Risk and the designer : an investigation into what affects risk-taking and errors in the design of hazardous offshore installations using an experience sampling methodology

This exploratory piece of work has disclosured certain predictive affects associated with a designer's use of risky protocols (,Risky') and cognitive error ('Error'). The implication of this organisational study on risk and cognitive error (Simon, Hillson & Newland, 1997) rests in the potential for theory development in the role of the offshore designer. The focus of this research has been to investigate how designers of hazardous installations, in particular offshore platforms, might influence the design end users' safety performance. The risk paradigm provided the conceptual framework for making sense ofthe designer's attitude to risk. This exploratory research investigates if individual personality differences and the individual perception of risk and other constructs affect cognitive error and the use of certain risky design protocols. This study has extended the use of the Experience Sampling Methodology (ESM) into the complex design enivironment. A sample population of 167 design engineers from 55 design teams was assessed in situ, up to four times per day over four working weeks. The sample that participated in this organisational study was drawn from industrial sectors that involved the high hazard nuclear and offshore oil and gas industry. This research has been conducted in a number of stages, applying both conventional questionnaires and the novel electronic diary based techniques. Questionnaires were used to measure stable factors through individual maturity, such as personality, and an experience sampling methodology, using personal digital assistants that were used to record momentary data. The stable factors were analysed using exploratory factor analysis to derive 14 emergent factors from the six constructs examined. Multilevel hierarchical linear modelling, using HLM6, was applied to these factors and the momentary diary data. Whilst the research was primarily interested in the individual designer, there were certain interactions between the sample units that characterised the multilevel structure of the investigation. The momentary data nested within individuals, and within design teams showed that personality is significant in predicting cognitive error reports and the use of risky design protocols. Analyses indicated that emotionally stable individuals commit fewer errors, whereas extraverted, open and agreeable personality traits and the key job characteristic of job autonomy and the organisational safety climate are significant predictors in the use of risky design protocols.