Any research that attempts to formulate a conclusion from a particular path of inquiry uses aspects of scientific method. Although its presentation and interpretation vary across the field, but overall premises are unaltered (Edmonds et al., 2014). Mixed method research is becoming increasingly articulated to research practice and recognised as the primary research paradigm. In this context, the present paper critically analyses the chosen research method, in this case, mixed design in the quasi-experimental study (Shadish et al., 2002) applied in a broad range of studies from natural as well as social sciences, to develop insights into and strategies for achieving this integration.
Mixed Methods and Mixed Design in Quasi-Experimental Study
Mixed research, as its name indicates, involves the mixing of quantitative and qualitative methods or paradigm characteristics (collecting, analysing and interpreting) in a single study or series of studies that investigate the same underlying paradigm (Leech & Onwuegbuzie, 2009). Quasi-experimental research (QER) designs like experimental design test causal hypothesis and this method are applied in the research when it is not possible to randomise individuals or groups to treatment or control group. Generally, such research can be observed in ex-post impact evaluation designs.
Usage of Mixed Methods & Potential advantages of Mixed Design in Quasi-Experimental Study
In general, MMR is of three types, i.e., exploratory sequential, explanatory sequential and convergent and other advanced designs (concurrent, parallel) (Creswell, 2015) that embed QER. However, the choice of design is based on the research question being posed along with the timing of collected data, the relative weight of the quantitative and qualitative to answer the research question (Creswell & Clark, 2011) and approach adopted to mix the two data sets. Combing methods within a single study, surpasses the advantages of having a singular methodology, by consolidating the development of triangulation (Robinson & Mendelson, 2012) Mixed design offers advantages as it seeks convergence and corroboration of findings, expansion i.e., seek to expand the range and breadth of inquiry, development, and initiation (Greene, 1989). MMR serves as an umbrella for several sub-questions, as it calls for separate sets of qualitative and quantitative followed by mixed method (Tashakkori & Teddlie, 2003). As emphasised, a qualitative design may succeed a quantitative to explore potential impacts, thereby it offers a rich understanding of the context and can strengthen each result. Secondly, explaining, enriching, refuting and confirming findings from one approach can complement the other through the concept of triangulation.
Examples of Mixed methods used in Literature
MMR has been applied in many types of research. Specifically it provides right insights, strengthening the policy relevance of impact assessments. For instance, Buckley, (2015) examined the effect of policy instruments contributing to indigenous firm growth using sequential mixed-method quasi-experimental design. The resulting analyses allowed policy recommendations thereby creating an ethos of ‘evidence-based’ policymaking. Further MMR research has also been widely used in clinical research and public health interventions in real-world settings when evidence-based interventions are known. Kong et al., (2018) employed a sequential hybrid model, with an embedded quasi-experimental model where the qualitative inquiry was conducted in the first stage followed by a quantitative evaluation to design an architectural intervention and determine its effectiveness in enhancing knowledge, attitude and behaviour among school students. The application of MMR design enabled the researchers to create highly robust research by investigating multifaced architectural issues that promoted innovation with the design and planning. Weinhardt et al., (2017) evaluated the impact of multi-level economic and food security intervention on health outcomes and HIV vulnerability. Authors compared who received the SAFE intervention, against those who did not by collecting quantitative data in baseline, 18-month and 36-month follow-up. The second sample was from a random sample community where interviewing was done in all households who were not direct participants in the SAFE and the control groups. Finally, in-depth qualitative interviews and focus groups were conducted to understand SAFE participant’s experiences. The intervention package designed had positive effects. Brenner et al., (2014) had an objective to increase the number of deliveries and had conducted a quality assessment using some indicators. In a second step, selected facilities that met quality criteria were selected for intervention. Finally, a qualitative study was conducted to understand the perceptions on quality care and service utilization. The design permits the assessment of organisational and behaviour changes.
Challenges of Mixed Methods
Although mixed method design offers advantages, still validity (internal and external) of the research been questioned and well documented in the literature. Studies had identified history bias, selection bias, maturation bias, lack of blinding, differential dropout, and variability in interactive effects (White & Sabarwal, 2014). Many researchers might not have the expertise to conduct MMR, challenges in analysing quantitative and qualitative. They also need to work for a diverse team on the identification of underlying assumption that they can bring to the research question. It is therefore plagued by the problems of representation, integration and legitimation. Despite the above challenges, still, MMR in quasi-design has been widely used in the study of impact assessment.
Comment here