0.2 Research design and methodology

• the research questions
• the interview guide
• themes, concepts and categories used in prior studies
• prior knowledge of the researcher
• the data itself

The mechanics for organising the data were based on the descriptive analysis technique as outlined by Terre Blanche, et al, (2006; 252)

• reading through the interview transcripts and jotting down notes and comments
• selecting the richest and thickest portions of the interview transcripts
• listing the emerging themes and clustering and encoding them
• identifying the major themes and sub-themes
• analysing categories of the themes and sub-themes and (making sense) of what they mean

The data analysis was managed thematically. The categories and themes were the categories and themes used in the interview guide and supplemented by the literature review.

Validity of the research findings

The production of a convincing narrative report of the research has most often served as de facto validation, even if the only thing it validates is the researchers writing skills rather than his or her observational capacities [Kuklick in Denzin and Lincoln (2008; 163)]. McMillan and Schumacher (2001; 478) concur when they state that selecting trustworthy data also involves an awareness of the researchers assumptions, predispositions and influence on the social situation.

Qualitative researchers generally validate their data by means of triangulation. Denzin in McMillan and Schumacher (2001; 478) explain triangulation as the cross-validation among data sources, data collection strategies, time periods and theoretical schemes. To find regularities in the data the researcher compares different sources, situations and methods to see whether the same pattern keeps recurring – a theme could be cross checked by comparing data found in artefact collection (minutes, memos and so on), informant interviews (principals, teachers, and so on) and field observations, as illustrated in figure 7. Flick in Denzin and Lincoln (2008; 7) state that qualitative research is inherently multi-method in focus, – the use of multiple methods or triangulation reflects an attempt to add rigour, breadth, complexity, richness and depth to any inquiry. Triangulation is not a tool or strategy for validation but and alternative to validation.

Figure 7: Triangulation for Logical Pattern

Artefact Informants

Collection

Social Scene

or

Process

Field

Observations

Source: McMillan and Schumacher (2001; 478)

The critics of qualitative research generally claim that qualitative research is too subjective, the number of cases sampled is too small, or that the analysis and interpretation is not scientific – to mention but a few criticisms. The questions that qualitative researchers need to ask themselves is: to what extent can the data analysis be trusted; how much confidence can the researcher place in his / her analysis; is the data analysis presented to the readers in such a way that they can verify and validate the findings for themselves. Lincoln and Guba (1985; 290) suggest that qualitative research should be assessed in terms of credibility, transferability, dependability and conformity of the data.

• credibility relates to how the researcher reconstructs the participants views into the process of the inquiry. One method is, according to Lincoln and Guba, prolonged engagement, that is the investment of sufficient time to achieve certain purposes, learning the culture (of the respondents), testing for misinformation introduced by distortions either by the self or by the respondents and building trust (with the respondents).
• transferability refers to the degree to which the findings can be applied to other contexts. Terre Blanche, et al, (2006; 381) state that to create a foundation for transferability and to allow other researchers to use the findings in making comparisons with their own work it should contain an accurate description of the research process, an explanation of the arguments for the different choices of methods and a detailed description of the research situation and context.
• dependability for the qualitative researcher is equivalent to reliability in quantitative research that is it should produce the same results in a similar context. In qualitative research this replicability is impossibility because of the flexibility of the research design. Guba and Lincoln (1989; 242) state that far from being a threat to dependability such changes and shifts are a hallmark of a maturing and successful inquiry – but such changes and shifts need to be both tracked and trackable.
• conformity is concerned with establishing that the data and the findings and interpretations are linked. Lincoln and Guba (1885; 326) refer to an audit whereby a third party reviews the audit trail maintained by the researcher including recordings, transcripts, interview guides, lists of interviewees, field and observational notes, and so on.

Summary

The chapter commenced with the rationale for the constructivist approach to the research study and the motivation for the use of interviews, in particular the semi-structured interview approach. The purposeful sampling of the research sites was elaborated upon with emphasis on the information-richness and ethical considerations. The data collection and data analysis methods were then discussed with an overview of the validity of the research findings. The following chapter focuses on the research findings.