Abstract

The present dissertation investigated the effect of individual variables on the perception of emotional nonverbal behaviors of an empathic virtual agent. Prior studies have reported the importance of implementing empathy in virtual agents regarding maximizing the quality of human–agent interactions. As studies in human–human interactions demonstrate, nonverbal behaviors play a significant role in the transfer of meanings and intentions between the interlocutors. Therefore, the aim of this dissertation was to examine how individual variables influence the perception of the nonverbal behaviors of an agent. The nonverbal behaviors were specifically created to express the empathizing intention of the virtual agent. In this regard, three empirical studies have been conducted to investigate the effect of users’ age (Study 1), users’ cultural background (Study 2), and users’ cultural stereotypes toward the agent (Study 3) on the perception of an empathic agent. The results of Study 1 provide important insights into the understanding of how users of varying ages perceive an agent’s emotional nonverbal responses. Older users observed to be more sensible to nonverbal behaviors of the agent since they perceived the agent as more empathic compared with younger adults. The higher perception of the older users was based on the agent’s displayed nonverbal behaviors, indicating the importance of implementing emotional nonverbal behaviors specifically for this target group. Moreover, there were differences between older and younger adults with respect to perceiving specific nonverbal behaviors as empathic. The culture of the users in Study 2 was also found to have an influence on users’ perception of an agent. Comparing Iranian and German users (as the examples of relatively collectivistic and individualistic cultures) showed differences regarding users’ perception of the agent’s nonverbal behaviors. While Iranian participants rated a wide array of nonverbal behaviors as empathic, their German peers perceived one specific nonverbal behavior only as the most empathic one. This is where the factor of culture becomes a matter of concern. In order for a virtual agent to look more emphatic, users’ culture should be considered while implementing nonverbal behaviors. And finally, in Study 3, it was observed that users do not show stereotypical attitudes toward the agent when they are aware that the agent is designed and produced in a different country. Users who were primed about the country of the agent did not show different ratings of empathic behaviors of the agent in the presence of emotional nonverbal behaviors. The findings of this study make valuable contributions to the research in the domain of bias toward the agent. It can be stated that as long as there are no cues in the appearance of the agent representing a different culture, users do not get biased toward its emotional responses. Taken together, this dissertation provides insight into understanding that user variables indeed matter and they can influence accurately perceiving the emotional nonverbal behaviors of virtual agents. Since the purpose of designing an empathic agent is to enhance users’ experiences and satisfaction, the factors that have the potential to impact their perception of an empathic agent should be considered when designing a virtual agent.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call