Many experimental situations lead to inverse sampling schemes with some random or non-random stopping rule, since at each experiment only a bounded number of observations can be made. This note discusses the problem of estimating the unknown probability in the underlying geometric distribution of such schemes. The author encountered this problem in a very special context, that of estimating the proportion of non-fare-paying passengers in a local transportation system. The methods derived were used on material collected during 1 month by the non-uniformed ticket controllers of the Gothenburg, Sweden, transportation system. Some 3,079 cars and 40,786 passengers were checked; 982 free passengers and 132 stowaways were found, yielding the estimates.

  • Corporate Authors:

    American Statistical Association

    806 15th Street, NW
    Washington, DC  United States  20005
  • Authors:
    • Jagers, P
  • Publication Date: 1973-12

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00054343
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jun 10 1981 12:00AM