How Not to Let Secrets Out When Conducting Qualitative Research With Dyads

Abstract

Confidentiality is one of the cornerstones of research involving human participants. Researchers are the frontline gatekeepers of their participants' right to confidentiality, and situations can arise that challenge this responsibility. This is the case when individuals who have shared a common experience (i.e., dyads) are interviewed separately, but interview results are disseminated within the context of dyads. Based on our experience of conducting research with dyads and given how little literature is available to serve as guide, we set out to write this article to share the knowledge we acquired and the solutions we found. We will describe both the ethical challenges and the methodological decisions involved in conducting qualitative research with dyads. The article also describes different modalities of dyadic analysis, their benefits and drawbacks. This endeavor seems especially relevant as research with dyads is emerging in several domains involving couples, families, caregivers and health.

Resolution of diagnosis and parental attitudes among parents of adolescent kidney recipients

“…These one‐time interviews took place at a location deemed most convenient by each participant, the nephrology unit (n = 3) or the participant's residence (n = 6). They were conducted in a flexible manner to maintain the focus on what emerged from participants' discourse as most relevant and important to them . Each lasted between 60 and 90 minutes and was digitally recorded to facilitate accurate transcription .…”

Multi family member interview studies: a focus on data analysis

“…After reviewing a number of example studies (e.g. Alexander et al ., ; Dancyger et al ., ; Eisikovits and Koren, ; Ummel and Achille, ), we advanced the following overarching data analysis strategy: (1)Interpretative Phenomenological Analysis (IPA) of all interviews (interview by interview) (2)Analysis of each family unit (3)Integration of themes and subthemes of each family unit, resulting in new cross‐family thematic categories (4)A continuous auditing process throughout all phases of the analysis. …”

“…In order to assure ‘network confidentiality’ when doing research with families, one needs to balance anonymizing details and maintaining authenticity (Harden et al ., , p. 447). Thus, for ethical reasons, the dissemination of the results should be at a general level, rather than at a dyadic or systemic level (Ummel and Achille, ). This means that the strength of this analysis (being able to offer an in‐depth understanding of shared family realities) at the same time encompasses the method's main limitation: we simply cannot exemplify all new insights at a systemic level when seriously considering our responsibility to protect (network) confidentiality.…”

“…Furthermore, Morris (2010) outlined the advantages of combining both separate and joint interviews for the study of the needs of patients and their carers and -in so doing -tuning into the participants' interview preferences and gathering the richest possible data (Morris, 2001). A number of researchers have outlined the ethical dilemmas that come with interviewing family members alone or together, sequentially or at the same time (see, for instance, Eisikovits and Koren, 2010;Forbat and Henderson, 2003;Norlyk et al, 2016;Ummel and Achille, 2016). The second aspect, however, data analysis, remains largely underrepresented in the literature (Eisikovits and Koren, 2010;Ummel and Achille, 2016).…”

“…A number of researchers have outlined the ethical dilemmas that come with interviewing family members alone or together, sequentially or at the same time (see, for instance, Eisikovits and Koren, 2010;Forbat and Henderson, 2003;Norlyk et al, 2016;Ummel and Achille, 2016). The second aspect, however, data analysis, remains largely underrepresented in the literature (Eisikovits and Koren, 2010;Ummel and Achille, 2016). While the complexity of this type of analysis can be overwhelming (Warin et al, 2007), only a few articles pay more than usual attention to data analysis (e.g.…”