Chapter 3. Methodology: Contexts, Concepts, and Methodological Processes 3.1 Introduction and Overview of 1 Introduction and Overview of Chapter
3.5 Research Design and Process
3.5.2 Mixed method design
for this study, I will refer interchangeably to UKZN, as well as South African higher education in general.
UKZN is currently structured according to four Colleges, namely the College of Agriculture, Engineering and Sciences, the College of Health Sciences (including Medicine), the College of Humanities (including Education), and the College of Law and Management Studies (UKZN, n.d.). Admission to study as an undergraduate student at UKZN, as at other South African
universities, is based primarily on the matriculation score derived from the national matriculation examination results.35 Admission is however subject to redress considerations in selected
programmes such as various alternative access and bridging/foundation programmes, and in the College of Health Sciences. Due to extremely high demand for spaces in the College of Health Sciences, in addition to matriculation score, admission is stratified by race to accommodate quotas that are reflective of the demographics of the province within which the University is situated (UKZN, 2013b).
incorporates three dimensions. The three dimensions pertain to level of mixing in the study (i.e., full or partial), sequencing of data collection (i.e., concurrent or sequential), and the status or emphasis of quantitative or qualitative method/s in the overall design (i.e., equal or dominant).
Each of these three dimensions will now be presented in relation to Figure 3-1.
This mixed method study was categorised as fully mixed (i.e., the first dimension) for three reasons:
firstly, both quantitative and qualitative components were used when designing the research objectives and questions (i.e., profiling/prediction through quantitative methods, and exploration through qualitative methods) – see Figure 3-1, where A = quantitative research questions, B = qualitative research questions;
secondly, both qualitative and quantitative data was collected and analysed – see Figure 3-1, where C and D = quantitative data, and E and G = qualitative data;
thirdly, both qualitative and quantitative inferences were made – see Figure 3-1 (F, H, and I).
In contrast, while a partially mixed design makes use of both quantitative and qualitative methods, mixing would only take place at the level of interpretation and inference (Leech &
Onwuegbuzie, 2009). When considering Leech and Onwuegbuzie’s (2009) second typological Figure 3-1. Graphic representation of the mixed method design developed for the study
dimension (i.e., sequencing) for the current study, the quantitative data was collected first (see Figure 3-1, Phase 1 quantitative), this then leading into a second qualitative data production phase (see Figure 3-1, Phase 2 qualitative). This was therefore a sequential mixed method design. Finally, the qualitative data production process was more extensive than the quantitative process, and was also deemed to hold more value than the quantitative process in terms of its potential to explore the phenomenon under inquiry. Specifically, qualitative data was produced via three modes (see Figure 3-1, Ei, Eii, and Eiii), while quantitative data was produced through setting up one database (i.e., the quantitative sample). As a result, the third dimension denoted that this was a study where one of the data modes (i.e., qualitative) had dominant status, this being indicated by capitalisation and increased font size in the graphic representation. In
addition, the third dimension (status or emphasis in mixed method studies) was also illustrated in Figure 3-1 via the smaller quantitative phase of the study being “nested” within the larger
qualitative phase. Therefore, although sequencing of the study positioned the quantitative data collection first, level of emphasis positioned the qualitative data with higher status.
As mentioned above, the fully mixed sequential dominant status design was adapted to include an iterative and recursive component (Nastasi et al., 2007) and to reflect the dialectical stance adopted in this study. The iterative and dialectical stance was most evident between three aspects of the research design, these being structured to enhance dialectical tensions between
quantitative and qualitative modes. Firstly, the dialectical and iterative nature of the study was evident in the design of the research questions (specifically the second and third research questions were positioned in dialectical tension with first research question). This is also
illustrated graphically in Figure 3-1 via the dotted bi-directional arrow between points A and B.
Secondly the dialectical and iterative nature of the study is also evident in the analysis and interpretation of the quantitative data. This will become explicit when the data production process is described in more detail below, but is currently represented in Figure 3-1 via the unidirectional dotted arrow from D to Ei, and the bi-directional dotted arrow between Ei and F.
Thirdly, as is evident in most mixed methods designs (Leech & Onwuegbuzie, 2009), the inferences drawn for this study were derived from both quantitative and qualitative modes (see Figure 3-1, bidirectional dotted arrows between F and I, and H and I). The dialectical and
iterative nature of the study was also manifest within the qualitative data analysis and
interpretation processes (see Figure 3-1, bi-directional arrows between Ei, Eii, Eiii, G, and H).
The bi-directional arrows between H and G, and G and E, for example, pertained to member checking of descriptive accounts, which in turn added to the accuracy or descriptive validity of the data (Maxwell, 2002).
Having firstly described the rationale for selecting UKZN as the institutional case study, and secondly, the mixed method design used for this study, subsection 3.5.3 will involve a detailed account of Phase 1 (the quantitative phase) of the study. This account will include the process involved in setting up the database (quantitative sample), as well as the data analysis.