ESTIMATING LONG-DISTANCE TRAVEL BEHAVIOR FROM THE MOST RECENT TRIP

Abstract

In the United States and Europe, increasing attention is being paid to the measurement of long-distance travel behavior. Such measurement poses special problems that are not encountered in the measurement of daily mobility in urban areas, which has been the focus of attention of most previous household travel surveys. Because long-distance travel is a relatively rare event, a major issue has been the selection of the period of observation. Unlike daily mobility surveys, where the most common survey period is a single 24-hr period, long-distance travel surveys have used survey periods from several weeks to several months. However, selection of too short a period means that many respondents have no long-distance trips to report, while selection of too long a period means that frequent long-distance travelers have many trips to report, some of which occurred a long time before the conduct of the survey. This results in problems of recall or, in the extreme, problems of non-reported trips or even non-response. As a result, many surveys have used survey periods of 2 to 3 months. However, this method can result in under-reporting of trips from infrequent travelers, who are recorded as non-travelers, and possible under-reporting from frequent travelers due to the recall and response problems mentioned above. In response to the above-mentioned problems, an alternative survey design is proposed by the authors which seeks to obtain information from all respondents about the most recent long-distance trip they have made, irrespective of when that trip was made. For frequent travelers that trip may have occurred today, while for infrequent travelers the trip may have occurred months or even years ago. Each respondent reports the details of only their most recent trip. In this way, trip information is obtained from all respondents, while limiting the response burden for the frequent long-distance travelers. The paper describes, theoretically and empirically, how the data obtained from this survey design can be used to obtain unbiased estimates of long-distance trip rates. It then compares these results with what would have been obtained from more conventional long-distance travel survey designs. It then suggests modifications to the basic design of long-distance travel surveys.

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 — 8 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.