Validating Smartphone- and Computer-based Technologies with GPS for Activity Tracking

Abstract

The space in which we live and complete our daily activities such as shopping, eating, and working is known as the activity space. Measuring activity space can provide insights into the relation between built environment and health outcomes. Historically, geospatial research in public health was conducted using paper-and-pencil travel logs. It has since moved to Global Positioning System (GPS)-enabled instruments and to computer-assisted interviews. The purpose of this study was to validate a newly developed computer-assisted instrument, Karma, against a traditionally used GPS instrument and a smartphone-based application, MapMyRun (MMR), to study activity space. 12 participants, recruited in the spring of 2018, were asked to collect data using the three instruments over the same three days. Four primary outcome variables were tested for each participant-day (n=29): dwell point count, active dwell duration (in minutes), travel time (in minutes), and track length (in kilometers). Statistically significant correlations were observed for active dwell duration, travel time and track length from Karma with both GPS (satellite-based instrument) and MMR (smartphone-based instrument). The only exception was the dwell point count variable that did not show significant correlation between Karma and GPS. Additional analyses suggested slightly different travel patterns for food shopping days vs. non-food shopping days and for weekends vs. weekdays. Limited sample size did not allow further stratified analyses. Despite a small sample size, the present findings suggest potential use of Karma to measure activity space in lieu of GPS instruments. Further studies are needed to test the use of Karma with a larger sample size and in population segments that depend on modes of transportation other than car for their primary travel.