For qualitative research, validity is about ensuring that the research instruments actually relate to the issues under investigation (Kumar 2011: p.184; Blaxter, Hughes and Tight 2004: p.221). Put differently, if researchers cannot demonstrate a logical fit between the broader research design and the choice of methods, the research will be invalid (Neuman 2007: p.115). Validity can be tested by asking participants to confirm and approve their interview transcripts (Marshall and Rossman 2011: p.42; Kumar 2011: p.278). Ideally, and according to Easterby-Smith, Thorpe and Lowe (2002: p.135), this should occur during the pilot stage but as this research adopts a phenomenological stance, each interview transcript will be validated (Sanders 1982). Furthermore, validity trails will be established (Richards 2005: p.143-144) to provide increased assurance in the research process. This will include:
- Analytical memos (see Attachment A as an example);
- An account of the research as a "whole" or a "pattern" through which concepts and ideas have been drawn together (see Chapter 6); and
- Draft research models (see the Integrated Management Model: Chapters 7 and 8).
Reliability, on the other hand, is a factor that contributes to the achievement of validity (Cooper and Emory 1995: p.153) and is confirmed when the research findings e.g. patterns, themes, worldviews, etc (Labuschagne 2003) can be repeatedly used by other researchers which lead to similar results (Kumar 2011: p.184; O'Leary 2005: p.59).
The constructionist approach adopted for this phase of the research recognises that all research participants will construct meaning according to their individual preconceived positions. Such research will always contain bias, both from the researcher and from the subjects of a study. It is anticipated that the effect of these biases can, at least to an extent, be mitigated by reflecting on assumptions made by the research participants (Kapoor 2004) and by the researcher referring back to original text to ensure the iterative process of the research is maintained.