From: Gary Ivory
Posted: Mon 4/02/2002

Can't resist this one . . . .

And while I'm on the subject, how do people distinguish, if at all, "qualitative research" and "interpretive research" and does this distinction have some impact on the issue of being "truly" qualitative?

These are great questions. Let me throw out the explanation I give my students and see if it helps at all, or if anyone wants to contest it.

I see the issue in terms of a matrix with two columns: some folks work with quantitative materials, others with qualitative. But the matrix has (at least) four rows (These differ based on which questions they think are most worth asking):

The empiricist row has researchers who work with quantitative and researchers like the late Matthew Miles with his colleague Michael Huberman whom I see as basically empiricists, cause they always kept asking, "Are our analyses 'right,' i.e., do they reflect 'reality.'"

The second row are the interpretivists, who seem to me to be far more interested in how people interpret phenomena inside their own heads than whether the phenomena accurately reflect some 'real' world outside. I think these researchers might use either qualitative or quantitative data also.

Next row we have the Criticalists, who seem to me to suggest that any research that ignores the fact that social systems are set up to create winners and losers is missing the point. These seem to me to concentrate on qualitative research, but I see no reason why they wouldn't find uses for quantitative.

In the fourth row are the Post-modernists, and I am not sophisticated enough about their perspectives to capsulize their views in a sentence. Nor do I understand whether they would ever want to use quantitative data.

So, I see your first question, "What is truly qualitative research" as unanswerable. I see your second question, "What is the difference between qualitative and interpretivist" as being misguided, since I think they are points on two different dimensions, and not really comparable.

Anyone? Anyone?

From: Tony Gallagher
Posted: Mon 4/02/2002

I think I tend to agree with the general thrust of Gary's comments, but perhaps approach the issue in a slightly different way. I think the two sides of the issue (qualitative and interpretative) reflect two different issues, that is, of method and methodology. Issues of method are really about technical questions on the appropriateness of a particular data-gathering approach to a specific problem. Issues of methodlogy, on the other hand, deal with the assumptions of the researcher on the social world and social knowledge.

A researcher who clings to a positivist worldview will privilege empiricism, quanitative measurability and a search for universals - such a person is likely to eschew the use of qualitative approaches altogether on philosophical grounds. Someone working within an interpretativist framework may draw on either tradition, but, as Gary suggests, there are some versions of this that are so tied up in relativist positions (perhaps some postmodernists fit in here) that they are left with little basis for judging appropriateness, never mind validity.

From: Joao Vieira da Cunha
Posted: Mon 4/02/2002

I think Gary strikes a chord here. 'Qualitative research' is a somewhat strange combination of words. There is qualitative and quantitative data. In quantitative data, researchers codify their observations using numbers and in qualitative data, researchers are, in a way less willing to rely on the accuracy of their observations and thus keep their observations intact. Thus I agree with Gary regarding his first dimension. Now, what I think merits reflection is the second dimension. Empiricism may be an elusive term here, especially because of its philosophical usage.

The choice between qualitative and quantitative here is driven by the textbook notion that 'method follows research question'. How and why questions are answered through qualitative methods. 'What' and 'how much' are answered through quantitative methods. This approach also relegates the use of qualitative methods to the exploratory (and thus initial) stage of research, to formulate propositions. This is the approach in most of the used of mixed qualitative and quantitative methods in the same research and this is the practice that sustains the subordination of qualitative methods to quantitative ones, especially in non-disciplinary fields like management. In the second column, I put those that (like me) do not believe that there's an independent social reality to be discovered by the eager social scientist. This is the standard use of qualitative research in anthropology and, to some extent, in sociology, which is often equated with the interpretative approach. One could think about using quantitative methods under this framework, but without really much thought, this does not seem very plausible because measuring things means categorizing them a priori, and this is incompatible with interpretivism (this second column).

Two questions arise from this:

1) How can we use quantitative methods to illuminate our qualitative and interpretive research. For myself, I have tried to used them to further explore patters, counting a posteriori, to use the dominant language
(numbers) to the service of the dominated. There is, of course, a political agenda here: to reduce the supremacy of quantitative / there's a reality out there' approach.

2) How does the use of qualit software such as NVIVO or N5 affect this?
Counting occurences of specific coding categories and bringing them into tables reify our observations?

I wonder....

From: Paul ten Have
Posted: Mon 4/02/2002

I agree with Gary Ivory that "qualitative research" and "interpretive research" are two different perspectives, although their may be an 'elective affinity of the latter for the former. But I also think that qualitative research can be characterized in broad terms that do not depend on any specific approach, for example along the following lines:

In his 'Constructing social research: the unity and diversity of method' (1994: 33 etc.), Charles Ragin has distinguished three major types of social research, based on their general goals and specific research strategies, qualitative, comparative and quantitative research.

Qualitative research is especially used to study what he calls ‘commonalities', i.e.
common properties, within a relatively small number of cases of which many aspects are taken into account. ‘Cases are examined intensively with techniques designed to facilitate the clarification of theoretical concepts and empirical categories.' In Ragin's version of comparative research (cf. also Ragin, 1987), the focus is on diversity and ‘a moderate number of cases is studied in a comprehensive manner, though in not as much detail as in most qualitative research'. It ‘most often focuses on configurations of similarities and differences across a limited number of cases.'

Quantitative research, finally, investigates the covariation within large data-sets, that is, a relatively small number of features of cases is studied across a large number of cases. So the focus is on ‘variables and relationships among variables in an effort to identify general patterns of covariation.' Elsewhere, Ragin (1994: 92) remarks: ‘Most quantitative data techniques are data condensers' and ‘qualitative methods, by contrast, are best understood as data enhancers.' The crucial feature of data treatment in qualitative research, I would say, is to ‘work up' one's research materials, to search for hidden meanings, non-obvious features, multiple interpretations, implied connotations, unheard voices. While quantitative research is focussed on summary characterizations and statistical explanations, qualitative research offers complex descriptions and tries to explicate webs of meaning.

'True qualitative research', then, would be those types that are focussed on 'complex descriptions' and the explication of 'webs of meaning'. And NVivo is apparently best suited for that kind of word.

'quasi-qualitative research', on the other hand, tends to depend on some kind of 'qualitative coding' of larger data sets. Threfore, in those types of research, one will most often strive for a relatively limited set of codes, that is one will use 'techniques (that) are data condensers', and the focus will be on qualiotative ‘variables and relationships among variables in an effort to identify general patterns of covariation.' For that type one could say that it tends to be 'quatitative without numbers'. For such types of research N4/N5 are apparently best suited.

Who is next?

From: Lioness Ayres
Posted: Mon 4/02/2002

Thank you, that is an entirely elegant distinction, and I like it very much. I think it is probably a bad idea to describe any kind of research as "quasi" any other kind - I spent years teaching from a research text that described qualitative research as a particular kind of "non-experimental design" - but of course I started that myself with the distinction between "truly" and "not truly" qualitative.

Have you read Renata Tesch and her ideas about de-contextualizing and re-contextualizing? Does this relate to Ragin at all?

From: Sarah Delaney
Posted: Mon 4/02/2002

continuums! continuums! every things a continuum!!

i do agree - there are different views of research per se and the philosophical framework informing it shifts according to which point on the continuuuuuuum you're at...