What drives older adults’ acceptance of virtual humans? A conjoint and latent class analysis on virtual exercise coach attributes for a community-based exercise program

As an example of health-enhancing physical activities (HEPA), structured physical exercise is valuable in promoting healthy lifestyles among community-dwelling older adults. Technology-driven virtual coaches have the potential to enhance and improve exercise programs, but the preferences of the aging population were not previously explored. This study examined and analyzed the attributes and levels related to the acceptance of virtual coaches among the aging cohort via experience-based conjoint and latent class analysis. Purposively selected respondents (<i>n</i> = 232) from two senior centers in the Philippines completed a conjoint activity followed by a computer-based survey focusing on attributes related to platform, appearance, gender, language, and music. Results revealed the subjects' inclination to a humanlike, feminine, local language-speaking virtual coach projected through a mixed reality platform with a contemporary music background. Additionally, latent class analysis outcomes produced three subgroups based on the pattern of preferences among the technology users. Study outcomes reinforce the importance of human-centered design and multidisciplinary approaches to health technology development.