Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
the purpose, should be reported. Interacting with the interviewer, however, should aid in improving the data quality attainable by telephone surveys. In recent years, Internet- based interactive survey methods have been proposed by Lyons and McDonald. Web surveys have the advantage of quick implementa- tion and cost reduction in survey administration, and, like CATI, the need for data coding can be eliminated. The obvious advantage is the absence of sampling frames and no control of respondent self- selectivity. Yet improvement in the accuracy of data records can be expected by providing a graphical user interface (GUI) that uses maps, graphic illustrations, and the like. For these interactive surveys, as with all the other methods mentioned so far, however, the accuracy of data on travel route and time is low because the surveys rely on respon- dentsâ memory (Hato et al. 1999). To solve some of these problems, survey procedures based on probe vehicles using GPS units have been devel- oped by Murakami and Wagner (1999) and Zitto et al. (1995). The ability of automatically recording the posi- tion of a vehicle over time has made it possible to observe travel speed and path for a long period. In contrast, a method has been proposed in which sensors are incorpo- rated into spaces of travel to record human travel behav- iors rather than attaching measuring instruments to transport modes for travel. For travel of pedestrians, a survey method using integrated circuit (IC) tags has been proposed. Hato and Asakura (2001) have developed a system that allows the measurement of migration behav- iors of subjects in which interactive information is pro- vided by subjects touching or passing readers for contact- type passive radio frequency identification (RFID) tags and active RFID tags installed in specific spaces in the Matsuyama metropolitan area. However, although such measurement methods have high accuracy for determining spaceâtime positions, they can be said to be survey methods for limited transport modes and spaces of travel. It is difficult to record travel seamlessly to analyze travelersâ travel behaviors trip by trip. There is a common problem that the system cannot measure trips that use other means of transportation than specific transport modes or such trips as transfers. Moreover, the purpose of a trip needs to rely on ques- tionnaire surveys. To solve such problems, a survey method using an automatic position and time recording system on the basis of mobile phones has been proposed by Hato and Asakura (2001). A travel behavior survey method that uses personal handset system automatically measures position data, and several methods have been proposed for recognizing human travel- activity patterns and deter- mining paths on the basis of position data alone. The numbers of observable trips are largerâ and long- term surveys are possible by these methodsâ than with con- ventional survey methods that rely on subjectsâ memory (Figure 1). However, complete automation will increase estimation errors in behavior patterns and paths if the accuracy of position determination is low (Hato et al. 2006). Therefore, it is necessary that the investigator per- form data correction. Because there is a limit to the auto- matic recognition of the facilities of stay by using a 189 DATA- ORIENTED TRAVEL BEHAVIOR ANALYSIS FIGURE 1 Day-to-day path selection activity patterns for one individual.
