Abstract
Health professionals can potentially take advantage of self-tracking data of their clients. To make effective use of those data, professionals need meaningful modeling beyond data summaries. Yet, transparency of such models is an important prerequisite for trust. Paradoxically, prior research indicates that high transparency levels potentially erode trust, and we hypothesize that this effect is moderated by users’ expertise levels. We conducted an online study with participants with various expertise levels in both health coaching and modeling methods. Domain experts showed higher congruency with the system when presented with high transparency over medium transparency, whereas novices showed higher levels of trust in medium transparency over high transparency. Modeling expertise did not influence trust across transparency levels. Our results emphasize the importance of tailoring the transparency level to the needs and beliefs of health professionals, to facilitate effective collaboration with health coaching support systems based on self-tracking data of their clients.
| Translated title of the contribution | Tailoring Transparency to Expertise: Health Professionals’ Need for Transparency in Representing Self-Tracking Data of Clients |
|---|---|
| Original language | English |
| Number of pages | 12 |
| Publication status | Published - 17 Mar 2020 |
| Event | 4th HUMANIZE workshop on Transparency and Explainability in Adaptive Systems through User Modeling Grounded in Psychological Theory (Workshop at IUI'20) - Cagliari, Italy Duration: 17 Mar 2020 → 17 Mar 2020 http://www.humanize-workshop.org/ |
Workshop
| Workshop | 4th HUMANIZE workshop on Transparency and Explainability in Adaptive Systems through User Modeling Grounded in Psychological Theory (Workshop at IUI'20) |
|---|---|
| Abbreviated title | HUMANIZE '20 |
| Country/Territory | Italy |
| City | Cagliari |
| Period | 17/03/20 → 17/03/20 |
| Internet address |