Elien Neimeijer

72 experiences of the clients. The duration of the interviews ranged from 25 minutes to 58 minutes with a mean duration of 37 minutes. The interviews were audio-recorded with the participants’ informed consent and transcribed (verbatim) for coding purposes. Afterwards, the audio recordings were deleted. Analysis The transcribed interviews were analysed using IPA. IPA can be used for a detailed ex- ploration of how people make sense of their personal and social world by exploring an individual’s personal perception or experience as opposed to an objective description of the object or event itself (Smith & Osborn, 2008). IPA is a dynamic process based on the assumption that the researchers have an active role in the research process and influence the extent to which they access the participant’s experience and how they interpret and make sense of that experience. The clinical experiences of the researchers are important to be able to properly interpret the experiences of the client in the light of their complex problems and unique context in which they reside (Zomerplaag, 2017). Therefore, we thought IPA better suitable than other procedures such as Grounded Theory (Glaser & Strauss, 1967) and thematic analysis (Braun & Clarke, 2006). Data analysis was carried out by the first and second author independently and fol- lowed the stages set out by Smith and others (2009). The first stage involved the close reading and rereading of the transcript to become familiar with the interview content. Second, the transcript was read line by line, noting points of interest and significance on a descriptive, linguistic and conceptual level. Third, the transcript and initial notes were reread, with emergent themes noted. At the fourth stage, themes that were considered as connected were grouped into overarching themes and given a descriptive label, after which these groups of themes were discussed within the research team. As a result, some additional changes were made in the grouping or descriptive labelling of themes. To ensure that the analysis was carried out in a rigorous way and that interpretations made by the first and second author were of an explicit nature, all stages involved a discussion with a third researcher to provide an audit of the analysis. These stages were repeated for each transcript after which the overarching themes for each interview were compared and discussed with the research team to find patterns across cases. After 12 interviews had been reviewed, no new theoretical aspects emerged from further coding and com- parison and saturation was reached (Mason, 2010). The dimensions and themes that