Beyond the understanding gained from analysis of secondary data obtained from the literature, reports and case studies, data analysis will involve taking the raw primary data gathered from interviewees and focus group participants, and organising them into important themes, categories and case examples. With regard to phenomenological analysis, data can be broken down into four discrete levels or steps (Sanders 1982). These are outlined in Table 2.3.
Table 2.3 Phenomenological Analysis of Data (adapted from Sanders 1982).
Step | Question | Action | Purpose / Explanation |
1 | How can the phenomenon or experience under investigation be described? | Transcribe interviews | Transcribed narratives identify and describe the qualities of human experience and consciousness |
2 | What are the themes emergent in the descriptions? | Identify the themes that emerge from the descriptions | Themes refer to commonalities present within and between narratives. Themes are identified based on the importance and centrality accorded to them rather than on the frequency with which they occur |
3 | What are the subjective reflections of the themes? | Develop noetic / noematic correlates | Noetic / noematic correlates represent the individual's perception of the reality of the phenomena under investigation. Interpretation of these correlations is fundamental to the identification of essences or of what an experience 'essentially is' |
4 | What are the essences present in the themes and subjective reflections? | Abstraction of essences or universals from the noetic / noematic correlates | Abstraction is accomplished through intuition and reflection or eidetic reduction. If noema is described as the 'what' of experience and noesis as the 'how' of experience, then essence may be described as the 'why' of experience |
As previously discussed and in conjunction with the steps outlined in Table 2.3, this research involves transcribing interviews with the edited transcripts being returned to the interviewees for validation before including the data as part of the research project (as per step 1) (also see section '2.5 Research Validity and Reliability', below). Following the advice of Saldaña (2011: p.45), transcripts will exclude informal and broken speech e.g. 'ahs' and 'ums' as well as influent speech strings e.g. 'there was a kind of a…', unless considered by the researcher that these types of speech patterns contain insightful inferences.
To assist with the data analysis process outlined in steps 2, 3 and 4, the software application NVivo (version 10) will be deployed (Richards 2005: p.106). As analysis proceeds, data can be expanded from transcripts and summaries into more detailed descriptions and where data and commentary can be inter-weaved (Richards 1999). This will aid the development of the integrative partnership, risk and performance management model process.
This research will use the open coding technique at Step 2 to identify themes from the interview transcriptions. This will involve developing initial classifications / labelling of concepts during a preliminary attempt to condense the acquired data into categories (Babbie 2007: p.385; Neuman 2007: p.330).
Axial coding will then be used during steps 3 and 4 to test the need for regrouping the data (forming a justification for changes that may be made to the data e.g. reducing the number of codes developed during open coding, the development of new codes, etc) (Saldaña 2009: p.160; Neuman 2007: p.330), as well as for identifying core concepts from the data (Babbie 2007: p.386). During axial coding, the following may be considered: potential causes and consequences, conditions and interactions, strategies and processes, and categories that could be merged together (Neuman 2007: p.331). Data analysis for this research will generally follow these four steps. Chapter 8 (see '8.4 Data Analysis Processes') expands upon the thematic development logic, which will be based on appropriate research questions from the data collection instrument (refer to '8.2 Design and Testing of the Data Collection Instrument').
Analytic memos may be used at any / all steps outlined in Table 2.3. This involves keeping a record of pertinent thoughts that may contribute towards developing and refining data interpretations (Saldaña 2011: p.98) as well as establishing a foundation for validating the analysis (Richards 2005: p.62, p.74). Reflection may extend, for instance, to how the researcher relates to the interviewees and the phenomena under study; emerging patterns and concepts; possible linkages, connections and overlaps; problems with the research and their ethical implications; and future directions for the study (Saldaña 2009: p.34-38). All memos will be given an appropriate heading and dated accordingly (Saldaña 2011: p.98; Richards 2005: p.75).