Harmonizing DevOps Taxonomies

We applied a novel GT process that extends the Charmaz's GT variant to allow multiple researchers to participate in the coding process, that is, collaborative coding. This GT process ensures consensus on the constructs that support the theory, thus improving the rigor of qualitative research (cf. here). To this end, we analyze the intercoder agreement (ICA) to measure the extent to which different raters assign the same precise value (code or category) for each item rated (qualitative data item or quotations).

The GT process is composed by the following stages: initial/open coding, selection of core categories, selective coding, sorting, theoretical coding, and theory writing (see Figure 1). The outputs of these phases are the coding, the memos, the categories, and the resulting theory, all managed through the tool Atlas.ti v9.