ACCURACY IN PUBLISHED STATISTICS

Background information is given to explain the effect of the precision of original measurements, sample bias and sample variation on the accuracy of the published Transport Statistics Great Britain (TSGB) 1995. The statistics are gathered from administrative sources, census sources and from sample sources. Census data, whilst rarely error-free, is subjected to fewer types of error than information based on sampling procedures. The limitations, benefits and requirements of each source are outlined. Sample surveys must be carefully designed and in general data collection costs tend to rise in proportion to the square of the required improvement in sample accuracy, with data storage, validation and analysis costs also increasing rapidly. The levels of accuracy required must be realistic bearing in mind the purpose for which the figures are intended. Advice is given on publications giving details of the statistical methodology employed. For the covering abstract see IRRD 875261.

  • Availability:
  • Corporate Authors:

    Her Majesty Stationary Office

    49 High Holborn
    London WC1V 6HB,   England 
  • Authors:
    • ROONEY, J
  • Publication Date: 1995-9

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00719868
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • ISBN: 0-11-551716-2
  • Files: ITRD
  • Created Date: Apr 26 1996 12:00AM