Selective coding
In grounded theory, selective coding is the stage where the analysis converges: the researcher chooses the one concept that best explains what is happening in the data, the core category, and organizes every other category around it. It follows open and axial coding.
How it works
By this stage, the analysis has many categories linked through axial coding. Selective coding asks which one concept explains most of what is happening: the core category. It must appear frequently, connect to the other categories and account for variation in the data. The researcher then refines every other category in relation to it, fills gaps with further data where needed, and writes the storyline that connects them.
The result is a grounded theory, an explanation of the process the study is about, built up from the data. Chun Tie and colleagues describe the findings as a set of interrelated concepts, with explanatory statements linking each category to the central core category, often written through a storyline that ties them together.
A worked example
After 25 interviews with cabin crew about a new tablet app, a researcher at an airline has 8 categories. Selective coding identifies the core category "protecting my reputation". Crew adopt the app when early use makes them look competent in front of passengers and colleagues, and avoid it when it risks a visible mistake. The rollout adds a practice mode crew can use at home before their first flight with it.
Related
Automate this in Pulse
A candidate core category earns its place if it accounts for most of Pulse's themes, and a theme it leaves unexplained is the case selective coding sends the analyst back to.