A thematic content analysis was used to analyse the data. This was an iterative process that involved several readings of the verbatim transcripts, field notes and listening to the audio recordings. Initial analysis was based around the discussion topics used in the question guide (these having been developed from earlier studies on similar population groups).
Key early themes (proto-themes) that emerged from the transcripts were labelled and coded. These proto-themes were flexible categories which, following further study of the transcripts, were expanded and modified to encompass the issues that arose from within the data upon comparison of segments of text. These initial codes were developed with an assistant (RB) who also took field notes and checked the transcription. The codes were marked on the margins of the transcripts against corresponding areas of text, to enable retrieval and further sorting of the text.
Emerging themes were reviewed by an academic anthropologist (MF) at the University of the West Indies. Using ‘MS Word’ software, further refinement of the themes was achieved by placing segments of similarly coded text next to each other to explore deeper meaning and context. Final themes were developed and reported with a description and related quotations to underline their meaning.