Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: The rapid development of robotic technology has gone far beyond the traditional industrial robots which rely on limited human-robot interaction techniques to the recent development of animal-like entertainment robots, which engage in significant peer-to-peer (human-robot) interaction and manifest a wide range of social features including personality. Personality is an essential feature for creating socially interactive robots and study on this dimension will facilitate enhance human-robot interaction. Using AIBO, developed by SONY, this study examines the personality dimension in human-robot interaction.
In order to test the hypothesis that people would not only recognize robot’s personality but also socially respond to such personality accordingly, a balanced, 2 (AIBO personality: introvert vs. extrovert) x 2 (participant personality: introvert vs. extrovert) between-subjects experiment (N=48) was conducted. We believe that with accurate identification of AIBO’s personality types, the matched and mismatched personalities between robots and participants will make difference in participants’ perception of AIBO (i.e. AIBO’s intelligence, pet-likeness, physical attractiveness, social attractiveness), perception of the interaction (i.e. enjoyable, interesting, fun, entertaining, boring(reversed coded) and exciting), their self bonding to AIBO, their likeability of it and their social presence. Implications of the current study on human-robots interaction and personality literature will be discussed.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Social responses to lifelike characters can significantly alter human evaluations of technology. This study tested the differences between a picture of an on-screen character interacting in three different contexts (retail purchase, health advice, reading survey) versus a three-dimensional robot conducting the same interactions off screen. A laboratory experiment (n=72) was conducted in the US and replicated in Japan which tested differences in social evaluations, credibility, and memory. Results showed significant interactions between nationality and embodiment across a range of social and cognitive responses including perceived credibility, perceived surveillance, memory, and valence. These results are interpreted within technological and cultural contexts.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This study investigated the relationship between culture, intercultural communication apprehension (ICA), and attitudes toward humanoid robots using a questionnaire assessing undergraduates’ cultural levels and attitudes toward robots (ATR). Results demonstrated ICA and interdependent self-construal having positive relationships with negative ATR. There was a negative relationship between previous interactions with robots and negative ATR. This research helps explain how individual-level factors affect ATR and suggests that further research exploring effects of ATR on behavior is required.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Virtual social entities in computer interfaces have expanded the traditional social identity spectrum by including humanoid nonhuman characters such as robots. Robot characters provide a unique comparison for studying attitudes towards and preference of human races. An experiment with 105 White participants assessed the participants’ implicit and explicit racial prejudice towards the Blacks, their interest in robots, and their ordered preference among 15 White, Black and robot computer-generated characters for five different purposes. The explicit racial prejudice predicted more liking of White characters than Black ones and more preference of White to Black characters as one’s avatar, virtual friend, and virtual tutor. The implicit racial prejudice, measured with the Implicit Association Test (IAT), provided further predictive utility for choosing avatar and virtual friend among participants who took less time in character selection. Interest in robots and racial prejudice predicted the preference of robots to White or Black characters for some tasks. Among the 65 participants (61.9%) who reported no interest in robots, racial prejudice negatively predicted the preference of Black to robot characters in liking and as virtual friend and tutor, showing a pattern that White individuals with high explicit racial prejudice progressively preferred robot characters to Black ones. Although the results suggest alarming strength of high racial prejudice, drawing implications about real humans and robots is cautioned against unless future research replicates the effects with images of real humans and real robots.

Publication Type: Session Paper
Abstract: According to Computers Are Social Actors (CASA) theory, people will respond to robots using the same rules and heuristics that they use when interacting with other people. Given this paradigm, there are five different interaction stances that a robot can be designed to present: person, foreigner, social actor, medium, and machine. For each model, I will present the linguistic, para-linguistic, and behavioral manifestations associated with each model. I will then discuss the experimental evidence that demonstrates the advantages and disadvantages of each model.