Narrative: In undertaking this research, we draw on a narrative tradition with a focus on biographical stories. Our work is unusual in this tradition in not looking at just one or two cases, but across a number of individuals as have Thomson & Holland (2003). Narratives make connections between events, show the influence of the passage of time in carrying the action forward, and demonstrate the goals and intentions of individuals (Coulter & Smith 2009). Narrative methodology has been used broadly in the social sciences (Elliott 2005) including to study identity in a range of fields, e.g., Ricoeur 1991; Sondergaard 2005. The underlying premise is that narratives, whether oral or textual accounts of life experiences, represent constructions of identity (Sfard & Prusak, 2005; Riessman, 2008). We view data collection as a co-construction between researcher and participant and the results of analysis as provisional (Coulter & Smith 2009; Sfard and Prusak, 2005). Our goal in analysis is to ensure the fidelity, coherence, plausibility, and usefulness of our constructions and interpretations of individuals’ narrative identities (Coulter & Smith 2009; Creswell, 2013).
The narrative approach we take is distinct from many others as it takes a longitudinal perspective, consistent with our belief that "without awareness of the flow of time and its influence on us, we cannot really understand agentic processes" (O’Meara & Campbell 2011, 449). We collect varied types of data beginning with an initial biographic questionnaire (e.g., age, field, role, prior work experience, possible future careers). This is followed by activity logs, capturing the experiences/activities and perceptions of an individual in a particular week and completed 3-5 times over a 10-month period. Next a pre-interview questionnaire provides a retrospective view of the year before an annual interview. The interview probes responses on both the pre-interview questionnaire and activity logs, introduces new questions about past and present experiences, and asks individuals to imagine their near futures as well as several years forward. At this point, those who wish to continue, begin the annual data collection cycle again. This data collection approach enables us to capture changes in intention as they occur rather than in retrospect. It is the accumulation and analysis of multiple narratives that enables a longitudinal perspective (the cameos here each represent summaries of between 15 and 25 documents including a minimum of two hour-long interviews).
 Without a team, it would have been extremely difficult to carry out this research.