Skip to main content

Table 1 Stages of thematic analysis

From: Developing a model of best practice for teams managing crisis in people with dementia: a consensus approach

Braun and Clarke stages Our methodology
Familiarising yourself with your data Data were transcribed verbatim by a transcription company, and quality checked by a researcher (ET). Two researchers (JY and MS) each read half of the whole set of transcripts and noted similarities, contrasting accounts, common patterns, and insights.
Generating initial codes JY and MS discussed these notes to develop initial codes, paying particular attention to aspects of crisis team practice or service provision that were mentioned, and the outcomes that participants reported as resulting from these activities. This drew on the research team’s existing knowledge from conducting a scoping survey (deductive analysis), but also left space to identify patterns of ideas discussed by the interviewees.
Searching for themes JY and MS discussed the codes and collated similar codes into potential themes. A theme index of the themes and subthemes was created, where each theme and subtheme was given a numerical identifier.
Reviewing themes Themes were mapped back on to transcripts in the right hand margin using the theme index numerical identifiers. Every instance of each theme and subtheme was identified and transferred to a framework, which consisted of a matrix for each theme, with a column for each subtheme and a row for each participant. JY and MS checked that all themes remained independent, and any that did not were combined with other themes. Thematic models were discussed with the wider research team as they were developed and refined.
Defining and naming themes JY and MS used the framework for each theme to summarise the content of each subtheme as a short statement. This enabled the themes and subthemes to be thoroughly operationalised and named accurately, capturing the essence of each theme.
Producing the report Examples that provided the best and most representative evidence for each theme were highlighted in the framework of each theme. Narrative summaries of the themes were documented and stored for use in further report writing.
For the purposes of the consensus process, all aspects of crisis team working and service provision were identified and documented, clustered by similarity or relatedness.