Smartphone Travel Survey-based Adaptive Mode Choice Experiment

This study presents the design and pilot evaluation of a smartphone-based adaptive survey platform that combines semi-passive GPS travel diaries with real-time, context-aware generation of stated preference (SP) scenarios. The custom mobile application automatically detects trips, infers mode and purpose, and triggers SP choice tasks directly from observed trip attributes, with the aim of reducing recall bias and respondent burden compared with conventional SP survey formats. The platform was piloted over a two-week period in Greater Montreal and Toronto, yielding more than 2,500 passively logged trips and 155 automatically generated SP scenarios. Mean response time per SP task was approximately 6 seconds, most users completed all prompted SP questionnaires, and more than half completed the embedded user-profile survey. Comparing users who did and did not receive context-aware SP prompts indicates that dynamic, trip-linked notifications are associated with higher overall survey engagement, even among participants who occasionally skip SP tasks. These results suggest the technical feasibility and operational viability of fully integrated, smartphone-based SP scenario generation and provide a practical template for future travel survey designs that seek to enhance realism, manage respondent burden, and mitigate survey fatigue.