124     QUANTITY ANP QUALITY IN SOCIAL RESEARCH
by various practitioners who have not had an epistemological axe to grind. Writers like Lupton (1963), Gans (1962), and Skolnick (1966), all of whom have written much admired monographs deriving from the use of such methods, seem to have exhibited few, if any, philosophical pretensions in their justifications for the use of qualitative research. These researchers were able to produce highly regarded ethnographic studies without recourse to the programmatic statements surrounding qualitative research. This very fact invites a questioning of the role of programmatic statements in relation to the pursuit of good social research. Rather, these researchers were concerned to get close to the people they were studying, to allow for the possibility of novel findings, and to elucidate their findings from the perspective of the people they studied. While the last of these three concerns is invariably taken as a keynote of the epistemological substructure of qualitative research, a preoccupation with meaning and subjects' perspectives is not exclusive to the qualitative tradition. Further, the methods with which qualitative research is associated have often been chosen on technical grounds rather than epistemological grounds.
The lack of a definitive link between broad epistemological positions and methods has also been suggested by Snizek (1976), who analysed 1,434 articles in sociological journals covering the period 1950 to 1970. Snizek was concerned to find out whether there was a connection between the three paradigms which Ritzer (1975) had suggested underpin sociology and the methods with which they are associated. If these three basic approaches to sociology really arc paradigms (T. S. Kuhn, !970), one would anticipate a link between the endorsement of the epistemology with which each is associated and the methods of research used. Two of the paradigms - the 'social factisť and 'social definitionist' orientations - correspond to quantitative and qualitative research respectively to a fair degree. However, Snizek was unable to discern a clear pattern which linked the general orientation of each paradigm with the methods of investigation employed.
The alternative position is to suggest that there ought to be a connection between epistemological positions and methods of data collection. This view would imply that researchers should be much more sensitive to the wider epistemological context of methods of data collection and that they are not neutral technical devices to be deployed under a variety of auspices. Choosing to conduct a survey or an ethnographic study would mean accepting a package of views about social reality and how it ought to be studied. Accordingly, it might be argued that researchers who claim to study subjects' views of the world with a survey (e.g. Coldthorpe et al., 1968) are misguided, since they should have chosen a method more suited to
THE DEB A TE     125
this perspective, like unstructured interviewing. It is not at all clear from the various writings on the debate about quantitative and qualitative research that the view exists that there ought to be a recognition of a mutual interdependence of epistemology and method (as against a view that there is such a connection). The problem with the 'ought' view is that it fails to recognize that a whole cluster of considerations are likely to impinge on decisions about methods of data collection. In particular, the investigator's judgements regarding the technical viability of a method in relation to a particular problem will be important, as the technical version of the debate about the two research traditions implies.
Methods are probably much more autonomous than many commentators (particularly those who espouse the epistemological versions of the debate) acknowledge. They can be used in a variety of contexts and with an assortment of purposes in mind. Indeed, the very fact that many qualitative researchers instil an element of quantitative data collection into their investigations underlines this point to a certain degree. Similar points can be made in relation to the connection between broad theories and methods. As pointed out in Chapter 2, Piatt (1986) has strongly questioned the supposed link between functionaüsm and the survey. Similarly, while some writers have found an affinity between a Marxist perspective and qualitative research (e.g. Sharp and Green, 1975; Willis, 1977), others have preferred the methods of quantitative research for the empirical elucidation of concepts associated with this theoretical perspective (e.g. Wright and Pcrrone, 1977). Symbolic interac-tionism, while typically associated with participant observation (Rock, 1979), is not universally identified with qualitative research and an anti-positivist epistemology: M. H. Kuhn (1964) used the techniques and research strategies of quantitative research in his attempt to use symbolic interactionist notions; many studies conducted within this theoretical tradition make substantial use of standard survey techniques alongside participant observation; and there is even a questioning of whether G. H. Mead's (1934) writings lead in a direction which is antithetical to the application of the methods of the natural sciences (McPhail and Rexroat, 1979). The tendency to associate particular methods with particular epistemological positions is little more than a convention (which took root in the 1960s), but which has little to recommend it, either as a description of the research process or as a prescriptive view of how research ought to be done.
