On Bayesian sample size determination

Abstract

Three Bayesian methods are considered for the determination of sample sizes for sampling from the Laplace distribution – the distribution of time between rare events – with a normal prior. These methods are applied to the sizing of aircraft mid-air collisions in a navigation system or large flight path deviations of aircraft in air traffic management scenarios. A computer program handles all computations and gives a good insight about the best suggested method.

Keywords:

• kth average coverage criterion
• kth average length criterion
• worst outcome criterion

Acknowledgements

The authors wish to thank the referees for providing constructive and helpful comments that have led to substantial improvement in the quality of the paper.