Decision making in a nuclear emergency response organisation : the access controller
This thesis sets out to examine decision making in emergencies by industrial on-scene incident commanders. The aim was to develop a better understanding of decision making strategies used in hazardous, dynamic situations, with ill-defined problems and complex organisational issues. The project was set in the UK nuclear power generation industry. The focus was on the role of the Access Controller (AC), who, as the operational level decision maker in a nuclear emergency response organisation (ERO), is particularly crucial when effective decisions must be made that may affect the potential escalation of the situation. The Naturalistic Decision Making (NDM) approach (Zsambok and Klein, 1997), which focuses on the context of the decision making environment, expertise, and the holistic decision making process, provided the theoretical framework for the thesis. Within this framework, the model of Decision Making Under Stress (DMUS) developed by Stokes, Kemper and Kite (1997) in the aviation domain was tested in this nuclear setting. The thesis consisted of five studies conducted with participants from various fixed nuclear installations throughout the UK, culminating in the development of a computer-based decision making task. The main findings of the thesis were that the majority of AC decisions are rule-based (80%). Written procedures did not appear to enhance decision making in emergencies, but prior experience of making a particular emergency decision was found to assist more competent decision making. Duration of experience in the AC role was not in itself a good indicator of decision making expertise, which was found to be dependent on a corpus of problem solving strategies from normal operational role, along with recognition of critical cues (situation awareness) and recall of a schema or script stored in LTM. The conclusions presented in this thesis, for both the role of AC specifically and a nuclear ERO in general, should go some way towards improved understanding of decision making in complex, hazardous environments. Moreover, these results may generalise to other industries and organisations where effective decision making during emergency response is essential.