|dc.description.abstract||In order to support people's everyday health and wellness goals, health practitioners and organizations are embracing a more holistic approach to medicine---supporting patients both as individuals and members of their families and communities, and meeting people where they are: at home, work, and school. This 'everyday' approach to health has been enabled by new technologies, both dedicated-devices and services designed specifically for health sensing and feedback -- and multipurpose --such as smartphones and broadband-connected computers. Our physical relationship with computing has also become more intimate, and personal health devices can now track and report an unprecedented amount of information about our bodies, following their users around to an extent no doctor, coach or dietitian ever could.
But we still have much to learn about how pervasive health devices can actually help promote the adoption of new health practices in daily life. Once they're `in the wild,' such devices interact with their users, but also the physical, social and political worlds in which those users live. These external factors---such as the walkablity of a person's neighborhood or the social acceptability of exercise and fitness activities---play a significant role in people's ability to change their health behaviors and sustain that change. Specifically, social theories of behavior change suggest that peer support may be critical in changing health attitudes and behaviors. These theories---Social Support Theory, Social Cognitive Theory and Social Comparison Theory among them---offer both larger frameworks for understanding the social influences of health behavior change and specific mechanisms by which that behavior change could be supported through interpersonal interaction. However, we are only beginning to understand the role that pervasive health technologies can play in supporting and mediating social interaction to motivate people's exploration and adoption of healthy behaviors.
In this dissertation I seek to better understand how social computing technologies can help people help each other live healthier lives. I ground my research in a participant-led investigation of a specific population and condition: adolescents and obesity prevention. I want to understand how social behavior change theories from psychology and sociology apply to pervasive social health technology. Which mechanisms work and why? How does introducing a pervasive social health system into a community affect individuals' behaviors and attitudes towards their health? Finally, I want to contribute back to those theories, testing their effectiveness in novel technologically mediated situations.
Adolescent obesity is a particularly salient domain in which to study these issues. In the last 30 years, adolescent obesity rates in the US alone have tripled, and although they have leveled off in recent years they remain elevated compared to historical norms. Habits formed during adolescence can have lifelong effects, and health promotion research shows that even the simple act of walking more each day has lasting benefits. Everyday health and fitness research in HCI has generally focused on social comparison and "gamified" competition. This is especially true in studies focused on adolescents and teens. However, both theory from social psychology and evidence from the health promotion community suggest that these direct egocentric models of behavior change may be limited in scope: they may only work for certain kinds of people, and their effects may be short-lived once the competitive framework is removed. I see an opportunity for a different approach: social tools for everyday adolescent health. These systems, embedded in existing school and community practices, can leverage scalable, non-competitive social interaction to catalyze positive perceptions of physical activity and social support for fitness, while remaining grounded in the local environment.
Over the last several years I have completed a series of field engagements with middle school students in the Atlanta area. I have focused on students in a majority-minority low-income community in the Atlanta metropolitan area facing above-average adult obesity levels, and I have involved the students as informants throughout the design process. In this dissertation, I report findings based on a series of participatory design-based formative explorations; the iterative design of a pedometer-based pervasive health system to test these theories in practice; and the deployment of this system---StepStream---in three configurations: a prototype deployment, a `self-tracking' deployment, and a `social' deployment.
In this dissertation, I test the following thesis: A school-based social fitness approach to everyday adolescent health can positively influence offline health behaviors in real-world settings. Furthermore, a noncompetitive social fitness system can perform comparably in attitude and behavior change to more competitive or direct-comparison systems, especially for those most in need of behavior change}. I make the following contributions: (1) The identification of tensions and priorities for the design of everyday health systems for adolescents; (2) A design overview of StepStream, a social tool for everyday adolescent health;
(3) A description of StepStream's deployment from a socio-technical perspective, describing the intervention as a school-based pervasive computing system; (4) An empirical study of a noncompetitive awareness system for physical activity; (5) A comparison of this system in two configurations in two different middle schools; (6) An analysis of observational learning and collective efficacy in a pervasive health system.||