MOSAIC: Providing Meaning through Multimodal Health Data Representations for Older Adults

Health self-monitoring systems increasingly incorporate multimodal feedback, which often distorts or fragments information when rendered across visual, auditory, or combined channels. In health contexts, such inconsistencies attenuate confidence and increase anxiety, particularly for older adults managing chronic conditions. We introduce semantic mirroring, a design approach that frames the meaning of health information as a modality-independent message. This meaning remains consistently interpretable across interaction channels rather than being tied to specific sensory encodings. We demonstrate this approach through the design of MOSAIC, a research prototype for blood pressure self-monitoring that realises semantic mirroring through synchronised visual and auditory representations. Through a formative study with older adults (n = 6 recruited; n = 4 analysed), we found that surface-level cues were easily perceived across modalities, but more complex interpretations, such as action-oriented cues, cross-modal synchrony, and emotional responses, were frequently lost. These breakdowns were influenced by participants' existing health beliefs and expectations.