5 | 135 Career, family, and sport participation: a simultaneous exhibition? & Culver, 2014; Tozetto et al., 2017). Interviewers invited the participants to elaborate on the stories behind the major life events they experienced and why and how this affected their sport participation, by asking the central narrative generative question, which reads as follows (translated from Dutch): “Can you tell me how your daily life and participation in sport and sportive activities looked like before and after the major life events you experienced?” For both samples, we employed a resource approach to enrich the interview guide with follow-up questions, which could be deployed if necessary, to probe into the underlying mechanisms of changes in resources and constraints via which major life events can affect sport participation, and how this affected sport participation. This acted as a set of back-up and reminder questions for the interviewers. However, all interviewers were trained to allow participants to set the conversational agenda, as narrative interviewing involves a subtle balance between retaining substantive focus and participant-driven conversational themes (Jansen et al., 2018). The central research team gave this ample attention in the training of interviewers, prioritising the conversational quality and richness of the narrative over strict application of the interview guide. Although interviewer styles were diverse, the central team verified whether the narrative character of the interviews was sufficient, and hence checked the reliability and quality of the data. A single interview typically lasted 30 to 45 minutes, with outliers ranging from 15 minutes to over one hour. Analytical strategy Interviews were audio-recorded, transcribed verbatim, and then coded and analysed using the software program Atlas.TI 8. Conceptually, interpretation of the narrative data comprised several distinct iterations in the analyses of both separate samples and the following overall analysis, maximizing internal validity of the results and minimizing the risk for interpretation bias while recognizing the meaningful subtleties in the stories (Bal, 1997; Boje, 2001; Derrida, 1979), as described below. Regarding the analyses of the two separate samples, first, the co-researchers placed marking points in the transcription where a discontinuity in the narrative occurred, for example because of a topic or tone change or verbal turn-taking, indicating the segmentation in the interview stream. A 45-minutelong interview typically yielded between 30 to 38 narrative fragments. Second, co-researchers labelled each of the segments applying appropriate codes
RkJQdWJsaXNoZXIy MjY0ODMw