Much research has been conducted on the effect of management and organizational factors on the productivity of organizations. Furthermore, these factors have been identified as significant during accident investigations for several NASA incidents. This paper presents a probabilistic, causal, reasoning approach to capture and quantify the dependencies between the key characteristics of management and organizational factors and the likelihood of command file errors. Bayesian Belief Networks are used for modeling these dependencies and expert elicitation methods are used for populating these networks. The result is a model that is representative of a significant body of knowledge and is executable for the purpose of decision support and if-then analyses. This model is presented in the context of the existing body of work in managing command file errors at JPL, the body of academic and theoretical research in the field, and the history of NASA accidents and investigations.
展开▼
