The Use of Triangulation To Examine the Value of Several Measures of Safety Performance.* J. Harvey, H. Bolam, G. Erdos, and D. Gregory, University of Newcastle upon Tyne, UK
It has long been an issue that examination of reports of accidents and incidents is unlikely to present the full picture in the understanding of safety at work, and such reports may have only limited value in suggesting changes which may enhance positive safety performance. This issue is compounded by the infrequency of accidents and incidents and the fact that reports often tend to contain much qualitative information which is difficult to analyse. This paper presents new ways of analysing qualitative safety information and shows how such information may be considered alongside more conventional quantitative data in order to improve the understanding of safety performance. Ten different data and information sets were collected at a nuclear energy generating plant, including safety reports and audits, accident and incident data, safety suggestions, attitude measures and event diagnosis data. The data sets were analysed using content analysis techniques and attributional coding as well as rating and ranking scales. Using triangulation principles, the data sets are considered together, firstly in terms of positive safety behaviours and secondly in terms of negative safety behaviours, in order to examine their value as measures of safety performance. It is concluded that the measures together provide much more information about safety performance than any one, two or three of them alone could have done.
*Best Paper Finalist.
Go to . . .
2002 SRA Annual Meeting Table of Contents
2002 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page