Virtual humans—computer-generated faces and avatars—are rapidly becoming part of our online social worlds. This project advances understanding of emotional communication and explores how humans build rapport with virtual beings. The research examines fundamental questions about human-computer interaction: What social roles can virtual beings successfully fulfil? How do people's responses differ when viewing virtual versus human faces? What specific visual information in faces causes expressions to be perceived as genuinely emotional? Is this information the same for virtual and human faces? By answering these questions, the project aims to improve the design of virtual humans for applications in education, healthcare, and customer service, while also deepening our understanding of how humans process and respond to emotional signals. This research will contribute critical insights for developing more effective and trustworthy virtual agents that can enhance human-computer interaction across diverse settings.