ADEQUACY OF CLUSTERED HOME INTERVIEW SAMPLING FOR CALIBRATING A GRAVITY MODEL TRIP DISTRIBUTION FORMULA

Abstract

INCREASED USE OF THE GRAVITY MODEL FORUMLATION HAS BROUGHT ABOUT A WIDE VARIATION IN STUDY PROCEDURES, MANY OF WHICH DEVELOPED WITHOUT THE BENEFIT OF BASIC RESEARCH, ARE CHARACTERIZED BY A REDUCTION IN THE DATA REQUIRED FOR MODEL CALIBRATION. IN SEVERAL INSTANCES ENGINEERS HAVE ADVOCATED USE OF A REDUCED SAMPLE HOME INTERVIEW SURVEY, CLUSTERED IN SELECTED ZONES, TO SUPPLY THE BASIC DATA FOR MODEL CALIBRATION. THE SELECTED ZONES ARE CHOSEN TO REFLECT A RANGE IN THE URBAN CHARACTERISTICS KNOWN TO BE CORRELATED WITH TRAVEL HABITS. FOR EXAMPLE, RESIDENTIAL DENSITY, CAR OWNERSHIP AND DISTANCE FROM THE CBD ARE AMONG THE VARIABLES SUGGESTED IN THE PAST TO DETERMINE WHICH ZONES TO SELECT FOR THE HIGH-RATE CLUSTERED SAMPLING. THIS PAPER REPORTS ON THE RESULTS OF A RESEARCH STUDY DESIGNED TO EVALUATE THE USE OF SMALL SAMPLE CLUSTERED HOME INTERVIEWS AS THE BASIC SOURCE OF GRAVITY MODEL CALIBRATION DATA. BASIC DATA SOURCES FOR THE STUDY WERE THE COMPREHENSIVE 1958 PITTSBURGH HOME INTERVIEW STUDY AND A SPECIAL HIGH SAMPLE RATE CLUSTERED HOME INTERVIEW SURVEY CONDUCTED IN 1960 IN 13 SELECTED ZONES. THE CLUSTERED SAMPLE PROVIDED VERY STABLE TRIP VOLUMES BUT DID NOT PROVIDE GOOD CALIBRATION DATA. THE CLUSTERED SURVEY DID PROVIDE SUFFICIENT TRIP DATA FOR THE DEVELOPMENT OF ZONAL ESTIMATES OF TRIP PRODUCTION BUT IT DID NOT PROVIDE ENOUGH TRIP DATA FOR THE DEVELOPMENT OF TRIP ATTRACTION ESTIMATES. TRAVEL TIME FACTORS DEVELOPED FROM THE CLUSTERED SURVEY WERE SIGNIFICANTLY DIFFERENT FROM THOSE DEVELOPED WITH THE TOTAL STUDY AREA DATA. ONE REASON FOR THE INADEQUACY OF THE CLUSTERED SURVEY DATA IS THE BIAS INHERENT IN CLUSTERED SAMPLING.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 3 citations exist upstream.