Mixing research methods—a conversation towards a negotiated account

Alan Bryman (2007: 22) did a content analysis of 232 articles, published between 1994 and 2003, with 'mixed' research methods. He found that only 18% of the articles succeeded to mix (integrate) qualitative and quantitative methods. He remarks (p. 8):

As result of the efforts of methodologists and researchers, there is nowadays considerable understanding of a variety of issues, such a various ways in which quantitative and qualitative research can be mixed … However, at the same time, the fundamental issue of the degree to which mixed methods researchers genuinely integrate their findings has not been addressed to a significant extent. In other words, how far do mixed methods researchers analyze, interpret, and write up their research in such a way that the qualitative and quantitative components are mutually illuminating? … "genuinely integrate[d]".

Bryman (2007: 9) conducted semi-structured interviews with 20 UK social scientists about "the researcher's views on and practices in relation to the integration of quantitative and qualitative research" and asked during the interviews (p. 10) "whether they found it straightforward to combine their quantitative and qualitative findings". Nine barriers to effective integration emerged:

1. The two sets of findings often documented for different audiences
2. Because of methodological preferences, researchers tend to emphasise what they comfortable with.
3. The structure of research projects sometimes complicates integration.
4. The timelines of either the qualitative or quantitative may be such that the one ends sooner than the other, if not taking place prior to.
5. In addition to No 2, skills specialisation or incapacities may impede integration.
6. The difference in the nature of data results in different ways of writing the results up.
7. Bridging ontological divides are sometimes challenging, for example, objectivist versus constructivist or discursive accounts.
8. Preference of some journals present publication challenges, if a particular methodological bias prevails it is often difficult to get an article accepted that does not fit.
9. "[R]elative absence of well-known exemplars makes it difficult for researchers to draw upon 'best practice' when it comes to combining findings" (p. 19).

Bryman (2007: 19) grouped the first eigth barriers into three types, namely:

Barriers related to intrinsic research aspects	Wider institutional context of mixed method research	Skills and preference of researchers
Structure of research projects (3) Bridging ontological divides (7) Role of timelines (4)	Different audiences (1) Publication issues (8)	Methodological preferences (2) Skills specialisms (5) Nature of the data (6)

The absence of good exemplars, relate to all three types of barriers.

Bryman argues for genuine integration, where qualitative and quantitative findings are mutually informative and talk to each other—a conversation—to construct an negotiated account. The triangulation metaphor has been a hindering factor; it should not be just a matter of testing data against each other, but "about forging an overall or negotiated account of the findings that brings together both components of the conversation or debate" (p. 21).

Bryman, A. 2007. Barriers to integrating quantitative and qualitative research. Journal of Mixed Method Research, 1(1), pp 8-22, January 2007.