Observation – The research used observation in two distinct ways; to develop an understanding of the physical symbols of relocation and to develop a greater understanding of the case organisations. With the exception of participant observation, it appears that ‘qualitative observation’ is an underdeveloped aspect of research. Indeed, it has been suggested that: “A common mistake among observers is to take the physical environment for granted”.
The preliminary fieldwork had drawn the researcher’s attention to a variety of observational cues. These included observation whilst in reception areas, waiting in the car outside the organisation, walking around public areas in the organisation and using facilities such as the library and canteen. Taken on their own, these ‘chance’ observations may have provided useful sources of information. However, whilst it was both undesirable and impossible to filter out these influences, it was important to take a more systematic approach to observation (Becker and Geer, 1982).
Documents – In addition to the main literature review, further dimensions of data were obtained through the analysis of documents. Moser and Kalton (1975) identify three different types of document: those giving information about specific research areas, those giving information about the population being studied, and personal documents. Alternatively, Scott (1990) classifies twelve different types of document based on the attributes of access and authorship. Whilst such demarcations may be suitable for general social research, they were in appropriate for this particular investigation. Consequently, for the purposes of this work, two types of document were identified: primary documentary data produced by the case organisations; and primary documentary data produced about the sector.
Questionnaires – In addition to questionnaires used during the preliminary fieldwork, a survey was conducted into mission and vision statements in the electricity supply industry. All three surveys were based on the use of semi-structured questionnaires which contained mainly open-ended questions. This approach was intended to allow respondents to reply ‘in their own terms’ 5 without predetermining the content of their contribution. It has been argued that data collected in this way is affected by: the respondent’s ability to communicate in writing; the inability to probe or review contributions; and the reliance on respondents providing full replies (Patton, 1980). However, each survey was followed up with written requests for information and telephone requests for responses or for the clarification of contributions
It has been suggested that: “There are no formal, universal rules to follow in analyzing, interpreting, and evaluating qualitative data”6. However, it has also been argued that: “We should continue to be concerned with producing texts that ex plicate how we claim to know what we know”7. Indeed, it appears that one of the potential dangers of adopting a predominantly qualitative research approach would lie in not explaining how the researcher turned the raw data into findings.
The research had an epistemological leaning towards inductivism, with the consequent broadly inductive approach to analysis meaning that: “… the patterns, themes, and categories of analysis come from the data; they emerge out of the data rather than being imposed on them prior to data collection and analysis”8. The resultant approach to analysis was therefore continuous, iterative, intensive and comparative. For example, the researcher’s immersion in each case led to the triggering of analysis and interpretation during the fieldwork. To control this process, such interpretations were entered into a ‘day -book’. These interpretations were subsequently examined more rigorously during the process of data analysis which followed the conclusion of the main field work. Consequently, one of the main strengths of the research strategy was the opportunity to develop continuously an understanding of the data while it was being collected.
Academic research is frequently assessed on the basis of questions regarding validity, reliability and generalisability. However, this language is believed to derive from quantitative research, which makes it difficult to apply in the context of qualitative re search. Alternative labels have been adopted by a number of researches who suggested that: “… the basic purpose of qualitative analysis is to provide use full, meaningful, and credible answers”, and Altheide and John son (1994) who argue in terms of ‘validity-as-reflexive –accounting’.
The underlying theme of these apparently contradictory positions is that academic research is distinctive in its method, even though these methods may vary quite considerably between different philosophical approaches. The following critique will consequently focus on the ‘logics of engagement’ which were used to link the researcher with the research objects, rather than the differences, which are some times used to defend one philosophical approach against an other. This critique will be structured around four overlapping categories proposed by Miles and Huberman (1994) which cover questions of confirmability, dependability, credibility, and transferability.
The ‘objectivity/confirmability’ of the research was dealt with in a number of ways. There searcher strove to present as full as possible a description of the methods used in the study. Furthermore, the researcher attempted to be ‘reflexive’ in terms of the personal experiences, which preceded the research project, and in terms of acknowledging the potential impact of these on the field work and subsequent analysis. In addition, the data was subjected to competing explanations, most notably in the application of different theories of change management. By addressing such issues the researcher sought to reduce any potential bias in the work.
However, the basis of the research was in ‘analytic realism’9. This means that the researcher presupposed that reality is social constructed and interpreted. Consequently, the researcher endeavoured to represent the social world constructed by those researched, and to highlight the influences on the way in which this had been interpreted by the researcher. As a result of this the reader should be better placed to make their own interpretation of the research findings.
