3. Incorporation of individual findings into a model of the organization under study; 4. Presentation of evidence and proof.
Whilst Becker (1958) was describing sociological participant observation, this approach is broadly applicable to Human Factors research and was adopted for this thesis. The analytical stages are now briefly discussed in turn.
1. Selection and definition of problems, concepts and indices
This research was undertaken with the intention of adopting sensemaking and distributed cognition as theoretical lenses through which to guide data collection and interpretation. As a result, the broad themes to be investigated were identified a priori, from the features of sensemaking and distributed cognition. These were summarised in Table 2.4 in Chapter 2.
Conducting sensemaking research from a distributed cognition perspective provides an approach to the problem of what data to collect, as distributed cognition researchers attempt to describe a range of system features in order to uncover cognitive processes. These can include system goals, organisational structures, individuals’ interactions with one another and with artefacts, the process of representation transformation and the role of the work environment (Rogers and Ellis, 1994). Provisional data collection was therefore carried out to begin to develop an understanding of the goals, organisation, artefacts and procedures of emergency response work and to establish whether the high-level themes were applicable to the domain. It was then necessary to identify the lines of enquiry to take during data collection in order to address the research questions. Focussing on the research questions helps to avoid the pitfall of being distracted by the mounds of interesting but superfluous data (Baxter and Jack, 2008). Preliminary data collection identified specific aspects for further investigation – for example, the roles of the notepad and the incident management system in supporting sensemaking during 999 calls. This stage also featured the development and implementation of a strategy for investigating the different areas of the emergency response C2 structure in order to develop appropriate case studies to investigate them.
2. Check on the frequency and distribution of phenomena
Becker (1958) describes this stage of analysis as a process of identifying the representativeness and pervasiveness of the concepts identified in stage 1. Lubbe (2003) notes that:
“…the case study methodology is not designed to measure the frequency of occurrence of events but rather to support or reject theoretical propositions.” (Lubbe, 2003, page 14)
As such, this involved comparing and contrasting findings to published literature in order to situate the study into the wider body of research (Baxter and Jack, 2008). The multiple perspectives approach to data collection also enabled corroboration of findings, to ascertain their representativeness.
Routine emergency case study corroboration
Where a feature of sensemaking during routine emergency responses was identified using one method, it was verified through other sources or data collection methods. For example, multiple Call Handlers in different Control rooms and police forces were observed using ‘active listening’ questioning techniques; these observations were then validated through interviews with call handlers and supervisors, as well as through a review of Call Handling training material. Similarly, the widespread and consistent uses of notepaper and notepads observed in police control rooms in different forces throughout the data collection period supported the conclusion that this was an established phenomenon, rather than a short-lived or local idiosyncrasy. Interviews with Call Handlers and supervisors supported these observations.
Points of divergence between data collection methods and sources are to be expected and may indicate areas worthy of further investigation (Bryman, 1988). However, careful investigation is necessary to establish what is actually happening, as well as the reasons for the divergence. For example, the Controllers spoken to stated that they allocate incidents to specific units, rather than asking for ‘any unit’ to attend; this was supported by a review of training materials and comments from supervisors, both of which stated what Controllers should do. However, over the course of control room observations, the researcher observed Controllers repeatedly radioing ‘any unit’. Later, during participant observation sessions with both Police forces, ‘any unit’ requests were regularly heard and the implication that this was a common occurrence was corroborated by interviews with Police Officers and Sergeants. Having made this observation and following a period of reflection, subsequent control room and participant observation sessions sought to establish why Controllers might feel the need to ask for ‘any unit’, through a review of the demands of their role and the information and resources available to them.
Major incident case study corroboration
Interviews with staff from the main response agencies provided the principal source of information on the sensemaking processes involved during the emergency response. The interviews were analysed in parallel, to identify points of commonality and divergence; participants were then contacted again (by telephone or email) in order to clarify points and to check for accuracy. These points of divergence were reflected on not only in light of sensemaking theory, but also following
interviews and observations with senior staff from The Fire Services College and WMP. Publicly available documentation on the incident was used to verify the accounts provided. Observations of tactical decision exercises and interviews with senior commanders at the Fire Services College, along with a review of Fire Service doctrine (e.g. HM Government, 2008) provided background knowledge regarding Fire and Rescue major incident practices, as well as validation of the account provided by the respondents involved in this major incident. The later visit to Gloucestershire Constabulary Headquarters and interviews with senior staff, along with the in-depth interview with the retired Chief Constable provided further background information on major incident practices and the countywide ‘water emergency’, as well as corroboration of the timeline of events provided by the case study interviewees.
Checking with practitioners
Patton (1990) suggests a further method of corroboration, which is to have members of the domain in question review the study findings. In line with this approach, checking with practitioners was used during interviews as a means for evaluating whether earlier observations represented typical events, although responses were treated with caution, as some discrepancies were noted between interviewees verbal accounts (what they would do) and observed events (what they did).
Checking with practitioners also enabled the researcher’s understanding to be reviewed, in order to give confidence regarding the trustworthiness of the developing analysis. Drafts of research reports and conference papers based on this research (cf. page xi) were submitted to successive WMP Force Communications Centre Commanders and senior Warwickshire Police Officers for comment. Additionally, the researcher prepared two reports (totalling 8,000 words) for Warwickshire Police on preliminary findings from the Communications Centre observations and the secondment with WMP; these were submitted to senior Officers for comment. As Krefting (1991) suggests, checking with practitioners became progressively less valuable towards the latter stages of the research process, as interpretations and higher conceptual analysis became the prominent consideration over descriptive data.