Making Predictions Tangible: Using Data Physicalization to Explore Expectations Around Health Predictions

As predictive technologies like Digital Twin become increasingly applied in healthcare, much focus has been given to improving accuracy, explainability, and trust in health predictions. However, little is known about how people expect to engage with predictions in meaningful ways. In this paper, we explore how people imagine interacting with their health predictions through the physicalization they constructed. We conducted three workshops with 18 design students, using fictional weekly allergy risk predictions and multisensory materials to elicit the embodied, contextual, and creative representations of predicted health risks. Our analysis of the physicalizations revealed that participants envisioned predictions not merely as data points but as embodied and contextual experiences. We identified five sensory mapping strategies and six design themes that illustrate how people engage with predictions. Drawing from these findings, we reflect on design implications to support agency, adaptability, and personally meaningful interactions. We contribute an empirical understanding of users' interpretations and engagements with health predictions, offering inspiration for expanding the interaction design space of future predictive health technologies.