Key points are not available for this paper at this time.
Background Modern open science initiatives have implications for qualitative researchers, particularly in relation to 'open' data. Aim We describe the process of data-sharing for a qualitative interview study, to make transparent our thinking and decision-making around open data for our particular project. Methods The initial project aimed to explore perspectives of young adults on what precipitates and maintains feelings of loneliness among this age group. Participants did not necessarily need to discuss personal experiences of loneliness. 27 virtual interviews were conducted with young adults between 18 and 25 years. Results 26 participants consented to make their anonymised transcripts available via the Irish Qualitative Data Archive; 24 transcripts were archived. One excluded transcript contained very rich descriptions of niche social activities and rare life events that could identify the participant. However, redacting these details meant that there was scant information left, and what was left was divorced from the rich context in which it was provided. For a further transcript, the participant had a relatively rare combination of demographic characteristics. These, combined with the detailed discussion of participants' own characteristics and activities in the interview, meant that risk of identification of the participant was too great, even if access was restricted to educational and research use. Interestingly, one participant (whose transcript was archived) used no identifying information that definitively merited redaction or replacement, suggesting that even when participants consent to data archiving, this might shape the way in which they engage with the interview. Conclusions Data-sharing is more complex than 'open' or 'closed'. Purposeful consideration of the why and how of data-sharing can support researchers to make optimal use of participant data while being mindful of the ways in which data sharing can shape the context of qualitative data collection.
Creaven et al. (Fri,) studied this question.