Use of a Smartphone GPS Application for Recurrent Travel Behavior Data Collection

Abstract

Smartphone-based household travel survey (HTS) studies to date have typically followed the two-part survey process that has historically been used for paper, computer-assisted telephone interviewing, and online HTS. In this two-part survey process, households provide demographic data in a recruit survey (part one) and record trips in a travel diary (part two) often at a later date. The Metropolitan Council, the planning organization serving the Twin Cities metropolitan area in Minnesota, has conducted a pilot study for their cyclical HTS, the Travel Behavior Inventory (TBI), that is one of the first large-scale fields of an all-in-one smartphone HTS design. For the 2018 TBI pilot, the traditional two-part survey was merged into a continuous survey experience within a smartphone app. The TBI pilot used a split sample to test this all-in-one design against a traditional two-part smartphone survey design. For the all-in-one design, households were invited to sign in directly to the smartphone application instead of first recruiting online or by phone. The pilot results provide a direct comparison of the two-part and all-in-one designs at the household-, person-, and trip-levels. The results showed a lower overall recruit and completion rate for the all-in-one design but showed clear promise for increasing representation of younger and lower-income populations—traditionally hard-to-reach groups who completed at a higher rate with all-in-one. The authors discuss several factors which may have contributed to the lower overall completion rate and describe planned updates for future waves of the TBI aimed at improving overall response while maintaining the developments that have improved representation from hard-to-reach groups.