Beyond the Response: How Timing and Context Shape Empathy, Responsiveness, and Social Presence in AI Conversations

This study investigates how conversational artificial intelligence (CAI) agents’ response delay and context awareness influence users’ perceptions of social presence, empathy, and responsiveness, and engagement, in the context of stress management and wellness for college students. Grounded in social response theory and social presence theory, this research focuses on the conceptual mechanisms through which these design features shape human-like interactions in CAI. Response delay refers to the intentional pause between a user’s input and the agent’s reply. Social response theory suggests that pauses reflect “thoughtfulness,” which aligns with human expectations of social interaction and thus foster trust and engagement. A short delay (e.g., 2 seconds, accompanied by a visual indicator) may signal CAI’s humanlike cognitive effort and deliberation, enhancing users’ perceptions of social presence or sense of being with an intelligent, socially engaging entity. Context awareness refers to CAI’s ability to recall and integrate conversational history. A context-aware CAI references prior user interactions, ensuring coherence, personalization, and emotional resonance in its interaction with users. Social presence theory suggests that the ability to build on prior interactions enhances social presence. The increased social presence through CAI’s thoughtful delay and personalized responses based on context awareness may create a relationally engaging interaction, fostering user engagement with the CAI. This study will employ a 2 (response delay: yes vs. no) × 2 (context awareness: aware vs. unaware) between-subjects experiment. Four versions of CAI, as a virtual wellness counselor for college students, will be designed to manipulate the four experimental conditions, and their impacts on users’ social presence, empathy and responsiveness perceptions, as well as engagement will be examined. Findings will provide theoretical and practical insights into optimizing CAI design for human-like interactions in wellness contexts.