Assessing Data Quality of Unvalidated Trip Segmentation Records from Smartphone-based Travel Survey

This paper examines the data quality of unvalidated trip segmentation records from smartphone-based travel surveys. The study addresses challenges related to incomplete responses and the high proportion of unvalidated trips in large-scale implementations of smartphone-based travel surveys. Three travel survey apps and Google Location History were used concurrently to record the same trips over multiple days. Unvalidated records were compared to validated records and analyzed across different apps. Data quality was assessed based on trip identification, segmentation, mode detection, and purpose imputation. The findings offer insights into improving the quality of unvalidated travel records, reducing respondent burden, and enhancing data usability for travel behavior analysis.