EVALUATIVE CRITERIA AND EPISTEMIC COMMUNITIES 111sented in the handbook edited by Denzin and Lincoln (2000), the kind of research done under the qualitative
label is typically consistent with the interpretive approach emphasized in this volume.
- As Miles and Huberman emphasize in their introduction, “the data concerned appear in words”
(1984, 21, emphasis in original). They do suggest converting words to numbers in some cases, but they
advise that “this is not always wise. Even when it does look like a good analytic strategy, our counsel is this:
keep the numbers, and the words you used to derive the numbers, together in your ensuing analysis. That
way one never strips the data at hand from the contexts in which they occur” (1984, 21, original emphasis). - Among the terms in the last column of Table 5.3, “trustworthiness” is used as a general rubric for the
quality of research, Moustakas’s “testing out with others” appears similar to the technique of “peer debrief-
ing,” and Seale’s “replication” and “validation,” like Atkinson, Coffey, and Delamont’s (2003) use of
“evidentiality” and “veracity,” appear to be criteria rather than techniques in that they do not directly address
the “how to” question. - Readers who might be under the impression that such complexity is unique to interpretive criteria
debates should consult Adcock and Collier’s discussion of the variety of definitions of, and approaches to,
“validity” in psychology and political science (2001, esp. 536–38). For more detailed discussion of criteria
and techniques, see Schwartz-Shea (2004). - Morse and Richards argue that “to claim that reliability and validity have no place in qualitative
inquiry is to place the entire paradigm under suspicion; such a claim has ramifications that qualitative in-
quiry cannot afford, and it diverts attention from the task of establishing useful and usable measures in the
qualitative context” (2002, 168). - Miles and Huberman, who have been arguing since 1983 for the development of distinct criteria
appropriate to qualitative-interpretive work, emphasize that “the problem of quality in qualitative studies
deserves attention on its own terms, not just as a justification device” (1994, 277, emphasis in original). - What I cannot claim to know based on this methodology is the extent to which these selected textbook
authors are members of particular disciplinary interpretive epistemic communities. That claim would require
considerably more research on text usage by discipline. For an example of such research showing the lack of
recognition of interpretive research in the discipline of political science, see Schwartz-Shea and Yanow (2002). - The labels of “first-order” and “second-order” were chosen over two other possible sets of labels for
these two categories. I rejected “criteria” and “technique” based on my analysis that these terms could not be
coherently distinguished in the literature (Schwartz-Shea 2004). I rejected “primary” and “secondary” be-
cause the connotations of these terms imply a normative ordering that does not reflect my intention of
describing what I found vis-à-vis my “index test.” - Atkinson, Coffey, and Delamont argue that the term “thick description” is “dreadfully misunderstood
and misrepresented” in that it means more than attention to detail but also “a disciplined appreciation of the
over-determination of cultural phenomena” (2003, 114). Although I do not disagree with their analysis, my
point is to emphasize the comparative power of the term—the clear way that it evokes a practice—which
derives in part from its implicit contrast to the variables research gestalt. - Autoethnography “refers to a particular form of writing that seeks to unite ethnographic (looking
outward at a world beyond oneself) and autobiographical (gazing inward for a story of one’s self) inten-
tions” (Schwandt 2001, 13). It is commonly the use of the researcher’s own personal experience with some-
thing, e.g., illness, to understand how others might similarly experience it, e.g., interactions with physicians,
hospitals, etc. See, e.g., Greenhalgh (2001). - There is a striking contrast between the interpretive understanding of triangulation and the term’s
usage within the variables gestalt. There, triangulation is attractive because of its promise of discovery and
precision—the pinning down of a concept (as achieved through composite measures) or the identification of
a “true position” (Neuman 1997, 151). In such approaches, corroboration is the theme; the point of triangu-
lation is to use methods that offset each other’s weaknesses (Singleton and Straits 1999, 393); this empha-
sizes consistent findings across methods, and inconsistencies are regarded as problematic. This perspective
on triangulation reflects the positivist preference for parsimonious explanations, which are often described
as “elegant” and “powerful.” For a critique of the argument about “offsetting weaknesses,” see Atkinson,
Coffey, and Delamont (2003, 144–48). - Pader (chapter 8, this volume) argues that, through field research experience, scholars can learn an
“ethnographic sensibility” that can improve their reading practices of others’ research, as well as of others’
experiences, over the lifetime of their careers. - Literally “going back” is not always possible, as is most evident in the case of historical research; cost
and other barriers (e.g., participant privacy) are impediments in contemporary studies. But researchers have