Incorporating Geographic Correlation When Sampling a Transportation Network

Abstract

Estimating characteristics of motor vehicles and their occupants based on sampling road segments and vehicles on those road segments is inherently a statistical sampling problem. An important aspect to sampling transportation networks that has not been previously examined in detail is addressed here; it incorporates the correlation between road segments in a traffic network into the sampling design. An approach for estimating the correlation between links in a transportation network is described, and a recommended approach for incorporating this correlation into the sampling design is presented. Also, the implementation and evaluation of these approaches relative to simple random sampling without replacement is illustrated by using four sampling scenarios typically encountered when sampling a transportation network.

Get full access to this article

View all access options for this article.

References

Kinateder

J. G.

, McMillan

N. J.

, Orban

J. E.

, Skarpness

B. O.

, and Wells

. Sampling Designs and Estimators for Monitoring Vehicle Characteristics When Observation Capacity Is Limited. In Transportation Research Record 1581, TRB, National Research Council, Washington, D.C., 1997.

Thompson

S. K.

Sampling. John Wiley and Sons, New York, 1992.

Brewer

K. R. W.

, and Hanif

. Sampling with Unequal Probabilities. In Lecture Notes in Statistics, Springer-Verlag, New York, 1982.

Chaudhuri

, and Vos

J. W. E.

. Unified Theory and Strategies of Survey Sampling. North Holland Press, 1988.