Victoria Groom

The influence of racial embodiment on racial bias in immersive virtual environments

Increasingly, people interact with others via digital representations, or avatars, that feature indicators of race. Nonetheless, little is known about the effects of avatar race on attitudes and behaviors. We conducted a study to determine how peoples implicit racial bias is affected by the race of their avatar in an immersive virtual environment (IVE). Our results indicate that the effects of avatar race extend beyond digital spaces. People embodied by Black avatars in an IVE demonstrated greater implicit racial bias outside the IVE than people embodied by White avatars. These findings have important implications for strategies to reduce racial prejudice and provide new insights into the flexibility of racial identity and racial attitudes afforded by virtual technologies.

I am my robot: the impact of robot-building and robot form on operators

As robots become more pervasive, operators will develop richer relationships with them. In a 2 (robot form: humanoid vs. car) × 2 (assembler: self vs. other) between-participants experiment (N=56), participants assembled either a humanoid or car robot. Participants then used, in the context of a game, either the robot they built or a different robot. Participants showed greater extension of their self-concept into the car robot and preferred the personality of the car robot over the humanoid robot. People showed greater self extension into a robot and preferred the personality of the robot they assembled over a robot they believed to be assembled by another. Implications for the theory and design of robots and human-robot interaction are discussed.

I'm sorry, Dave: i'm afraid i won't do that: social aspects of human-agent conflict

As computational agents become more sophisticated, it will frequently be necessary for the agents to disagree with users. In these cases, it might be useful for the agent to use politeness strategies that defuse the persons frustrations and preserve the human-computer relationship. One such strategy is distancing, which we implemented by spatially distancing an agents voice from its body. In a 2 (agent disagreement: none vs. some) x 2 (agent voice location: on robotic body vs. in control box) between-participants experiment, we studied the effects of agent disagreement and agent voice location in a collaborative human-agent desert survival task (N=40). People changed their answers more often when agents disagreed with them and felt more similar to agents that always agreed with them, even when substantive content was identical. Strikingly, people felt more positively toward the disagreeing agent whose voice came from a separate control box rather than from its body; for agreement, the body-attached voice was preferred.

Critic, compatriot, or chump?: responses to robot blame attribution

As their abilities improve, robots will be placed in roles of greater responsibility and specialization. In these contexts, robots may attribute blame to humans in order to identify problems and help humans make sense of complex information. In a between-participants experiment with a single factor (blame target) and three levels (human blame vs. team blame vs. self blame) participants interacted with a robot in a learning context, teaching it their personal preferences. The robot performed poorly, then attributed blame to either the human, the team, or itself. Participants demonstrated a powerful and consistent negative response to the human-blaming robot. Participants preferred the self-blaming robot over both the human and team blame robots. Implications for theory and design are discussed.

A toolkit for exploring the role of voice in human-robot interaction

This paper describes an open source speech translator toolkit created as part of the “Survivor Buddy” project which allows written or spoken word from multiple independent controllers to be translated into either a single synthetic voice, synthetic voices for each controller, or unchanged natural voice of each controller. The human controllers can work over the internet or be physically co-located with the Survivor Buddy. The toolkit is expected to be of use for exploring voice in general human-robot interaction.