STATISTICALLY BASED METHODS FOR EFFICIENT SAMPLING OF INLAND WATERWAY FREIGHT CHARGES

Preliminary results of work on the estimation of freight charges for commodity movements in the Ohio River Basin are presented. The U.S. Army Corps of Engineers had rate quotations on 875 of the approximately 12,000 movements in the basin and funds to obtain an additional 325. Consad Research Corporation was engaged to apply statistically based analysis to specify the sample that would give the most accurate estimate of freight rates for the given sample size. The traffic universe and original rate quotations were stratified by commodities and annual tonnage. There were 18 commodity strata, each of which was broken down into 3 volume substrata. The original rate quotations were used to calculate the sample size required for each of the 54 cells at several levels of sampling error and confidence intervals. This indicated the additional points needed. These points were also apportioned among the strata on the basis of freight charges to achieve efficiency in estimating total system charges. Points were also allocated among the three cells of each stratum for purposes of statistical reliability. The resulting sampling plan had a projected error of 5-39 percent for each stratum and an error of 12.9 percent for the system as a whole. Mathematical relations were also derived and applied to estimate freight rates for water and alternative overland movements. The methods used should result in greater reliability in the estimation of freight charges.

Media Info

  • Media Type: Print
  • Features: Tables;
  • Pagination: pp 14-19
  • Monograph Title: Inland Waterway Studies
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00330032
  • Record Type: Publication
  • ISBN: 0309031052
  • Files: TRIS, TRB
  • Created Date: May 21 1981 12:00AM