3. CHAPTER THREE: RESEARCH METHODOLOGY
3.6. Data Analysis: Thematic Analysis
Once the interviews with the participants were completed the recordings were then transcribed by a professional transcriber who adhered to the confidentiality agreement (See Appendix G). The transcripts were then analysed by employing the method of thematic analysis.
The analysis of the qualitative data is to first examine the data for themes within each interview, and then secondly to conduct an across-interview analysis to identify important themes about psychotherapist development. Braun and Clarke (2006) indicate that thematic analysis can be implemented in studies that are held in a social constructionist paradigm. With this in mind, thematic analysis was undertaken by applying Ryan and Burnard’s guidelines (2003). As such the qualitative data were analysed to achieve “thematic description”
(Sandelowski & Barroso, 2003, p. 913). Analysis of qualitative data needs to utilise “explicit, systematic, and reproducible methods” (Greenhalgh & Taylor, 1997, p. 740).
The broad aim of the qualitative approach is to analytically work through texts by dismantling the text, for example from transcribed interviews, into smaller units of meaning and deriving description of the phenomenon from the identified meaning units (Sparker, 2005). Thematic analysis, in the current study, was the method of choice to analyse the huge corpus of data as it is regarded as a flexible method which is able to hold complexity and in the process provide a thick description of the data (Braun & Clarke, 2006). Research methods are “living entities” (p. 399) and as such qualitative research at its best is an organic process.
Both deductive and inductive thematic analysis were undertaken within the current study. Inductive thematic analysis is often used where there are few prior studies in the field and the themes and categories are derived directly from the data (Hsieh & Shannon, 2005), this was important as few studies have been conducted in South Africa, utilising Skovholt and Rønnestad’s (1993, 1995, 2003a, 2003b, 2003c) international model of psychotherapist
development, in order to be sensitive to possible unique findings in a new context. In contrast, a deductive approach in analysing the collected data was also used within the current study in order to compare the findings with similar studies conducted internationally.
To be illuminative, analysis of research data needs to be connected to the originating context. After all, social constructionism has prioritised that each individual is woven into their
76 context (Yang & Gergen, 2012). This has serious implications for the current study for the
contextual frames from which the psychotherapists practice weaves their individual
developmental processes. The researcher’s understanding of this larger context contributes needs to constructing a meaningful construction of the phenomena whereby the data can answer the research question posed (Downe-Wamboldt, 1992; Krippendorff, 2004). Thematic analysis offers both a systematic and meaningful analysis of the data within a specific context (Loffe & Yardley, 2004). This qualitative study makes possible a contextualised in-depth understanding of the developmental experiences of a small number of South African psychologists as
psychotherapists.
Tesch’s (1990) eight-step approach was implemented to analyse the research data in a comprehensive manner. Firstly, I read through all 34 transcripts in five sets (ten student psychologists, seven intern psychologists, seven qualified psychologists, five experienced psychologists and five senior psychologists) to cross-check each transcript with each voice- recording of the interview, to verify accurate transcription from the recording, to obtain a sense of the data collected as a whole and to initiate the process of generating ideas about the
information gathered. Secondly, I randomly chose one interview transcript from each set and started to annotate the meanings underlying the responses given. I then chose another transcript from each sub-sample of participants randomly and implemented the same process of annotation. Thirdly, I compiled a list of tentative themes. Similar themes (units of meaning) were grouped into clusters, and possible major and minor themes were outlined and notes about divergent themes and contradictions where noted. Fourthly, with the list as an organising scheme, I then returned to all the transcripts and began to carefully highlight, code and sort across interviews all appropriate segments of the text, aiming to capture the essence of what the participants shared in their interviews.
The literature review provided “sensitising concepts” and a “skeletal framework” in order to meaningfully sort and hold the data (Boeije, 2010, p. 23) to begin a concentrated analysis. Concepts were drawn and coded from an individual sentence or even paragraph of the data from the data corpus of over 600 pages of transcribed interviews. Concepts were then organised to identify distinct themes or patterns which placed the concepts into a meaningful interlinked whole. Importantly, a theme is a crystallisation of what participants conveyed in their words. Themes are established by “bringing together components or fragments of ideas or experiences
77 which often are meaningless when viewed alone” (Leininger, 1985, p. 60). How the themes cohered, ultimately, reflects the researcher’s subjectivity and rigour in tandem with the current knowledge base of the field (Leininger, 1985).
Fifthly, to organise and hold the data, categories were constructed. Categories were initially constructed deductively from the literature on psychotherapist development and finalised based on themes and sub-themes inductively found in the data analysed during the research process. In consultation with my two research promoters I “sounded out” the labels of the main categories in order to enhance the credibility of the current study (Niewenhuis, 2013, p. 109). Over-arching categories were identified and formed domains of the research which extensively organised the large data set.
Sixthly, the categories were sorted in alignment with the how psychotherapist
development is seen to unfold. Seventhly, each category was analysed to generate meaningful discussion points and to allow for formulations about psychotherapist development. Thematic analysis involves looking for patterns within the data or the common threads as well as variances within each interview as well as across the various interviews. The common and diverging threads enable a holistic yet nuanced account of the analysed data. Finally, I reviewed the analysed data, together with the stakeholder’s checks to see where recoding was required to ensure accurate and credible findings.
Anderson (2010) indicates that the findings and discussion of the qualitative data can be meaningfully combined together when reporting on a study. With this in mind, the Findings and Discussion of the current study will be reported together in a holistic manner in Chapter Four.