4.3.1 Mixed-Methods Research Worldview
While purely quantitative research often subscribes to the positivist/post-positivist epistemology and qualitative research may follow a plurality of worldviews, mixed-methods research is best served by a pragmatic approach to the data. Pragmatism works from the belief that the effects are of primary import, rather than the causes often investigated in a post-positivist worldview. To summarize the worldview, both qualitative and quantitative models are important for understanding data, but only insofar as they are able to consistently and predictably produce a desired practical outcome.
While other epistemologies may indeed allow for both qualitative and quantitative data gathering and interpretation, mixed-methods are often focused on application as opposed to a more purely research oriented framework. In taking on mixed-methods research, the researcher is often hoping to work in a flexible fashion and interpret results as they arise as need be. For this end, a pragmatic worldview is often useful.
Pragmatism is fundamentally concerned with the idea of what works, and secondarily how and why it works. In looking at how phenomena exist and interact in the real world, mixed-methods are inherently concerned with practicality. Through gathering both quantitative data for empirical verification and qualitative data for subjective interpretation, the goal of this framework is to provide both understanding and actionable points, and thus fulfill the needs of practitioners.
In educational research, this worldview places emphasis on the outcomes while documenting both measurable and subjective qualities influencing learning and instruction. It is the approach to both theory and practice advocated by Dewey (1948), as well as underlying much of the work by Brophy (2004; 2005). This approach matches well with the methods used by mixed-methods researchers as it provides a flexible way to approach data.
4.3.2 Mixed Methods Practices
While previous generations of researchers have stated that qualitative and quantitative methodologies are inherently incompatible (e.g., Guba, 1990), current theorists have countered that the two methodologies offer more similarities than differences (e.g., Johnson
& Onwuegbuzie, 2004). Less polar views on the qualitative side argue that qualitative inquiry can offer ways to clarify and enrich quantitative data (Hesse-Biber, 2010a; 2010b), and thus has a place in presenting the human experience associated with empirical findings. Indeed, as
noted previously, quantitative research requires good theory and qualitative controls even in the experimental hard sciences (Shadish, Cook, & Campbell, 2002).
Within social sciences such as education, both quantitative and qualitative methodologies are inherently interested in providing well-grounded hypotheses and their answers regarding individuals and groups, especially with regard to school context, learning, and development (Onwuegbuzie & Leech, 2005). Qualitative observation forms the foundation and basis for theory, which then can be used as a way to generate further hypotheses to generalize on the data (Creswell, 2008). As discussed, quantitative decision processes are further inherently qualitative, from the generation of theoretical observation to methods for isolating observations to setting of objective cutoff points for alpha scores and fit indices.
In mixed-methods procedures, four main factors influence the overall shape of the research: weighting of qualitative and quantitative focus; timing of data gathering; mixing of data interpretation; and the role of theory. In designing a research project, these 4 factors must be clarified in order to ensure effective analysis. Researchers must first define the primary objective as either qualitative or quantitative, followed by procedures for data gathering. Figure 4.2 displays the possible design tracks for a research project. The research question defines the role of a specific methodology in the project, either primarily qualitative or quantitative, or equal weight on the two. In the notation of the project design (Creswell, 2008; Johnson & Onwuegbuzie, 2004), this is often shown using all capitals for the dominant methodology (“QUAL,” “QUANT”) or all lower for the less dominant paradigm (“qual,”
“quant”).
Figure 4.2. Possible pathways for monomethod and mixed-method investigation. From Johnson &
Onwuegbuzie, 2004.
Next, the timing of the data gathering must be resolved. Projects within a mixed-method design may be sequential, concurrent, or embedded. In sequential studies, one type of data is gathered followed by another type, and are usually shown using an “→” to denote the order of events. In concurrent studies, two different types of data are collected simultaneously, potentially from differing sources, and are generally shown using “+” to denote simultaneity. Embedded designs gather both types of data simultaneously from the same source, usually denoted by stacking the two on top of one another. Sequential designs allow for follow-up to deepen knowledge, while concurrent methods allow researchers to nest different types of questions within a larger collection of data (Creswell, 2008). Figure 4.3 illustrates how both the dominant paradigm and sequence of data gathering may be documented.
The final consideration for research is the role of theory. While many qualitative paradigms traditionally do not work with theory, many have recently come to accept the role of pre-existing theory for generating hypotheses and interpreting events (Creswell, 2008). On
the opposite side, good quantitative research in the social sciences often requires theory for creating instruments and generating hypotheses in line with previous research (Kline, 2009).
While the generation of new theory and creation of new instruments based on observation may not require a background, the same approach may be taken with the application of a theory to a new context. For this purpose, researchers must clarify the theoretical position taken, be it strongly theoretically oriented or oriented towards generating a theory, in order to establish their research orientation for the reader.
Sequential research designs
QUAL → quant (Primarily qualitative with quantitative followup) QUANT → qual (primarily quantitative study with qualitative followup) Concurrent research designs
QUAL + QUANT (equal focus with data gathered simultaneously)
qual + QUANT (primarily quantitative study with simultaneous qualitative data)
QUAL
quant (primarily qualitative study with embedded quantitative data)
qual
QUANT (primarily quantitative study with embedded qualitative data)
Figure 4.3. Research method design documentation. From Creswell, 2008.
4.3.3 Theory in Mixed-Methods Research
As discussed in the previous section, theory may potentially have numerous applications in mixed-methods research. In order to apply these theories to both the qualitative and quantitative data, a transformative mixed-methods approach to the interpretation and application of data is necessary. An underlying theory guides what research questions to ask, how observations are to be taken, and how to approach the interpretation. In declaring a theoretical perspective, the researcher is clarifying any pre-existing biases that may be held by virtue of utilizing this research method.
In some instances, social science theory may be a particular advocacy worldview or ideology, such as feminism or internationalization, but this is not always necessary. As many psychological and educational theories carry with them corollaries and sub-theories, these may also allow researchers to deepen their understanding of the world and document phenomena using a specific lens. In most cases, the theoretical framework provides some access to the methods, and the theory is a stronger guide to analysis than the methods themselves (Creswell, 2008, p. 212).
For the purposes of applying theory to practice, mixed methods research offers the greatest chance of capturing both a valid empirical framework while documenting classroom events of clear relevance to elementary elementary foreign language teachers. By approaching data from both qualitative and quantitative perspectives, mixed-methods allow researchers to adopt a flexible approach, borrowing the concept of “best of all, worst of none” (Page, 2012).