CAREER: Adaptive, Collaborative User Interfaces for Chronically Ill Adolescents' Personal Data Management

这个项目是关于开发跟踪个人信息的协作技术，在帮助患有慢性病的青少年跟踪数据以管理他们的疾病的背景下。越来越多的人可以通过健康跟踪器和智能手机传感器来推断睡眠时间和体力活动等数据，但症状跟踪、疼痛和睡眠质量等个人体验仍然需要手动跟踪。这很费力，特别是对于自理能力不成熟的青少年。因此，该项目的一个关键目标是设计合作策略来收集自我报告数据，这些策略改变了收集数据的时间、方法和人员，以最大限度地提高为特定青少年收集的数据的质量和一致性，考虑到他们的条件、自我照顾能力和家庭状况。第二个关键目标是将这些自我报告的数据与智能手机和健身跟踪器感知的数据进行匹配，开发算法和可视化显示，帮助青少年和家庭一起审查数据，并就青少年的状况做出正确的决定。对于数据收集和审查，开发的工具都需要对一些问题敏感，例如父母和青少年对病情的看法不同，以及青少年独立和建立个人身份的需求。为了做到这一点，研究团队将与患有慢性病的青少年家庭密切合作，对家庭目前如何管理这种跟踪进行实地研究，一起和单独采访青少年和父母，以及在他们建立和改进这些工具时使用原型工具进行短期研究。该项目将对这些工具进行长期评估，以了解它们如何影响青少年与自我护理相关的态度和能力，以及他们如何与家人和医生合作。这项工作将为首席研究员的人机交互和个人健康信息学课程提供信息，并支持向K-12学生开展夏令营外展活动，利用健康信息学提高他们对STEM研究的兴趣。研究工作将分三个主要阶段进行，与亚特兰大地区多个医疗中心的青少年患者和医生密切合作。第一阶段包括围绕第一个目标的形成性工作，即获取个人健康护理信息并定义体现协作数据收集、呈现和管理技术的移动传感应用程序的需求。这将涉及对青少年、父母和照顾者的访谈研究，以了解他们对协作式个人健康管理技术的态度和看法。它还将利用通过生态瞬时评估和经验采样收集数据的战略的试点部署，频率和时间部分由自动感知的数据指导。第二阶段旨在建立一个功能齐全的原型系统，包括三个主要组成部分：(1)实施在第一阶段被证明有效的数据收集战略；(2)开发利用遥感数据对自我报告进行三角测量的增强数据融合技术；(3)创建用于分析融合数据的算法和支持这些分析的可视化。然后，这个原型将推动第三阶段的研究，这是一项为期18个月的研究，有60个家庭分成三组：一组将使用完整的原型，一组用现有的标准抽样策略取代协作启发策略，另一组作为对照组。分析将考察该系统如何影响青少年和父母的隐私问题，总体上接受自我报告和感知的移动数据收集，特别是使用该工具，使用该工具所需的努力，以及青少年的自我效能和健康控制点。评价将利用标准调查工具进行量化测量，并在青少年随着时间的推移使用这些工具时对他们进行访谈。

This project is about developing collaborative technologies for tracking personal information, in the context of helping teenagers with chronic health conditions track data to manage their illnesses. Increasingly, data such as sleep duration and physical activity can be inferred through fitness trackers and smartphone sensors, but personal experiences such as symptom tracking, pain, and sleep quality still need to be tracked manually. This is effortful, especially for teenagers with immature self-care abilities. Thus, one key aim of the project is to design collaborative strategies for collecting self-report data that vary the timing, method, and person used to collect the data to maximize the quality and consistency of data collected for particular teens, given their condition, self-care abilities, and family situation. A second key aim is to align this self-report data with data sensed by smartphones and fitness trackers, developing algorithms and visual displays that help teens and families review the data together and make good decisions about the teens' condition. For both data collection and review, the tools developed will need to be sensitive to problems such as differences in parents' and teens' views of the condition and teens' needs to become independent and establish personal identity. To do this, the research team will work very closely with families with chronically-ill teens, doing field studies of how families currently manage this sort of tracking, interviews with teens and parents both together and separately, and short-term studies with prototype tools as they build and improve them. The project will end with a long-term evaluation of the tools to see how they affect teens' attitudes and abilities related to self-care, as well as how they work with their families and doctors. The work will inform the lead investigator's courses on human-computer interaction and personal health informatics, and support the development of summer camp outreach activities to K-12 students, using health informatics to increase their interest in STEM research.The research work will proceed in three main phases, working closely with adolescent patients and doctors at multiple medical centers in the Atlanta area. The first phase involves formative work around the first aim of eliciting personal health care information and defining requirements for mobile sensing applications that embody collaborative data collection, presentation, and management techniques. This will involve interview studies with teens, parents, and caregivers, to elicit their attitudes toward and perceptions of collaborative personal health management technologies. It will also utilize pilot deployments of strategies for collecting data through ecological momentary assessments and experience sampling, with frequency and timing guided in part by automatically-sensed data. The second phase aims at building a fully functional prototype system, with three main components: (1) implementation of data collection strategies proven effective in the first phase; (2) development of augmented data fusion techniques that triangulate self-reported with sensed data; and (3) creation of both algorithms for analyzing the fused data and visualizations that support these analyses. This prototype will then drive the third phase, an 18-month study with 60 families in three groups: one that will use the full prototype, one a version that replaces the collaborative elicitation strategies with an existing, standard sampling strategy, and one as a control group. The analysis will examine how the system affects both adolescents' and parents' privacy concerns, acceptance of both self-reported and sensed mobile data collection in general and with this tool in particular, the effort required to use the tool, and adolescents' self-efficacy and health locus of control. The evaluation will utilize both standard survey instruments for quantitative measures and interviews with adolescents as they use the tools over time.