Estimating Survey Weights with Multiple Constraints Using Entropy Optimization Methods

Household travel surveys are the main source of information for understanding individual travel behavior and also for conducting a variety of travel analyses, from reporting descriptive statistics to calibration and validation of advanced travel forecasting models. A challenge often faced by transportation professionals is to accurately expand the survey households to represent the population. The problem becomes even more complicated when dealing with household travel surveys, because the goal is to find household weights such that the distributions of the characteristics in the weighted sample not only match given distributions of households but also those of persons. This paper presents an Entropy Maximization methodology to estimate household survey weights to match the exogenously given distributions of the population, including both households and persons. The paper also presents a Relaxed Formulation to deal with cases when constraints are not feasible and convergence is not achieved. The methodology is applied to a large geography - Maricopa County region, Arizona, and a small geography – blockgroup, and estimation results are presented. Estimation results show that the Strict Formulation can be used to estimate the weights when constraints imposed by distributions of population characteristics are feasible. Relaxed formulation can be used to estimate weights when the constraints are infeasible such that distributions of the population characteristics are satisfied to within reasonable limits.

Language

  • English

Media Info

  • Media Type: DVD
  • Features: Figures; References; Tables;
  • Pagination: 22p
  • Monograph Title: TRB 88th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01127605
  • Record Type: Publication
  • Report/Paper Numbers: 09-1354
  • Files: BTRIS, TRIS, TRB
  • Created Date: Apr 30 2009 8:28AM