In comparison with the rather stark contrasts between quantitative and qualitative research which have permeated the pages thus far, the next chapter examines some of the ways in which the two traditions may be used in tandem. The epistemological version
126    QUANTITY AND QUALITY IN SOCIAL RESEARCH
of the debate does not readily admit a blending of quantitative and qualitative research since the two traditions are deemed to represent highly contrasting views about how social reality should be studied. The technical version of the debate much more readily accommodates a marriage of the two since it acknowledges the respective strengths and weaknesses of the two approaches as methods of data collection. Many writers shuttle uneasily back and forth between the two ways of thinking about the two traditions. It is little wonder that confusion ensues when there is a lack of clarity about what quantitative and qualitative research are. In this context, the view of a leading writer on the ethnography of schooling is instructive:
It is not surprising that some work called 'ethnography' is marked by obscurity of purpose, lax relationships between concepts and observation, indifferent or absent conceptual structure and theory, weak implementation of research method, confusion about whether there should be hypotheses and, if so, how they should be tested, confusion over whether quantitative methods can be relevant . . . and so forth. (Spindler, 1982, p. 2)
Precisely because many qualitative researchers have failed to sort out whether the style of research to which they adhere is an cpistcmological or a technical position, it is possible for such confusion to reign. However, when quantitative and qualitative research are jointly pursued, much more complete accounts of social reality can ensue, as many of the examples cited in the next chapter imply.
Notes
1    Symbolic interactionism, by contrast, is a theoretical position developed largely within the social sciences, but which has its roots in an cpistcmological position, namely, pragmatism.
2    Frequently, discussion in the literature of such issues takes the form of evaluating qualitative research in terms of its validity and reliability (e.g. LeCompie and Goetz, 1982). I have resisted such an approach in this book, because I feel that it imposes a cluster of standards upon qualitative research which to a large extent are more relevant to the quantitative tradition, within which such terms were originally developed.
6
Combining Quantitative and Qualitative Research
The rather partisan, either/or tenor of the debate about quantitative and qualitative research may appear somewhat bizarre to an outsider, for whom the obvious way forward is likely to be a-fusion of the two approaches so that their respective strengths might be reaped. The technical version of the debate more readily allows this solution to be accommodated because it is much less wedded than the epistemological version to a view that the two traditions reflect antagonistic views about how the social sciences ought to be conducted. In this chapter, the focal concern will be the ways in which the methods associated with quantitative and qualitative research can be, and have been, combined. As noted in Chapter 3, there arc examples of investigations carried out by investigators who locate their work largely within the tradition of qualitative research, but who have used survey procedures in tandem with participant observation (e.g. Woods, 1979; Ball, 1981). Such research will be employed as an example of the combination of quantitative and qualitative research, because the chief concern of the present chapter is with the methods with which each is associated.
The focus on methods of investigation should not lose sight of the significance of a distinction between quantitative and qualitative data. For example, some of the findings associated with an ethnographic study may be presented in a quantified form. In their research on the de-skilling of clerical work, Crompton and Jones (1988) collected much detailed qualitative information, in the form of verbatim reports, on the work of their respondents. In spite of considerable reservations about coding these data, they aggregated people's accounts of their work in terms of the amounts of control they were able lo exercise in their work. Even among qualitative researchers who prefer to resist such temptations, the use of quasi-quantitative terms like 'many', 'frequently', 'some', and the like, is common (e.g. Gans, 1982, p. 408). Further, survey researchers provide the occasional verbatim quotation from an interview, or one or two case examples of respondents who exemplify   a  particular pattern.   Sometimes,  the  reporting  of