Addressing Barriers to Reflection in Physical Activity Tracking

Abstract

Self-tracking is the practice of gathering personal data, typically using wearable devices and smartphones to gain knowledge about personal health and improve one’s wellbeing. Nowadays, many people track data related to their physical activity to set goals and track progress. Often, people track to discover behavioural and contextual patterns to adjust their behaviours and make more informed decisions. Such endeavour requires self-trackers to make sense of their data and reflect, creating space for learning, taking actions and changing behaviours. Reflection has been described as a core process in self-tracking. Consequently, prior work evaluated techniques to motivate reflection, such as visualisations, providing contextual information, and providing reflective prompts to encourage self-reporting. However, self-trackers feel demotivated and even abandon self-tracking tools when they do not know the appropriate goals or do not find meaningful insights from their data. These unmet expectations cause frustration, disappointment and uncertainty. Although resolving uncertainty is considered one of the purposes of reflection, the impact of uncertainty on reflection in self-tracking has not been yet explored. Therefore, this thesis addresses such a gap by investigating the challenges of reflection under uncertainty in self-tracking through a qualitative study. The findings suggest that uncertainty due to unmet expectations and reduced meaning hinders reflection. Interestingly, self-trackers leverage uncertainty to facilitate reflection. They use strategies such as identifying relationships between their data and other non-tracked factors and reconstructing activities to recall events that potentially affect their data. As a result, two interventions emerging from these adaptive strategies were designed, implemented and evaluated. The first intervention calibrated self-trackers’ expectations by incorporating their self-efficacy in the tracking process and moderating their physical activity goals based on their self-efficacy. The second intervention fostered reconstruction on physical activity data with the aid of spatio-temporal and contextual cues. Both interventions adopted an explicit approach to support reflection on physical activity data, goals and the contextual factors influencing them. The results suggest that promoting a sustained reflective practice on data and expectations helped self-trackers set more attainable goals. Additionally, moderating physical activity goals increased goal achievement and satisfaction with physical activity. On the other hand, reflection was enriched when suitable contextual cues were provided. Explicitly fostering a reflective practice holds great potential in supporting people’s commitments to tracking and achieving their desired tracking goals.