This paper covers a detailed examination of the adequacy of the highway performance monitoring system (hpms) sample for making needsestimates at the district level in texas and recommends increases in the current sample based on the results of the examination. To test the accuracy of sample sizes, a simulation model was developed to calculate the errors of a given sample size. Since the fhwa procedure uses average annual daily traffic (aadt) to calculate the requiredsample size, the simulation model was used to compare the accuracy of needs estimates with the assumed accuracy using aadt. It was found that in general, the errors were larger than the assumed error range at the functional class level, but the errors decreased substantially as functional classes were aggregated. The simulation model wasalso used to test the usefulness of stratifying the functional classes by volume group. It was found that in most cases stratifying didimprove the accuracy of the sample estimates. Further, it was foundthat in most cases calculating the sample at the functional class level and distributing the sample to volume groups proportionately bymileage performed better than calculating the sample at the volume group level. This paper appears in transportation research record no. 1236, Transportation data 1989.
Abstract