The second question concerns the ‘reliability/dependability/auditability’ of the findings, which was regarded in terms of whether or not the re search systematically studied what it claimed to study. As has been out lined previously, the research was based on a number of fundamental research questions which linked the fieldwork and theory. These questions were used to maintain the focus of the work, while at the same time allowing relevant (but unanticipated) is sues to be encompassed during the field work. The preceding discussion of the fieldwork has described the role of the researcher within the case studies i.e. as a researcher collecting data for a doctorate.
Efforts were made to collect data from a variety of data sources and using a variety of data collection instruments which would be consistent with the research questions. Furthermore, data were assessed on a number of dimensions including the degree to which meaning was shared between the researcher and the informant, and consistency with data gathered through other means. Finally, findings were subjected to peer review in terms of the continuous process of supervisory review, internal reviews of written work, responses to internal and external presentations, and the publication of articles. The dependability of the research was therefore safe guarded by aiming to collect data systematically over time, and systematically within each case.
The ‘internal validity/credibility/authenticity’ of the findings was taken to mean the degree to which the findings have meaning for those interested in them. One of the main tactics for dealing with this was to present comprehensive descriptions of the cases studied which were closely linked to both their inner and outer con texts. Triangulation was used extensively to synthesise multiple perspectives and clarify meanings.
In addition, findings were drawn from several cases which covered a number of different contexts. Furthermore, efforts were made to identify disconfirming evidence, to interrogate the data using rival explanations, and to link the findings with extant theory. The findings resulted from a systematic process of examining the propositions and questions generated by the research. Finally, where areas of uncertainty arose, these were noted and identified in the text. Consequently, the issue of credibility was addressed in terms of description, interpretation and theory.
Many academics feel that they never really produce as many pieces of work from their works as they should have done. There are many reasons for this failure, ranging from a lack of confidence in the value of preliminary results, to feeling the need to escape from a beast which has occupied much of a researcher’s home and work lives for several years. However, starting to produce distinct outputs from the early stages of a qualitative management research project can be a useful way of improving the quality of later work.
Because of the grounded theory approach to this research project, findings were subjected to numerous iterations of analysis and interpretation. These iterations encompassed a range of formal and informal reviews by peers. For example, the whole research process was subjected to regular reviews of progress by the researcher’s academic supervisor.
Towards the conclusion of this paper the output has portrayed the ways in which a specific qualitative management research project was brought to a successful conclusion. Multiple instruments could be used to collect data including over 100 interviews, three semi-structured surveys, structured observation, literature reviews, and document reviews. These individual data collection instruments have been reviewed critically and the extensive triangulation used to synthesise multiple views and clarify meanings has been described.
We have outlined the combination of intuition, data saturation, time, and contextual changes which were used to bring the fieldwork to a close. It has also explained the iterative and continuous process of data analysis and interpretation. In particular, it has argued that although the research was inductive by nature, the focus of the research was maintained through the generation of research propositions and questions from the main literature review and initial fieldwork. Furthermore, the overall research strategy was critically examined and defended in terms of the confirmability, dependability, credibility and transferability of method and findings. However, this examination was undertaken in terms appropriate for the research philosophy adopted and not in terms imposed by other philosophies.
Finally, several outputs from the research have been high lighted including published articles. The study supported the contention that: “Interpretive studies of change in complex business organisations are relatively rare”10. In view of the apparent lack of research precedent (both of the subject and the cases) the research strategy incorporated a ‘controlled opportunism’11 to data collection.
Furthermore, although the research was based in inductive and grounded theory it did not ignore existing theory (see, for example, Strauss and Corbin, 1994). Indeed, the research presented an opportunity for ‘theory elaboration’12 whereby extant change management theories could be developed in apparently novel settings. Additionally, the iterative and comparative nature of the research strategy was also important in conducting a more thorough and transferable interrogation of the research propositions and questions. Ultimately, conducting the chosen research project was hugely rewarding and, despite the challenging, un certain and emergent nature of qualitative management research, we would encourage others to follow suit.
1. Altheide, D. L. and John son, J. M. (1994). Criteria for assessing interpretive validity in qualitative research In: Denzin, N. K. and Lincoln, Y. S. eds Handbook of Qualitative Research. London: Sage, 485- 499
2. Becker, H. S. and Geer, B. (1982). Participant observation: the analysis of qualitative field data. In: Burgess, R. G. ed. Field Research: A Sourcebook and Field Manual. London: George Allen ; Unwin, 239- 250