Thematic analysis begins with deep familiarisation of your data. Read and re-read your transcripts or field notes before generating any codes. Premature coding leads to superficial themes that will not survive examiner scrutiny.
From initial codes, develop candidate themes that capture meaningful patterns across your dataset. A theme should be substantive — not just a topic that appears frequently, but one that tells us something important about the research question.
The final analytical write-up must link your themes explicitly back to your research objectives, using direct participant quotes as evidence. Themes without evidence are interpretations without data.
