Understanding how to Design Health Data Visualizations for Chilean Older Adults on Mobile Devices

Abstract

Mobile devices, including activity trackers and smartwatches, can help older adults monitor health parameters passively and unobtrusively. Most user interactions with small devices consist of brief glances at the time or notifcations. Consuming information from small displays poses challenges, which have been seldom studied from the perspective of older users. In this paper, we worked with older adults towards creating health data visualizations for them for small devices. We conducted a mixed-methods study with 30 older adults, in which we (1) conducted group discussions to understand participants' opinions, (2) measured times taken to interpret health data visualizations with and without progress information, (3) measured how much information they could manage to see during brief glances. When data was visualized without progress indicators, participants took less time to understand the data and made fewer errors. Participants preferred health data visualizations that featured peaceful, and positive pictorial representations. We present design opportunities for older adults' data visualizations in small devices.

GerontoVis: Data Visualization at the Confluence of Aging

“…Several empirical studies (e.g., [JHG*12,WM12,MSC*19]) advocate for the adoption of simple and high data‐ink ratio visualizations as a means to alleviate cognitive load and demand on older adults. In a participatory design project conducted by Cajamarca et al [CHD*23], older adults expressed their preference for health data apps on smartwatches to have “familiar,” “clear,” and “clean” designs. In two separate studies, Le et al [LRY*15, LCC*18] also found that older adults tended to focus on aggregate measures and data overviews unless trying to explain an observed pattern or change, indicating that older adults' information‐seeking behavior may conform to Schneiderman's visualization mantra “overview first, details on demand” [Shn03].…”

“…In addition to investigating various use cases of data visualization in health care, these also offer valuable insights that can inform the design of visualizations for older adults. For instance, Ahmed et al [ATRG*19] observed that older adults consistently removed a sad‐looking face emoji from a heart health dashboard due to its negative connotation and concluded that such designs may discourage continued use, while Cajamarca et al [CHD*23] observed that older adults preferred health visualizations that instilled tranquil feelings. While studying how older adults would use data visualizations for self‐assessment of their abilities to use a computer mouse, Jones et al [JHG*12] observed that participants noticeably found charts depicting hypothetical unwanted health scenarios to be unfavorable and avoided interacting with them.…”

Characterizing Glanceable Visualizations: From Perception to Behavior Change

“…While prior work on glanceable visualization and aging is remarkably sparse [14], the existing literature provides a starting point that informs and motivates this study. In this section, we discuss relevant current work on glanceable visualization (Section 3.1), visualizations for older adults (Section 3.2), as well as smartwatches and older adults (Section 3.3).…”

“…Fan et al [23] identified challenges for older adults in understanding interactive COVID-19 data visualizations, such as insufficient information clarity. Additionally, Cajamarca et al [14] observed that older Chilean adults interpreted health data on smartwatches more accurately without progress indicators, which otherwise distracted them from essential information. They also found that older adults with higher technology proficiency had better accuracy and speed in data interpretation.…”

“…The existing but limited body of work on visualization for older adults highlights critical aspects of design and graphical perception specific to this demographic. However, only one previous work was performed using smartwatches or glanceable visualizations [14]. This gap in the current literature underscores the pressing need for further exploration to determine whether the established understanding of graphical perception remains applicable in the face of significant shifts in screen size (small) and design objectives (rapid information retrieval).…”