 these questions yielded data in which participants detailed their experiences in activities such as IBC marketing events. In comparing these data across cases I noticed that participants repeatedly described boundaries they placed on their own behavior which seemed to be guided by loyalties beyond their institutional identities: a morally-imbued set of personal responsibilities that influenced how they carried out their work. I realized that

participants were outlining for me a model of layered identity prioritization. The extract in Figure 3 below from an email to my supervisors captures my early theorizing about this topic:

Figure 3: Extract from Email to Supervisors on May 11, 2018

I've noticed that in interviews when we talk about participating in marketing activities (e.g., open days) many participants taking pains to justify (perhaps to me as an assumed fellow academic?) their reasons for participating and where they draw the line in terms of what they will and won't do in these encounters. One participant made a big deal about stressing that he does this only because he truly believes in the university's mission. Another emphasised that he's not a "snake oil salesman" and talked about subverting the process by recommending the home campus rather than the IBC when students seemed a better fit for this... These people are constructing for themselves an identity that does justice to their values (stemming perhaps from the group habitus of academia) while allowing them to fulfill their job responsibilities.

This extract showcases the theory I was beginning to develop at the time about what was influencing these lecturers’ behaviors: the idea that they were constructing an occupational identity separate from their organization-related identity, with the former mediating their behavior as part of the latter. I also realized in my initial research that

participants seemed to enjoy talking about identity directly, finding it a helpful shorthand for describing their orientations. Thus, in the updated version of my interview guide for Sites 3 and 4, I included a new question that asked them directly about their professional identities, as seen in Figure 4 below.

Figure 4: Question Added to Interview Guide for Data Collection at Sites 3 and 4

When you think about your professional identity—your sense of yourself in your profession—do you feel like PU [Parent University] is part of that identity? And if so, do you identify more with PU local or PU global?

PROBE FOR: professional identity, identification with PU, distinction between professional and institutional identity.

FOLLOW-UP QUESTIONS:

… so is it academic first, then PU local, then PU global?

Has your professional identity changed since you started this job?

This new interview question for Sites 3 and 4 allowed me to glean further data that eventually helped me to saturate properties of the category “Prioritizing professional identity over institutional identity.”

Another example of mid-research theoretical adjustment was around the concept of “spoonfeeding,” a term that participants at Site 2 introduced to refer to the educational practice of oversimplifying course content and delivering it to students in overtly

accessible ways, such as providing sample answers to exam questions. This phenomenon intrigued me in its demonstration of how course delivery could differ while materials were ostensibly globally aligned. I began to theorize that spoonfeeding related to how lecturers envisioned their IBCs: they were re-imagining their IBCs as aligned with this practice and setting related expectations with students. This marked the beginning of my thinking of IBC lecturers as co-constructing their IBCs. In data collection at Sites 3 and 4, I adopted IBC construction as an additional sensitizing topic, probing for related insights in various parts of the dialogue.

These examples of my early theorizing around professional identity and spoonfeeding demonstrate the ways in which insights gained in analysis of early data influenced my later data collection. The analytical period bisecting my two stages of data collection helped me to refine my thinking and my interview questions to maximize the

effectiveness of data collection at Sites 3 and 4. By the completion of these interviews, I had a substantial amount of usable data, with a good portion of this data already tailored to further illuminate emerging phenomena identified in mid-research analysis. Following initial coding of the full set of data, I began the next official stage of CGT research: focused coding.