STATISTICAL CONTROLS IN RIDESHARING DEMONSTRATION PROGRAMS

The application of scientific experimental designs in ridesharing demonstrations is discussed. A review of typical designs, particularly those that use test and control groups and over-time observations of behavior, suggests that the numerous problems that jeopardize the validity of studies could be reduced or eliminated. Nine possible outcomes of demonstrations are reviewed and interpreted against the need for experimental designs. Two applications in the ridesharing area are then described: one conducted during a period of rapid background change (1979 energy crisis) and the other in a recent period of stability (1981). In both cases (conducted at employer and community sites in the Albany, New York, area), the use of a control group and before-and-after backgroung surveys permitted isolation of the true effects of the demonstration. In the first case (1979 energy crisis), this included the direct effect of the program (from coordinator records), indirect effects (from the existence of the ridesharing program itself), and external effects (from the energy crisis). In the second case (stable background), the indirect and external effects were found to be negligible. From this study it is concluded that the use of scientific designs in ridesharing analysis should be increased and expanded.

Media Info

  • Media Type: Print
  • Features: Figures; References; Tables;
  • Pagination: pp 1-9
  • Monograph Title: TRANSPORTATION INNOVATIONS: RIDESHARING TECHNIQUES AND PUBLIC-PRIVATE COOPERATION
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00381497
  • Record Type: Publication
  • ISBN: 0309035570
  • Files: TRIS, TRB
  • Created Date: Feb 29 1984 12:00AM