Using Mobile Phones to Reduce Missing Data in Youth Activity Monitoring Studies
Funded by National Heart Lung Blood Institute Grant R21HL108018
(Dunton and Intille, Pls)
Physical inactivity during adolescence increases risk for a number of serious health conditions. Surveillance, epidemiological, and intervention studies seeking to increase physical activity (PA) and/or decrease sedentary behaviors (SB) in adolescents could be enhanced through more informative and accurate methods of measuring these behaviors. Concerns about the validity of retrospective self-report of PA and SB have led to increasing use of objective measures, e.g., accelerometers and Global Positioning System (GPS) loggers, which can be complemented by subjective or contextual information using self-initiated event-contingent electronic ecological momentary assessment (EMA) after exercise or other critical activities. Regardless of the technique used, device non-wear, equipment malfunction, and participant non-response result in missing and ambiguous data that complicate statistical analysis. Adolescents recruited into objective PA monitoring studies will increasingly have “smartphones,” which are miniature computers with built in motion sensors and location-finding capabilities. Sophisticated programs (i.e., “apps”) can be easily installed on the phones. The phones are rarely far from the adolescents, and adolescent affinity for the phones creates new opportunities for activity monitoring in surveillance and intervention studies. The overall objective of this project was to develop new software for common mobile phones that can automatically detect and elicit information about activity, location change, and data loss episodes through real-time EMA or end of day recall. This technology supplemented objective monitors already used today, with minimal additional cost. Our solution will have three novel components:
(1) A phone “app” that used the mobile device’s built-in sensors to detect major transitions in type of movement or location, after which timely, context-sensitive questions and reminders are triggered that will provide information about activity, location change and data loss episodes, regardless of whether built-in or external objective monitors are used,
(2) A second phone “app” that had an entertaining, game-like feel and allowed adolescents to interactively “fill in gaps” in their own data at the end of the day using cues from automatically-detected major transitions to explain activity, location change, and data loss, and (3) Server-side software that remotely collected data from the two apps in real-time and provided researchers with a cost-efficient way to reduce missing data and improve characterization of transition in activity.
(3) Server-side software that remotely collected data from the two apps in real-time and provided researchers with a cost-efficient way to reduce missing data and improve characterization of transition in activity.