CHAPTER FOUR METHODOLOGY
4.4 RATIONALE BEHIND USE OF MIXED METHODS
when to use each. Both approaches are seen as complimentary but still as representing distinct universes (Rossman & Wilson, 1985). Pragmatism contends for integration of methods, and there is a growing impetus asserting that there are no good reasons for researchers to fear forging ahead with “what works” (Howe, 1988).
The present study concurs with the pragmatist approach. Denzin (1970, p.119) commented that “a false dichotomy exists between these two types of data”, and it should be noted that quantitative methods are not necessarily positivistic nor are qualitative ones necessarily phenomenological. Sechrest & Sidani (1995) claimed that the problems that are perceived to exist between the proponents of qualitative and quantitative appear to stem from misunderstandings and misstatements of the positions involved. Empirical inductivists and interpretists differ in their philosophical assumptions, and consequently the ways they collect and analyse data. Yet their ultimate tasks and aims are the same: describe their data, construct explanatory arguments from their data, and speculate about why the outcomes they observed happened as they did. Bazeley (2004, p.2) pointed out that “If one uses numbers, interpretation is still involved. If one’s data are texts, counting may still be appropriate”. Onwuegbuzie & Leech (2005) commented that mono-method research is the biggest threat to the advancement of the social sciences, extending this by commenting that as long as we stay polarised in research, how can we expect stakeholders who rely on our research findings to take our work seriously. They believe that the debate between the two fields has been divisive and counterproductive.
In the literature, there are extensive arguments clarifying how mixing methods allows improvement of the accuracy of conclusions by relying on data from more than one method. Among the rationale for integration of mixed methods includes:
Corroboration – It brings together data collected through more than one method to see if there is convergence in the findings.
Elaboration – It provides richness and detail, expanding understanding of the phenomenon through refinement and development. It can provide strength to an argument and provide a different perspective on the same phenomenon. It can be paralleled to “fleshing things out” or “putting meat on the bones”.
Clarification of results from one method with the results from the other method is possible. Mixed methods increase interpretability, meaningfulness, and validity of constructs and inquiry results by both capitalizing on inherent method strengths and counteracting inherent biases in methods (Greene, Caracelli & Graham, 1989).
Initiation (holistic triangulation) – Aims to uncover paradox and contradiction, and possibly lead to a substantial alteration in the overall perspective of the problem. Divergent findings may set up dissonance, doubt and ambiguity leading to further analysis or even a recasting of the entire research question. Jick (1979, 603) commented “it is here that qualitative methods, in particular, can play an especially important role by eliciting data and suggesting conclusions to which other methods would be blind”. Hence there is a search for the confirmatory and the provocative.
Development – This involves using the results of one study to help to inform or develop the other method, which helps to increase the validity of the constructs and inquiry by capitalising on inherent method strengths.
Expansion - Mixed methods can increase the breadth and depth of inquiry results and interpretation by analysing them from different perspectives
Weinholtz, Kacer & Rocklin (1995) asserted another advantage of mixing methods, namely saving quantitative data by using qualitative data. Weinholtz (1995) described the “salvaging power” of qualitative methods in reference to their ability to ensure some return on an investment that might otherwise be partially or completely lost.
The present study followed a “Triangulation design model” (Creswell, et al 2004), in which neither method is given priority, both are of equal importance. Morgan (1998) gives a good description of how the level of priority and sequence of methods should be considered in mixed methodology research design. Both methods were implemented and analysed concurrently but separately. Integration of data occurred following individual analyses, and findings were considered in relation to one another, and conclusions will be formed from this, as shown in Figure 7. Integrated results encompassed both confirmatory/supportive and conflicting findings.
IMPLEMENTATION
ANALYSIS
INTERPRETATION
Figure 7: Use of mixed methods in the present study.
The many advantages of adopting a mixed methodology that were particularly relevant to the aims of the research can be viewed in Figure 8. It was argued that the objectives outlined earlier were best answered using mixed methods. Mono-method was enhanced, elaborated on, weaknesses were decreased by the addition and juxtaposition of an alternative method. Qualitative methods augmented quantitative components by giving a sense of completeness and plentiful illustrations. Mixed methods honed in on unexpected results more effectively than either method in isolation.
Semi-structured interviews Outcome measures
Content Analysis Significance testing
Findings
Conclusions
Figure 8: Benefits of adopting mixed methods in the present study.
Randomised controlled trials are frequently considered the gold standard of research methodology, excluding the meta-analytic approach (Silagy, Jewell & Mant, 1994) as they can most effectively isolate treatment effects, predict effectiveness of treatment, have a control group for comparison, and can be statistically analysed. However, such a design for the present study was both inappropriate and unfeasible.
The throughput of clients completing a DBT programme within the study data collection would still be too small, even with the involvement of several services, to justify the implementation of such an approach. With the limited number of participants available, the present study utilised a mixed methods design with two phases, as this approach most effectively tackled the desired objectives. A mixed methods design was applied to take an in-depth analysis of the impact of DBT on clients. Although admittedly the approach has some limitations its benefits outweigh such a drawback.
Why use mixed methods in this study?
Triangulation / initiation–
Process– qualitative methods gave insight into the process involved in change