THE USE OF MULTIPLE REGRESSION AND CORRELATION IN TEST ROAD DATA

INQUIRIES PERTAINING TO AN EXPERIMENTAL ROAD TEST ARE DISCUSSED: (1) WHAT GOVERNS THE SELECTION OF DEPENDENT AND INDEPENDENT VARIABLES? (2) HOW MAY MATHEMATICAL MODELS BE CONCEIVED FOR EXPRESSING THE CONCLUSIONS? (3) WHAT LEVELS OF THE INDEPENDENT VARIABLES NEED TO BE INVESTIGATED? (4) WHAT TECHNIQUES MAY BE USED FOR ESTIMATING CONSTANTS IN THE MATHEMATICAL MODELS? (5) HOW PRECISE ARE THE ESTIMATED CONSTANTS DETERMINED BY THE EXPERIMENT? (6) HOW VALID IS THE MATHEMATICAL MODEL: (7) HOW FAR, IN SPACE AND TIME, MAY THE EXPERIMENTAL RESULTS BE GENERALIZED? EXPERIMENTAL FINDINGS CAN BE CHARACTERIZED BY MATHEMATICAL MODELS. NO EXTEMPORANEOUS MODEL SHOULD BE CONSIDERED AS REPRESENTING ANY MORE THAN POSSIBLE POINTS OF DEPARTURE FOR FURTHER EFFORTS TO INTERRELATE THE MANY RESULTS FROM A ROAD TEST. THE EXPERIMENTAL OBJECTIVES, THE EXPERIMENTAL DESIGN AND THE DATA ANALYSIS ARE SO CLOSELY CONNECTED THAT NONE OF THESE THREE ASPECTS CAN BE ADEQUATELY DESCRIBED WITHOUT REFERENCE TO THE OTHER TWO. A POSSIBLE ANALYTICAL APPROACH IS OUTLINED, ASSUMING THAT THE OVER-ALL EXPERIMENTAL OBJECTIVES ARE TO PRODUCE CERTAIN RELATIONSHIPS BETWEEN OBSERVED BEHAVIOR VARIABLES AND THOSE VARIABLES WHICH ARE EITHER SET BY THE EXPERIMENTAL DESIGN OR WHICH ARISE THROUGH THE FORCES OF NATURE DURING THE EXPERIMENT. AN EFFORT IS MADE TO GIVE A DETAILED DISCUSSION OF THE CONCEPTS AND ROLES OF RANDOMIZATION AND REPLICATION. DATA FROM THE WASHO ROAD TEST ARE USED TO DEMONSTRATE ANALYTICAL TECHNIQUES. /AUTHOR/

• Record URL:
  https://onlinepubs.trb.org/Onlinepubs/hrbproceedings/35/35-020.pdf
• Supplemental Notes:
  • Vol 35, pp 268-298, 6 FIG, 8 TAB, 3 REF. Distribution, posting, or copying of this PDF is strictly prohibited without written permission of the Transportation Research Board of the National Academy of Sciences. Unless otherwise indicated, all materials in this PDF are copyrighted by the National Academy of Sciences. Copyright © National Academy of Sciences. All rights reserved.
• Authors:
  • Irick, P E
  • Leathers, R C
  • Lill, R A
  • MERTZ, W L
• Publication Date: 1956

Media Info

• Media Type: Digital/other
• Monograph Title: Proceedings of the Thirty-Fifth Annual Meeting of the Highway Research Board, Washington, D.C., January 17-20, 1956
• Serial:
  • Highway Research Board Proceedings
  • Volume: 35
  • Publisher: Highway Research Board

Subject/Index Terms

• TRT Terms: Correlation analysis; Experimental roads; Mathematical models; Randomization; Regression analysis; Road tests; Testing; Variables
• Uncontrolled Terms: Experimental data
• Old TRIS Terms: Replicating
• Subject Areas: Data and Information Technology; Highways; Pavements; Vehicles and Equipment;

Filing Info

• Accession Number: 00207342
• Record Type: Publication
• Files: TRIS, TRB
• Created Date: Aug 8 1971 12:00AM