Designed Generalization from Qualitative Research

In our earlier work on generalizing from qualitative research (GQR) we identified our two-decade struggle to have qualitative research outcomes formally" listened to" by policy personnel and bureaucratic systems in general, with mixed success. The policy sector often seems reluctant to acknowledge that qualitative research findings can be generalized, so impacts tend to be informal or simply ignored. The" official" methodological literature on generalizing from qualitative research is epitomized by Lincoln and Guba's (1985) still oft quoted," The only generalization is: there is no generalization"(p. 110). We now understand there are many alternative possibilities for generalizing. In this paper we hope to provide a platform for discussion on GQR. We suggest Normative Truth Statements (NTS) as a foundation. NTSs, used in our proposed generalizability cycle, are a potential key to ensuring designated qualitative research methodology provides a capacity for generalization—and therefore be considered as a valid form of evidence in
policy decisions. In other words, we need a platform to articulate how to design qualitative research to maximize the type and scope of generalizability outcomes, referred to here as Designed Generalization from Qualitative Research (DGQR). Five steps of DGQR, using progressive NTSs in the generalizability cycle, are proposed as a way forward in understanding how generalizing from qualitative research may be made more transparent, accountable, and useful. The five steps are illustrated by reference to two example studies.