Identifying latent mode-use propensity segments in an all-AV era

• Published in: Transportation research. Part A, Policy and practice Vol. 130; pp. 192 - 207
• Main Authors: Kim, Sung Hoo, Circella, Giovanni, Mokhtarian, Patricia L.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2019
• Subjects: Autonomous vehicles; Behavioral response; Factor analysis; Latent class cluster analysis; Mode-use propensity; Perceptions; Perceptions; Mode-use propensity; Factor analysis; Behavioral response; Latent class cluster analysis; Autonomous vehicles
• ISSN: 0965-8564; 1879-2375
• DOI: 10.1016/j.tra.2019.09.015

This study offers an early glimpse of how individuals perceive the advantages/disadvantages of AVs, their mode-use intentions, and potential market segments with respect to mode use, should AVs eventually become the only way to travel by car. To do so, we implemented a statewide survey of Georgia residents (N = 2890) and using that data, we applied factor analyses to two blocks of AV-related statements. The first block measured 12 perceptions of AVs, and yielded two psychological constructs: AV pros (advantages/ benefits) and AV overuse cons (negative outcomes specifically associated with the excessive use of AVs). The second block of statements measured respondents’ inclinations between AV and non-AV options for 12 hypothetical transportation “needs”, and factor analysis identified four mode-use propensity constructs: AV(-inclined) over walk/bike, AV over flight, zero-occupant AV over occupied AV, and AV over transit. The main goal of the paper was to segment the sample on the basis of these four mode-use propensities, to identify clusters with similar propensity profiles or response vectors. We applied latent class cluster analysis to do so, and identified seven potential market segments: some preferring AV options in general, others preferring non-AV options or having unique propensity patterns based on certain contexts (e.g. long distance travel and vehicle occupancy). In the model, socio-demographics, geography, attitudes, and perceptions of AVs help characterize those market segments, and this provides a basis for deeper interpretation and consideration of policy implications.