Table 2 Phases of data analysis

1. Data familiarisation

Transcripts from interviews were read several times and initial notes and thoughts were written down

“Positive experience, no obstacles” (notes Participant 6)

2. Generation of initial codes

Coding was performed separately by two researchers (EB, CB), and data extracts collated within each code. Codes were then compared and discussed between the two researchers

”I have a craving for it (the strength training) now, because I think it’s so fun, so I still exercise at least three times per week”

Initial code: Developed craving for strength training

3. Generation of initial themes

Codes were collated into initial themes and subthemes independently by two researchers (EB, CB), a thematic map was constructed for each interview

Initial code: “Developed craving for strength training”

Initial theme: Lifestyle change—exercise addictive

4. Initial themes reviewed

The initial themes were reviewed by reading all data extracts belonging to each initial theme and the initial themes reworked to generate a preliminary thematic map of the whole data set. The themes were further reviewed by re-reading the whole data set and ensuring the accurate representation of the raw data in the themes

See Fig. 1a and 1b for example of the preliminary thematic map, and reworked thematic map

5. Refining and defining themes

The themes were refined in an analytic process of discussing the meaning and definition of each theme, and checking the final themes against the raw data set

See Fig. 2 for final thematic map

6. Report writing

The themes presented in a report with definitions of each theme and the story they tell, along with data extracts represent the themes