NINE ESTIMATORS OF GAP-ACCEPTANCE PARAMETERS

THERE IS A VAST ARRAY OF DIFFERENT METHODS AVAILABLE FOR ANALYZING GAP-ACCEPTANCE DATA. THIS PAPER CONTAINS A COMPARATIVE STUDY OF A SELECTION OF NINE OF THESE METHODS. FOR THE PURPOSE OF COMPARISON, THE FOLLOWING SIMPLE MODEL IS USED. EACH PERSON HAS A CRITICAL SIZE OF GAP. THESE CRITICAL GAPS VARY FROM PERSON TO PERSON. ONE PERSON ALWAYS REJECTS GAPS LESS THAN HIS OR HER CRITICAL GAP, AND ALWAYS ACCEPTS LARGER GAPS. DESIRABLE PROPERTIES OF ESTIMATORS ARE DISCUSSED. EACH METHOD IN TURN IS DESCRIBED. TO COMPARE METHODS, 100 SETS OF ARTIFICIAL DATA HAVE BEEN GENERATED. EACH METHOD HAS BEEN APPLIED TO EACH SET OF DATA. OF THE NINE METHODS, FOUR ONLY USE INFORMATION ON LAGS ACCEPTED OR REJECTED, WHILE THE OTHER FIVE METHODS USE ALL LAGS AND GAPS CONSIDERED. FIVE OF THE METHODS GIVE RESULTS WHICH ARE SERIOUSLY BIASSED AND CANNOT BE RECOMMENDED UNLESS CORRECTIONS FOR THE BIAS ARE MADE. THE METHOD SUGGESTED BY ASHWORTH IS FAIRLY EFFICIENT AND CAN BE RECOMMENDED PROVIDED THAT THE OFFERED GAPS ARE NOT HIGHLY CORRELATED. THE METHOD OF MAXIMUM LIKELIHOOD IS SHOWN TO GIVE SATISFACTORY RESULTS. A FEW IDEAS FOR HANDLING MORE REALISTIC MODELS ARE GIVEN AT THE END OF THE PAPER.

  • Corporate Authors:

    Traffic Flow and Transportation

    ,    
  • Authors:
    • Miller, A J
  • Publication Date: 0

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00227402
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Apr 11 1974 12:00AM