3.2.2.3 Procedure and Analysis

The collection of extensive and complex textual data, i.e. long interview transcriptions and relevant documents, required a careful sorting of the data as well as the application of the constant comparison method (coding) prior to data analysis. During this initial phase, descriptive data were separated from meaningful data. The descriptive data containing background information regarding the cases – organisations and the issue or problem under examination – were used to frame the case and the examined issue. Next, codes were created by means of open in vivo coding using software for qualitative analysis, i.e. ATLAS.ti, to constantly compare and contrast the ideas captured in the text contained in the meaningful data and to sort and group the data on the basis of periodic recurrences and patterns in the interviewees’ responses. These were later integrated into categorised themes (see the ‘initial codes’ in full in Table 3.5).

We then developed a coding template including codes that had emerged deductively from the theoretical analysis (see the ‘legal’ codes in full in Table 3.5), which we subsequently refined with the emerging codes from the interview transcriptions (see the ‘refined codes’ and ‘progress themes’ in full in Table 3.5). Combined with our objective to capture experiences in the text and to add quality to the data, the use of an analytical technique called ‘template analysis’ was preferred.1 This template analysis was used to constantly compare and contrast the pre-existing legal concepts in the template which had emanated deductively from the theoretical analysis with real-life concepts in the template that had emerged from the data inductively, which we integrated into themes with respect to stakeholder and employee participation in the governance of social enterprises.2 In a follow-up round, the respondents were asked to provide feedback and validation with respect to the case study. This feedback was included in the text of the analysis.

Table 3.5: Codes and themes in the two Greek case studies