Effects of interactivity and expressiveness on perceived social presence, memory and persuasion in interactive health communications

As interest in human-agent social interaction has increased, attention has increasingly been devoted to the embodiment of intelligent and emotional agents to enhance affective communication and in turn to maximize social response from users. Emotional displays and interactive responses are key in evoking social response from users to the agent appearing to interact with them. This experiment primarily focused on how media variables such as interactivity and agent expressiveness affected user's perception of the social presence of computer agent. Secondarily, we were interested in exploring how both interactivity, expessiveness and social presence affect the user's memory, attitude, and behavioral intention to a health communication message.; This study employed animated intelligent characters to guide young adult users (N = 124) in a drunk-driving prevention program. Users who encountered an emotionally expressive agent perceived higher levels of social presence of the agent. Users felt higher levels of mutual awareness, understanding, and more empathy and felt a strong sense of expressive animated agent's behavioral dependency on them. Interactive media users also felt higher levels of mutual awareness between them and the animated characters as well as others in the environment. They also felt higher levels of mutual understanding and animated character's behavioral dependency on them. Participants lacking two-way interaction with the animated character seemed also to have felt empathy. This phenomenon mirrors para-social relationships in traditional media, as when television audiences react to TV personae.; Interactive media fared better in enhancing the learning of the health information, while expressiveness did not figure significantly in learning. Users who used interactive media perceived animated agents to be more credible and perceived the expressive agent as more sociable and likeable. Furthermore, user's perceived presence of animated agents mediated away agent credibility directly linking to interactivity, while agent sociability/likeability directly associated with expressiveness stays un-swayed by user's perceived presence of the animated agents.; The findings suggest that interactivity was, to a higher degree, associated with cognitive factors such as learning, agent credibility, and mutual understanding while expressiveness has more impact on affective components, such as empathy, agent likeability and sociability. Agent sociability supported by expressiveness may not be associated with learning and that agent credibility supported by interactivity may be more likely to be associated with learning as interactivity was a strong predictor of learning. Future research may choose a topic that has not been as widely promoted as drunk driving prevention to have a better chance of getting attitude change, as the message receivers would be more susceptive to change if they did not have a set of attitude or beliefs with which to defend the attitude/beliefs advocated in the messages.