Join us for Patrick Loa’s MASc thesis presentation, “The Design and Empirical Evaluation of the Core-Satellite Framework for Urban Passenger Data Collection.” All are welcome.
For decades, landline-based household travel surveys have played a vital role in the understanding and forecasting of passenger travel behaviour. Primarily due to technological changes, this approach is producing increasingly insufficient data. The continued application of this data collection paradigm typically leads to two key issues.
The first issue is the inability of travel surveys that rely on computer-assisted telephone interviews (CATI) to produce results that are adequately representative of the target population. This issue is reflected in the age distribution of survey respondents, which tends to differ from that of the target population. Specifically, younger members of the population tend to be underrepresented in household travel surveys, while older persons tend to be overrepresented.
The second issue stems from the data requirements of contemporary approaches to analyzing and modelling passenger travel behaviour. While travel demand analysis has traditionally utilized revealed preference data, newer approaches include the use of attitudinal, stated preference, and passive data. It is difficult to collect these types of data through traditional data collection paradigm, as increasing the length of a questionnaire can adversely impact completion rates.
These issues require the modernization of the traditional data collection paradigm. This thesis builds on the work of Goulias et al. (2011), who proposed a “core-satellite” framework for passenger travel data collection. This framework is a modernized version of the traditional approach to data collection, wherein passive data and smaller, more targeted travel surveys (“satellite surveys”) to both complement and supplements the data obtained through traditional household travel surveys. This thesis will define each component of the proposed data collection framework and provide guidelines for the design of the small-sample surveys. Also, this thesis proposes approaches to ensure that the different types of data collected using the proposed framework are compatible with one another.
The value of the two newer components of the data collection framework, namely the satellite surveys and passive data, is demonstrated through two empirical applications. The first uses data obtained through the StudentMoveTO survey, which collected travel diaries and other information from students attending one of Toronto’s four universities. This study investigated the factors that influence the location choices of university students who use public transit to participate in discretionary activities to understand the determinants of the accessibility of said students. The empirical model that was estimated for this purpose was also used to derive utility-based measures of accessibility, which were compared to count-based measures of accessibility. The second empirical application combined zonal transit trip generation values obtained from the Transportation Tomorrow Survey with trip data provided by Uber, via the City of Toronto. This study investigated the impacts of zone-level transit, socio-economic, and land use attributes on the level of Uber and public transit usage in each zone. The findings of both empirical works were used to inform policy recommendations.