Computer-Assisted Web-Based Personal Interview Survey for Capturing Urban Mobility: A Contextual Approach in Resource-Constrained Settings

Abstract

This study demonstrates the operational deployment and feasibility of a computer-assisted web-based personal interview (CAWPI) approach in a resource-constrained urban setting. Piloted in Chattogram, Bangladesh, the method hybridizes face-to-face interviewing with a low-cost web platform under a bring-your-own-device (BYOD) strategy. It shows how such tools can be pragmatically adapted where pure computer-assisted web interviewing (CAWI) or an agency-provided device computer-assisted personal interviewing (CAPI) is infeasible. Data were collected from 550 households across 33 administrative wards, capturing household demographics, vehicle ownership, and daily travel behavior. The survey recorded travel history from 2,104 individuals, covering 3,884 trips. The CAWPI platform was optimized for low-bandwidth settings and integrated open-source tools, such as OpenStreetMap, for embedding geolocation during the survey data collection. Enumerators received training and real-time support throughout the survey. The effort achieved a 96% completion rate, with strong data quality reflected in consistent and logical trip timing, mode use, and activity patterns. The data were validated against nationally representative surveys. Challenges included limited geospatial search, occasional connectivity issues, and human input errors. The study identified best practices, including localized enumerator deployment, flexible data entry workflows, and continuous technical support. Findings suggest CAWPI has the potential to generate demographically representative and behaviorally coherent mobility data while remaining operationally feasible in resource-constrained settings. The approach is demonstrated as a contextually viable implementation for urban travel surveys in settings where structured data collection efforts are infrequent, particularly in the Global South, with future potential for offline capabilities and speech-to-text integration.