Estimating of bootstrap confidence intervals for freight transport matrices

Abstract

Freight transport studies require, as a preliminary step, a survey to be conducted on a sample of the universe of agents, vehicles and/or companies of the transportation system. The statistical reliability of the data determines the goodness of the outcomes and conclusions that can be inferred from the analyses and models generated. The methodology contained herein, based on bootstrapping techniques, allows us to generate the confidence intervals of origin-destination pairs defined by each cell of the matrix derived from a freight transport survey. To address this study a data set from a statistically reliable freight transport study conducted in Spain at the level of multi-province inter-regions has been used.