This workshop position paper argues for building interfaces that integrate people's real-world needs in personal health tracking. To achieve this, this paper proposes building user interfaces that support people's curiosity towards self-knowledge and their desire to socialize goal-oriented progress. Drawing on examples from popular applications and recent social computing systems, this paper suggests adding more learning and collaborative tools to researchers' prototyping toolbox.
let’s build user interfaces using ideas from learning and social psychology. Two such levers Innovations in hardware and software have include people’s curiosity in self-knowledge created immense opportunities for citizens and and their desire to socialize goal-oriented progress. healthcare experts to track useful condition- To help get started with, this paper draws specific health data. However, real world de- from both recent research systems and popployments of such tools consistently finds pat- ular commercial applications with successful terns of non-use. Virtual rewards and gami- field deployments. ifcation elements have demonstrated mixed success [1]. This position paper argues for building interfaces that integrate people’s real- 2. Supporting Curiosity world needs in personal health tracking. Promis- and Knowledge ing approaches include serious games that integrate real-world contexts and digital coaches Production that support goal-driven progress [2, 3]. However, many such approaches use costly and Back in the 90s, learner-centered design and cumbersome user-facing elements, like 3D in- education proposed designing interfaces that terfaces for handheld platforms or automated support learning in the classroom and the proagents; further rigorous evaluation needs to fessional setting [4]. Fast forward three decades, check for the real-world success of such ideas. learning resources and collaboration platforms This paper proposes a complementary approacha:re ubiquitous on the internet; however, learnercentered ideas have had limited efect on the design of prototypes across many domains, J1o3i-n1t7,P2r0o2ce1e,dCionlglesgoefSthtaetiAoCn,MUSIUAI 2021 Workshops, April including healthcare. One glaring omission " vineet@seas.harvard.edu (V. Pandey) is the absence of structured support for informal learning beyond youtube (or similar) videos. Health tools and apps have focused on collecting data from people but less on build© 2020 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 InterCPWrEooUrckReshdoinpgs IhStpN:/c1e6u1r3-w-0s.o7r3g (nCCatEiEoUnUaRlR(C-CWBYSW4..o0o).rrgk)shop Proceedings ing capacity in people. ful insights, rather than just sharing data. Widely
Capacity-building tools require appropri- deployed health apps and tools provide the ate learning support for complex activities that opportunity to convert people’s questions and draw on people’s needs like improving self- intuitions into valid scientific knowledge that knowledge, meeting goals, and contributing might be useful for individuals and for broader to society. First, many people–e.g. Quanti- society. How do we meet these three objecifed Selfers–demonstrate a need for improved tives? self-knowledge about how their choices af- Making progress in a new task requires apfect their health. This curiosity provides the propriate resources. Purely conceptual knowlopportunity to create a setup for problem- edge might provide lesser gains in people’s based learning: the person is primed to re- quality of work than providing concrete proceive guidance on how to draw value from cedural guidance [5]. Commercial apps expersonal health tools. Second, people draw emplify this need for complex learning supvalue from meeting personal goals; e.g., sim- port. For instance, sitting still with few thoughts ple curiosity about meditation might get peo- for 20 minutes during a meditation session ple started; making progress towards their goalsrarely comes simply to many. To support peo(e.g. developing confidence or reducing anx- ple in this challenging task, the Headspace iety) might keep them going. Third, people’s meditation app (headspace.com) (Fig 1a); proaltruistic nature—as demonstrated in citizen vides 1) conceptual support with an extensive science projects— can also contribute useful catalog of learning resources for both novices knowledge that can further our understand- and experts, 2) procedural support with guided ing of health and well-being. Personal con- meditation, 3) emotional support with monks’ text provides a great setup to educate people reflections, and 4) motivational support with about scientific knowledge production; inter- celebrities’ words of wisdom. ested people might contribute back with use- Personal health provides an immediately accessible and applicable area for people to ple to set annual book reading challenges that learn and implement scientific skills. While are then broadcast on their friends’ goodreads people’s contextual knowledge about them- feed (Fig 1c); seeing others’ goals and the books selves can yield potentially novel insights, cit- they’re reading provides further motivation. izen participation in research rarely goes be- Such goals can also be tinged with friendly yond expert-defined tasks. Structuring knowl- competition. Consider Fitbit’s weekly steps edge production to make it directly useful for challenge (Fig 1b);: people share weekly steps people might improve citizen contributions. counts with other (willing) friends. ComparPeople have gone beyond being data donors ing the actual number of steps is both fun to creating personally-meaningful insights by and motivating as the leaderboard changes running experiments using systems like Tummywith each day. Another way is to build social Trials and Galileo [6, 7]. By supporting sci- accountability with more rigid mechanisms entific explorations, such tools act as in-situ than what we have now. Prior social computcognition aids that support "thinking through ing work demonstrates two insights: 1) a bus doing". Other avenues to suppot scientific work stop model–where participants move through includes testing existing scientific theories for learning stages in cohorts–provides greater themselves, e.g. about running and heart rate. retention and achievement and, 2) explicitly setting activation thresholds make reciprocity obvious: failure to comply will doom the entire group’s eforts [8, 9].
Many current health applications support shar- Acknowledgments
ing updates on prominent social media
platThanks to the anonymous reviewers for their
forms like Facebook and Twitter. Such
"reachthoughtful comments that improved this
prooptimal" eforts are useful in reaching broad
posal’s link to existing research.
audiences but they are sub-optimal for
receiving useful feedback: posts typically receive
low-efort feedback such as likes and "Keep References
going!" comments. Furthermore, such broad
social sharing soon gets lost in the cacophony [1] L. Sardi, A. Idri, J. L. Fernández-Alemán,
of social media posts from hundreds of friends A systematic review of gamification in
eand followers. This proposal argues that peo- health, Journal of biomedical informatics
ple’s need to socialize goal-oriented progress 71 (
One way to motivate people is to convert a ence on Pervasive and Ubiquitous Comprivate goal into a broadcast goal with others puting and Proceedings of the 2015 ACM performing similar tasks; such communities International Symposium on Wearable of practice implicitly value actual real-world Computers, 2015, pp. 1063–1067. progress. The goodreads service nudges peo- [3] V. Wattanasoontorn, I. Boada, R. Gar