Freight Mode and Shipment Size Choice: Joint Approach versus Sequential Approach

There is preference toward adopting simulation based or analytical joint models in analyzing freight mode and shipment choice. However, in joint models the information of shipment size cannot be used for mode choice and vice-versa. In this research, the authors investigate if a sequential approach that allows them to use the information from the other dependent variable in their model can enhance the model fit. In their study, the authors systematically compare the performance of the joint copula based approach with a latent segmentation approach that probabilistically allocates each record to the two sequences: (1) mode first-shipment size second; (2) shipment size first-mode second. Mode choice analysis was performed employing both random utility (RU) and random regret (RR) based MNL and shipment size was analyzed using ordered logit model. Within the copula model structure, the authors also allowed the dependency to vary across each observation. The proposed models were estimated using freight flows within Florida and Piedmont Atlantic region obtained from 2012 U.S. Commodity Flow Survey (CFS) data. The estimated results reveal that copula model performed better than latent class model. A host of comparison metrics are employed to compare the performance of the two models and it is found that copula model outperformed the latent class model in both RU and RR paradigms.

  • Supplemental Notes:
    • This paper was sponsored by TRB committee AT015 Standing Committee on Freight Transportation Planning and Logistics.
  • Corporate Authors:

    Transportation Research Board

  • Authors:
    • Keya, Nowreen
    • Anowar, Sabreena
    • Eluru, Naveen
  • Conference:
  • Date: 2019


  • English

Media Info

  • Media Type: Digital/other
  • Features: References; Tables;
  • Pagination: 9p

Subject/Index Terms

Filing Info

  • Accession Number: 01697575
  • Record Type: Publication
  • Report/Paper Numbers: 19-05742
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 7 2018 9:31AM