There is no standardised approach to the analysis of qualitative data (Saunders et al, 2009). The approach used refines data in these four steps:
- Categorisation
- Unitising data
- Recognising relationships
- Developing and testing hypotheses to reach conclusions
Categorisation
The interviews were recorded and transcribed in order to be analysed. Meaningful categories and themes were developed from initial readings of the transcripts and were not predetermined. It was common for the same factors to be the influences of different questions especially as the interview question was on a focused branch of an established model. Observing the frequency of themes and factors mentioned across the interviews can diminish the effect of anecdotes. Anecdotal evidence are a common criticism of qualitative research which refer to the subjective views of the participants own understanding of their individual social world and one-off occurrences that are unrepresentative (Bryman & Bell, 2007).
Unitising data
This stage involves the further refining of data by reorganisation them into the categories determined in the preceding step using identifiable packets of usable information, also known as ‘open coding’. Using the computer, transcripts were cut up, tagged and sorted to produce a collection of categorised information. This rearrangement changes the raw data into a more manageable and comprehensible form while cutting down on noise irrelevant to the research question.
The researcher has decided to categorise and unitise data in an interview map that resembles a mind map
Recognising relationships
Both Miles & Huberman (1994) and Yin (1994) deem the generating of categories and reorganisation of data as the process of analysis. During these steps categories may be further refined, subdivided or split as data is further analysed and new insights gained. This process of reworking, resorting and redefining the categories covered during the interview is a crucial stage in the recognising of relationships, hierarchies and mental pathways within your data. This further rework of the prior open coding stage to look for relationships can also be called ‘axial coding’
Developing and testing hypotheses
As the content of the interview becomes more familiar a hypotheses may be developed. “The appearance of an apparent relationship or connection between categories will need to be tested if you’re able to conclude that there is an actual relationship” (Saunders et al., 2009). Marshall & Rossman (1989) state that testing the hypothesis involves seeking alternative or contradictory examples that do not conform. Both positive and negative examples found during axial coding can serve to develop your hypothesis keeping in mind the limitations of sample size and cross-sectional timing etc. of your research.
Next section: https://jl00282.wordpress.com/dissertation-part-2/companies-perspective/qualitative-methodology/summary/
The New Exchange 