Qualitative Coding Tool

Coding is the step where raw transcript text becomes analysable data. A code is a short label attached to a passage that captures what the passage is about, and a coded segment is the passage together with its label. Codes can stay close to the participant's own words (sometimes called in vivo codes) or describe a concept the researcher sees in the data. Most projects move through several rounds: an open first pass, a review that merges and splits codes, and a final pass applying the settled codebook to every document.

The codebook is what makes coding transparent and repeatable. Each entry should give the code name, a one-sentence definition, the inclusion criteria (when to apply it), the exclusion criteria (when a neighbouring code fits better) and an example quote from the data. Exclusion criteria are the part most often skipped and the part that does the most to stop two coders, or the same coder a month later, from filing similar passages under different codes. When a second coder is involved, agree the codebook on a sample of transcripts before coding the rest.

Codes are then grouped into themes: patterns of shared meaning that answer the research question. A theme is more than a topic summary; it says something about the data, and it is supported by codes from several participants. Sub-themes capture distinct facets of a larger theme. The frequency table helps you see whether a theme rests on one talkative participant or recurs across the sample, but frequency is a check, not the measure of importance. The step-by-step process behind this, from familiarisation to writing up, is set out in our Braun and Clarke style thematic analysis walkthrough, and the more counting-oriented alternative is covered in the guide to coding frames in content analysis.

Memos are the researcher's running notes: why a code was split, a hunch about how two themes relate, a participant who contradicts the pattern. Kept alongside the codes, they become the audit trail for your methods section and the raw material for the discussion. Desktop packages such as NVivo, ATLAS.ti and MAXQDA offer these same building blocks with more advanced querying; if your institution requires one of them, our overview of working in NVivo explains how the same codebook carries across.

Coding primary transcripts is different from synthesising published qualitative studies. In a systematic review of qualitative research, the "data" are the findings reported by included studies, and methods such as thematic synthesis and meta-aggregation apply, described in our article on qualitative evidence synthesis methods. Appraise the included studies first with the JBI checklist for qualitative studies, then use the Research Gold systematic review screening tool to take the review from screening through to synthesis. To compare themes study by study, the synthesis matrix generator lays them out in a grid.

When the dataset is large, the deadline is close or a journal expects a fully documented analysis, our PhD methodologists can carry out the full qualitative data analysis for your project, from codebook development and double coding to the written findings chapter.