STATISTICAL METHODS APPLIED TO HIGHWAY-RESEARCH EXPERIMENTATION /WITH DISCUSSION/

THE APPLICATION OF THREE OF FOUR BASIC STATISTICAL PRINCIPLES TO A HYPOTHETICAL PROBLEM IN TESTING PAVEMENT DESIGN IS DEMONSTRATED. AN EXPERIMENT IS DESIGNED IN WHICH A NEW TYPE OF PAVEMENT IS COMPARED WITH A STANDARD ONE, AND A NEW TYPE OF SUBBASE IS COMPARED WITH A STANDARD SUBBASE. THE EXPERIMENT IS PLANNED FOR THREE LOCATIONS. SIMPLE HYPOTHETICAL DATA CONSISTING OF MAINTENANCE COSTS IN DOLLARS PER MILE OF HIGHWAY PER YEAR ARE USED FOR ILLUSTRATIVE PURPOSES AND AN ANALYSIS OF THE DATA IS GIVEN IN SOME DETAIL. ONE VALUABLE PROPERTY OF THIS EXPERIMENTAL DESIGN IS THAT THE RESULTS NOT ONLY AFFORD THE REQUIRED COMPARISONS BETWEEN THE NEW AND STANDARD TYPES OF PAVEMENT AND SUBBASE, THEY ALSO AFFORD PRECISE MEASURES OF THE RELIABILITIES OF THESE COMPARISONS. ANOTHER USEFUL FEATURE OF THE EXPERIMENT IS THAT THE RELIABILITY OF EACH OF THE DESIRED COMPARISONS IS MAXIMIZED. SIMILAR ADVANTAGES AFFORDED BY STATISTICAL DESIGNS ARE ALSO INDICATED. THESE METHODS AND THEORY ARE LARGELY CONCERNED WITH THE PROBLEMS OF DESIGNING AND ANALYZING EXPERIMENTS AND MAY BE OF CONSIDERABLE VALUE IN THE FIELD OF HIGHWAY RESEARCH. /AUTHOR/

  • Supplemental Notes:
    • Vol 33, pp 112-120, 5 TAB, 6 REF
  • Authors:
    • Duncan, D B
    • Irick, P
    • Goode, R B
  • Publication Date: 1954

Media Info

  • Monograph Title: Proceedings of the Thirty-Third Annual Meeting of the Highway Research Board, Washington, D.C., January 12-15, 1954
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00205918
  • Record Type: Publication
  • Files: TRIS, TRB
  • Created Date: May 16 1972 12:00AM