Prescription for research practice? Grounded theory in qualitative evaluation

Abstract

We commend Bailey et al (1999) for vitalizing the discussion of evaluation in qualitative social geogra phy. The authors focus on strategies such as critical thinking and reflexive management as part of a grounded theory approach to research and evalu ation, and they are quick to point out that this is not the only way to achieve research that is worthy of attention. We agree that the tension between 'science' (quantitative research, structured evalu ation) and 'creativity' (qualitative research, theoreti cal development) found in the ways in which researchers design, carry out and evaluate qualitative social research may not be as pronounced as is sometimes portrayed in the literature (Hodge 1995; Lawson 1995). Nevertheless, intermeshing the two remains more problematic than Bailey et a! seem to claim. We would like to further the discussion on the issue of evaluation by commenting on Bailey et al in the following areas: a potentially blurred distinction between standardized evaluation and systematic evaluation; mechanisms for conveying rigorous prac tice to readers; the relationship between researcher and participant interpretations; a tension between 'data-led' and 'reflexively managed' research, when criteria for evaluation are formulated; and the need for discussion on the capacity for transferability of interpretations. We want to be clear that we are not dismissing the grounded theory of Bailey et a/-we use many of their practices (such as reflexive management) in our own work-but we are con cerned to highlight some potential problems with the relationship between qualitative evaluation and grounded theory. It is important to be clear about what is meant by evaluation. Bailey et al argue for a distinction between standardized evaluation and systematic evaluation, favouring the latter. We have made similar arguments (Baxter and Eyles 1997), but are conscious of the possible fusion of the two notions. Standardization implies prescription of practices, while systematic evaluation involves more general principles for evaluation. It seems that some of the principles for evaluation outlined in Table 1 of Bailey et al are prescriptions for particular practices. For example, the requirements of 'reflexive manage ment', 'constant comparison' and the 'use of archives for data' may be regarded as specific and potentially prescriptive. This runs contrary to the idea that evaluation in qualitative research should con cern general criteria. Indeed, the three practices may well appeal to some general principle(s) for evaluation-credibility, traceability, honesty-that are not described in the table. While we should not prescribe a single set of qualitative research practices, it is the practices themselves that are typically used as a shorthand way of conveying to others why our research is worthy of attention. This shorthand may become problematic if practices are reified and not scrutinized. The practice of triangulation may, for example, be reified by virtue of its common use in qualitative research. This may be precisely why Bailey et al avoid including it explicitly in their list of 'examples of practice' in Table 1. A mere mention of the use of triangulation is rarely sufficient for assess ing trustworthiness, since there are many forms of triangulation and many methods of performing them (Baxter and Eyles forthcoming; Denzin 1978). Thus, we agree with the tenor of their Table 1, which concerns the documentation of research practices in order to demonstrate safeguards against