The purpose of this paper is to describe a methodology based on vignettes and its application to measure the consistency of railway delay attribution, in order to report on experiences of using the methodology in this context.
A survey is used to ask personnel performing delay attribution about how they would report the delays described in the vignettes of the survey. The development, application and analysis of the survey are thoroughly described here.
The methodology proved useful for measuring the consistency of delay attribution.
The methodology also supports a further investigation of the accuracy of delay attribution systems.
A survey similar to the one presented here can be used by railways to estimate the accuracy of their delay attribution systems, as well as to continuously improve them. Changes to computer software and training will improve the delay attribution system under study.
By investigating current delay attribution practices using surveys based on vignettes, drawbacks of the current delay attribution system can be identified and remedied. The methodology is applicable to a wide class of deviation attribution applications.
