Realising CWA Designs within the Network Paradigm
Abstract
The present report reflects attempts to convert the outputs of one form of Human Factors analysis, Cognitive Work Analysis (CWA), into another, referred to here as the network paradigm. Examples of the use of networks in sociotechnical systems design include the use of the HFI DTC’s WESTT (Workload, Error, Situation Awareness, Time and Teamwork) tool.
The motivation for this work lies in providing a form of early Human Factors assessment that can be performed during the Concept phase of the CADMID process. CWA is formative in nature, which is to say it suggests a range of possible forms a system could take. The network paradigm typically deals with sociotechnical systems that are already extant on the basis of empirical data. By applying network analysis techniques to CWA created but otherwise putative sociotechnical systems, there is the prospect of carrying out a deeper investigation of the qualities that system has well in advance of any instantiation of it.
Three sets of social networks (actor-to-actor) and use-case networks (actor-to-function) were produced:
- A Maximal rendering of the CWA SOCA (Social Organisation and Cooperation Analysis) diagram that depicts the full range of possible functional activities over time; in other words all actors doing everything they possibly can (or would have to do).
- An Intermediate rendering of the CWA SOCA diagram that depicts actors performing the functions they are most likely to have to do in given situations.
- A re-drawing of the Maximal rendering that excludes a specific function (namely, “Information Exchange”) that may have obscured latent structures within the networks. This also demonstrates the use of the network approach to examine the potential failure modes of a sociotechnical system where, for example, a specific form of information technology fails.
In addition, by cross referencing the SOCA diagram against an Abstraction Hierarchy of functions that relates physical functions with specific forms of technology (artefacts), it was possible to plot networks that depict which artefacts are shared by actors in different situations.
It was found that:
- The command team networks were highly interconnected both in terms of mappings between individuals and between individuals in tasks.
- Therefore, digitisation for distributed working may be difficult as each member of the team has complex patterns of interaction (both in terms of collaboration on functional activities and in terms of their sharing of artefacts). While some artefacts are clearly used more than others (particularly “Requests for Information”), the pattern of connectivity means that to facilitate distributed working all the artefacts used by the command team would likely have to be digitised.
- The two previous findings summate to the finding that there are no natural fault lines within the command team meaning that there are no clear candidates for distributed working; rather, the networks imply that collocation would be strongly favoured.
- In general terms the translation of CWA outputs into network diagrams seems a useful activity that offers insights above and beyond those offered by the CWA figures themselves. The clearest benefit is in understanding how a specific entry in a SOCA diagram or Abstraction Hierarchy fits into a ‘wider picture’ context.
contact us