6.4 Research Process and Design
6.4.5 Transcription, Coding and Analysis
All the interviews were recorded to avoid inaccuracies resulting from poor recall, and were, then, accurately transcribed in Greek. Key sections and sentences of the transcribed interview were translated into English for presentation in the thesis. The process of translation is embedded within the socio-cultural context (Halai, 2007), as is the process of transformation of verbal or spoken conversation into textual form. Lapadat and Lindsay (1999, p. 82) emphasised that such a transformation is a theory-laden process and the decisions or choices made during it are inevitably influenced by the analysis and interpretation of findings (Halai, 2007).
The translation was conducted by the author, a native speaker of the Greek language, with a deep understanding of the wider social and political Greek context; this improves the validity and the reliability of the study. Although the majority of the interviewees spoke in English, the possibility of speaking Greek meant that the respondents were more open and expressed their opinions more honestly than would have been possible if the interviews had been
118
confined to English language use. Furthermore, the author was able to combine her technical accountancy competence with an understanding of Greek society and the Greek business environment, to provide a unique insight into the implementation process of IFRSs. A literal translation (i.e. translation word-by-word) was chosen in order to fully appreciate what the participants said and to ‘make one’s readers understand the foreign mentality better’ (Honig, 1997. p. 17). However, sometimes the interviewees’ accounts required a free translation approach, in order to render the meaning of the interviewees’ accounts into English.
Although translating and editing quotations involves a risk of misrepresenting the meaning and message an interviewee conveys (Rubin & Rubin, 1995, p. 273) or loss of information from the original, the interviews as already discussed were not intended as a solid body of knowledge, but as inspiration to problematise and creatively build on certain views.
The large amount of unstructured interview accounts collected were then grouped and organised. In order to analyse the interview material, computer-assisted qualitative data analysis (CAQDA) methods were tested. Although such software packages can be useful for the purposes of certain studies when analysing interview and other audio-visual data, the systematic and linear codification of the interview accounts resulted in a fragmentation of the information, as emphasis was placed on the key portions of the interviews. In order to gain a holistic understanding of the structure, the contexts within which views were expressed and the meaning of the discussion was important, making a manual approach to categorising the evidence and extracting the main themes more appropriate.
During the reading of the transcripts notes were taken and ideas and relevant extracts were highlighted. Care was taken to ensure that all the information was considered by crossing out the parts of transcript once it has been incorporated into the analysis. This procedure meant that the risk of selective use of data was avoided and that none of the themes were suppressed. The transcripts were coded first using a relatively general coding scheme, containing some ideas or perceptions that were regarded as interesting and pertinent to the research objectives, such as claims about benefits, criticisms, the local flavours added to IFRSs adoptions and reasons explaining the gap between the ideals of quality and whether and how the latter is achieved in everyday practice. Later, more specific themes emerged during analysis of the interview evidence.
119
In the analysis of the interview evidence some major themes, such as views on the ‘benefits of the use of IFRS’, ‘users of financial reports and their information needs’, ‘environmental factors affecting IFRSs’ application’, etc. emerged. The theoretical framework was revisited using interview evidence. Based on the theoretical framework adopted, the transcripts were categorised according to themes, such as perceptions about ‘the role of the state in financial reporting’, ‘inconsistencies between IFRSs’ objectives and practice’, ‘common sense’
benefits of accounting information based on IFRSs’, and so on. The evidence was compared with theory to develop a coherent and informed understanding of the use of IFRSs in the specific national context. By thinking in a dialectical way, divergence between established ideas and conceptions and alternatives to these ideas and practices were identified. Although the preparers and users participating in the study did not automatically raise identical concerns, they referred to similar themes and adopted similar patterns when reflecting upon key issues. The discussion revolved around emerging themes (Appendix 7 provides an illustration of the rationale of the analysis of interview evidence).
Quotations were added to the text for analysis when it was felt that they added credibility to the analysis and enhanced understanding of the various themes. Consistent with the research focus of this study, the findings illustrated in the following sections aimed to reflect a variety of perspectives, rather than reporting on relative frequencies by stating percentage figures.
This was in line with methods of data presentation in qualitative analysis, which are generally focused on understanding socially constructed concepts, and contain a minimum of standardisation, statistical methods and quantitative measurements (Sarantakos, 2005, p.
344). Terms such as ‘many’, ‘some’, and ‘few’ can be used when describing data. However, these proportions are not provided as statistical information but rather aim to show theoretical saturation. A variety of responses were also expressed in the final analysis, although in some parts, the reliance on reflections from a particular group, for example auditors or bank managers may be greater, due to the importance of their insights relative to a specific theme.
In order to provide a more critical approach to dominant views on the subject under investigation, and to encourage the exploration of different lines of interpretation the analysis was also systematically guided by an alternative understanding of the interviewees’ roles or points of view. For instance, as Alvesson (2011) observes interviews may involve difficulties of representation and normative pressure for adopting certain talk. They may require adaptation to normative pressure by mimicking standard forms of expression, such as “IFRSs are comparable, more transparent and of better quality than local GAAP” which mimics the
120
rhetoric promoted by the IASB. There is a challenge associated with interview subjects who are located in political arenas and driven by interests when promoting certain kinds of truth.
Auditors of the Big 4, government agents or academics may have different motivations or develop differing interests or rationales when participating in an interview.