The process of data analysis is not straightforward since it includes not only the practice
of recording ideas and tentative interpretation, but also a comparison of data with a
review of literatures in the field notes. To a certain extent, data analysis shapes the
the ethnographer not only depends on field notes or interview transcripts, but also on his
field experience, unwritten but embedded in his memory (Geertz, 1973; Okely, 1994;
Rosaldo, 1993).
The main assumptions of the ethnographer are the meanings ascribed by participants in
the field to what they do and their reasons for doing this, their interpretation of their
situation and their perception of their own identities (Silverman, 2001). These have
implications for data analysis in terms of identifying means and methods through which
the participants in the field perform their activities (Silverman, 2001; Schatzki, 2002).
The analysis should at least comprise of a description of how the participants perform
their activities, the strategies that they employ to achieve their purpose, how they relate to
each other, and how they make use of objects and artefacts in their activities.
The data was analysed using qualitative procedures; the open coding technique as
suggested by Strauss and Corbin (1998) was adopted at the initial stage of data analysis to
identify issues, actions, interactions and key concepts through careful reading of
interview transcripts, documents, notes and reflections of the context in which they
occurred.
Interview transcripts were arranged according to the division and department visited, and
the analysis was carried out manually using Microsoft Word: codes were recorded in the
distinguished by different coloured highlighting. In the field notes, the codes were
marked in the notebook margins.
Subsequently, these analytical codes were arranged into the categories suggested by the
data. This was done by transferring similar codes from both interview transcripts and
field notes into new documents and placing them according to their appropriate headings.
Following Silverman (2001), in developing categories to represent the data, attention was
given to the actions, the meaning underlying those actions and wider phenomena to
which these actions respond and which they shape. Categories are defined as concepts,
derived from data that stands for a phenomenon, or as abstract explanatory terms (Strauss
and Corbin, 1998, p.114). Ahrens and Dent (1998) refer to this process as pattern making,
i.e. the activity of grouping the data into different categories and tracing the
interconnections between them.
The interview transcripts, documents and field notes were therefore scrutinized in order
to identify statements or discourse that reflected the perceptions and views concerning the
use of accounting control systems in the everyday activities of different departments in
the organisation. Emerging patterns or themes were also derived through comparative
analysis, which is the act of finding similarities and differences in the data by comparing
conceptually similar events, incidents or issues with those previously coded. To arrive at
an understanding of the circumstances of the organisation, much reflective thought and
obtained from document review and observation. The use of the notion of triangulation
did not aim to achieve validity, as suggested by Yin (2003), but to provide multiple
sources of evidence in making sense of the phenomenon studied.
The insight that I gained from continuous reading of related literatures allowed me to
view the data from various angles. As new ideas emerged from on-going reflection and
new interpretation, data was revisited and some initial codings were then changed.
Thought emerged surprisingly sometimes; at the bus stop, while watching movies,
cooking, in the morning; not necessarily when I am in the office with a computer and
piles of books.
The relationship between data and ideas is crucial in carrying out analysis in the
qualitative study; it should be an iterative process in which the back and forth movement
between ideas and data is expected. Indeed, Weick (1995) suggests that this is a basic part
of theorising activity: the process of theorising consists of activities such as abstracting,
generalising, relating, selecting, explaining, synthesising and idealising. These on-going
activities intermittently deliver reference lists, data, lists of variables, diagrams and lists
of hypotheses, emergent products which summarise progress, give direction and serve as
place markers; they have vestiges of theory but are not themselves theories (p. 389).
“All interpretations are provisional; they are made by positioned subjects who are prepared to know certain things and not others. Even when knowledgeable, sensitive, fluent in the language, and able to move easily in an alien cultural world, good ethnographers still have their limits, and their analyses always are incomplete” (Rosaldo, 1993, p. 8).
This assertion is very true for my study, which attempts to demonstrate that my
knowledge of accounting, my life experience and my awareness (through reading) of
workplace behaviour, which I believe to be as equally important as systematic
preparation of fieldwork.