Model Results

This chapter presents the freight mode choice models obtained by the team. Since the descriptions of independent variables and model specifications for various types of models were provided in Chapter 7, this chapter only provides the model results. In deciding which models to present in this chapter, the team used two criteria. The first was conceptual validity, which in most cases consisted of verifying that the models had negative parameters for rate, transit time, and generalized cost. The constant term in the models should be positive, as it represents the bias towards the truck mode. The team, to the extent possible, compared the values of the parameters for the various models to ensure that comparatively speaking, the results make sense. The second criteria was statistical significance to ensure the models were statistically solid. In a number of cases, however, the team decided to include models that though conceptually valid, had parameters that were not statistically significant. These models have been labelled as “Not Recommended: Conceptually Valid but Statistically not Significant.” These models were included for reference purposes.

This chapter is divided into four sections. The first presents the shipment size models as a function of great circle distance. The second section explains the market-share models. The subsequent three sections present the shipment-level models with three types of weights, unweighted, domestic, and total weights respectively.

8.1. Shipment Size Models
8.2. Market-Share Models
8.3. Shipment-Level Models: Unweighted
8.4. Shipment-Level Models: Weighted – Domestic Cargo (Dom-Weight)
8.5. Shipment-Level Models: Weighted – Total Cargo (Total-Weight)

Previous   Home       Next   

524 Total Views 3 Views Today