Biases in incident reporting databases: an empirical study in the chemical process industry

The use of incident reporting schemes is becoming increasingly widespread in many domains such as industry, transportation, and health care; but the extent to which these schemes really capture a representative sample of actual events remains a topic of debate. While this has led several researchers to provide organisational design guidelines for implementing such schemes, much less is known about the individual reporter's perspective on contributing. Based on a literature survey and a diary study in which chemical plant operators reported events involving recovery from self-made errors, complete with reasons why they would or wouldn't normally report this, we identified six categories of reasons for not reporting. The distribution of the diary study events over these categories highlighted a difference between the plant's operators and management, with regard to the perceived importance especially of successfully recovered events with no remaining consequences, or events considered to be nothing new. This difference should be addressed by communicating the learning potential of these types of events. (C) 2003 Elsevier Ltd. All rights reserved.