CHAPTER 3 RESEARCH DESIGN AND METHODOLOGY
3.4 TRUSTWORTHINESS OF THE STUDY
In assessing the rigour of this qualitative study, the usual criteria of reliability and validity used to evaluate positivist research were not wholly appropriate for a qualitative study (Denzin &
Lincoln, 2011; Rule & John, 2011). The concepts of reliability, validity and rigour were therefore referred to as trustworthiness, credibility, transferability and confirmability in this study, as a qualitative study does not lend itself to these concepts as easily as a quantitative study would. Guba (1981) emphasized trustworthiness as the overarching criterion, which in the interpretive paradigm comprises the constructs credibility and transferability. Credibility is the ability of the researcher to take full account of complex data patterns and produce plausible findings (Guba 1981, p. 84). Confidence can thus be placed in the authenticity and plausibility of the data and analysis (Guba & Lincoln, 2005; Koul, 2008). Credibility can also be enhanced through the confirmation of conclusions by the participants and there should be congruency between the findings and observations (Merriam, 1991). Koul (2008) reiterated that the credibility of the study is also heightened when other researchers and readers, because of the accurate description, encounter and recognise the same experience.
Transferability can be achieved in detailed descriptions of the data as well as the context, so that the reader can make comparisons to other contexts (Rule & John, 2011). My use of purposive sampling can also serve to promote transferability (Guba, 1981). Confirmability centres on the minimisation of bias on the part of the researcher to ensure that the data are presented in an objective and neutral manner (Guba, 1981). In terms of confirmability I undertook to act in good faith throughout the research process and did not intentionally allow my background, experience and personal values to influence the way in which the research was undertaken and the results thereof. This also ensured that I was able to distance myself and
58
ensured that my personal experience did not influence my ability to see all the possibilities presented by the data.
To ensure the trustworthiness of this study the process of open coding was used to analyse the selected textbooks. The process of open coding involves dismantling, examination, identification of concepts and the categorisation of data (Strauss & Corbin, 1990). Line by line coding forces the analyst to verify and saturate categories, thus minimising the chance of missing important categories and ensuring the grounding of data categories beyond mere impressionism (Glaser & Barney, 1992). This also facilitates the breaking down of large sections of textual data into themes and relationships (Mouton, 2001), allowing for the categorisation of data and identification of the discourses, which formed part of the objectives of this study. The result should be a rich, dense theory assuring that nothing had been omitted.
Henning (2007) suggested that the process of open coding should follow two broad principles.
First the researcher should read the chapters to get an overall impression of the content. Price (2005) called this “reading with the text” to try and understand the writers’ positioning. Then the chapters should be re-read in a critical manner to identify “units of meaning” (Henning, 2007), raising questions about these meanings and establishing how they could be constructed differently. In the second reading Price (2005, p. 7) advised the analyst to read ‘against’ the text using CDA to oppose the “text’s apparent naturalness”. In this reading the analyst asks questions on the positioning of the text, whose interests are being promoted, whose interests are being ignored, and the possible consequences of this positioning. Here analysts are actually asking how these texts are used to reproduce or transform the status quo in society (Janks, 1997).
During the process of open coding I allocated codes to the different ‘units of meaning’, which in this study revealed two major units: the first was the inward-looking economic policies of protectionism and import substitution; and the second was the outward-looking economic policies comprising globalisation, free trade, international trade and export promotion. On a methodological note I must add that often the textbooks combined the policies of free trade and globalisation, for example, on page 95 of textbook D the section begins with these words:
Protectionism and free trade (globalisation)
These codes were then categorised into the discourses to be discussed as the findings of this study. It is at this point that the discourses of globalisation were identified. This process was
59
facilitated through the use of particular words and phrases relevant to that discourse. For example, in the analysis of data the identification of ‘financescapes’ was facilitated through the use of certain keywords, such as capital, trade, investment and profit. Lastly the discourses were then identified into the two themes showing the text’s stance, whether in favour of or biased against the economic policies. This process of open coding facilitated the generation of meanings and ideas and also deepened my understanding of the data that yielded multiple meanings. I used the analytical framework to guide this process.
The use of the analytical framework was labour-intensive. Even though the data sample involved a limited amount of extracts dealing with outward-looking economic policies and inward-looking economic policies, the detailed examination of words, clauses, sentences and paragraphs was intense because of utilisation of the tools of the framework. This was because, in order to explore the representation of globalisation, I had to compare the representation of other economic policies (inward-looking economic policies) against the representation of globalisation and economic policies linked to globalisation (outward-looking economic policies). Subsequently, within each major discourse the data analysis is separated into two themes, that of the romanticising of outward-looking economic policies (free trade, globalisation and export promotion) and that of the inward-looking policies (protectionism and import substitution) being portrayed as deficient. However, these themes often overlapped within each discourse because of the comparative nature of the presentation in the textbooks.
Moreover, the interconnectedness of the discourses sometimes did not allow for an easy separation and for them to be treated as discrete discourses.
To add further trustworthiness, ‘member checks’ in the persons of independent coders, my supervisor and the PhD cohort programme were undertaken. I also presented a paper based on this study at the SAERA (South African Education Research Association) conference in Cape Town in 2016, where my presentation was interrogated by other CDA scholars. In addition, during the time period of this study two attempts were made to publish articles based on this study. This required intense scrutiny to ensure that the literature review, research methodology and findings of the papers were rigorous and that the data samples were not misinterpreted to serve a particular political bias.
60