Open coding
Open coding is the first stage of coding qualitative data, in which the researcher reads the material closely, often line by line, and labels each meaningful fragment with a short code. It opens the data up before any categories are imposed.
How it works
The analyst works through a transcript asking of each passage: what is happening here, what is this person doing, feeling or describing? Each answer becomes a code, often in the participant's own words (an in vivo code) or as a short descriptive phrase. Early codes are provisional and numerous; later passes merge, rename and drop them.
Open coding comes from grounded theory, where it is followed by axial coding, which connects codes into categories, and selective coding, which finds the core category. Constructivist grounded theory calls the same first stage initial coding. Outside grounded theory, a similar close first pass starts most inductive analysis, including thematic analysis.
A worked example
For its first round, a pharmaceutical brand team interviews 8 people managing a chronic condition. Open coding the first transcript alone yields 46 codes, among them "hiding the injection from coworkers" and "refill dread". After all 8, the team has 210 codes, which later passes merge and prune to 38.
Related
Automate this in Pulse
Open coding line by line needs the participant's exact wording, and each Pulse interview arrives as a full transcript to lift in vivo codes from.