Modal Split Analysis Using Logit Models

Abstract

This study attempts to identify the mode and the factors that contribute to the selection of a particular mode for commodity movements provincially based on average shipment size, control, loads, hire, and type of commodity. The main objective of this study is to propose a more comprehensive and statistically credible method to analyze the vast data required in transportation planning. It involves the application of standard statistical techniques such as the log‐linear and logit models. The data are collapsed to form multidimensional contingency tables in order to develop these models. The data from both shippers and consignees are taken from a survey conducted by Alberta Transportation, Canada. From the analysis it is found that truck and rail are the only two major carriers of freight across the province. Truck mode dominates over rail in transporting all commodities. The less‐than‐full‐load market belongs exclusively to truck mode. The rail mode shares a very small percentage $(<15\%)$ of the full‐load market, and is used to transport specific bulk commodities under higher average shipment size.