Some points on mixed methods research Mike Griffiths m.griffiths@gold.ac.uk
Questions Is it legitimate to combine quantitative and qualitative research? In what ways have people combined quantitative and qualitative research? Is a mixed method approach always useful?
Q1: The arguments against multistrategy research* Embedded methods epistemological and ontological commitments Data gathering method is a commitment to a philosophical position Paradigm argument “conceives of quantitative and qualitative research as paradigms ... In which epistemological assumptions, values and methods are inextricably linked” (p. 453) *Bryman, 2004
Bryman’s (2004) argument for MMR Recognise different epistemological and ontological assumptions But “the connections are not treated as fixed and ineluctable” “A research method from one research strategy is viewed as capable of being pressed into the service of another” Indeed, there may not even be a ‘leading’ strategy in a given piece of research (p. 454)
Common contrasts: quantitative and qualitative research
Philosophical differences (Bryman, 2004) Table 1.1, page 20
Ontological differences: my suggestions
But… These are just guides to what approach might best fit a given question Different approaches might suit different aspects, or phases, of a research question.
Q2. Classifying approaches to MMR Hammersley (1996) Triangulation Facilitation (either way round) Complementarity Morgan (1998) Bryman, 2004
Triangulation Can apply within one approach Cross-checking results Can be planned or unplanned Possibility of failing to corroborate Trust one of the results rather than the other? (Newby, 1977, cited in Bryman: “instinctively trusted” participant observation above surveys) Try to work out the reason? (Deacon et al., 98, cited in Bryman: conflicting ideas about how well the press report research findings)
Qualitative research facilitates quantitative Providing hypotheses E.g. Phelan (1987): incest in different kinds of family Aiding measurement E.g. construction of a questionnaire Solving a puzzle (unexpected findings)
Quantitative research facilitates qualitative Selecting a sample, e.g. identifying suitable people to interview Filling gaps, e.g. structured interview or questionnaire to find information not available from observation
Complementarity Information not available through one method or other Quantitative for ‘static’, qualitative for ‘processual’ Quantitative for ‘macro’, qualitative for ‘micro’ Researchers’ vs participants’ perspective Behaviour vs attitudes
Is a mixed method approach necessarily best? Bryman’s (2004) conclusions: Not intrinsically superior Must be well designed Must be appropriate High demands on resources Needs skills in both approaches