Collecting Data with GPS: Those who reject, and those who receive

Abstract

Most surveys using GPS devices to collect data have focused on the device and its effectiveness for collecting reliable and accurate data. Another dimension to be considered with this form of data collection is whether the people who consent to carry the GPS devices differ from those who do not. Using active or passive GPS devices for data collection entails some degree of respondent burden. Even a passive device, must be carried by the respondent and be kept charged as necessary. In this paper, we test the hypothesis that there are statistically significant differences between GPS participants and nonparticipants and consider whether differences may introduce response bias to the GPS survey. The data are from the ongoing Sydney HTS and an associated project comparing the quality of personal travel data collected through the HTS with data collected through the use of GPS devices. Because the participants in the GPS study were recruited from those who already agreed to participate in the HTS, sociodemographic characteristics of both participants and non-participants are available. Identifying potential response bias will assist in developing specialised programs to encourage participation of those most likely to refuse, raising the response rate, and reducing the response bias.