Recent studies showed that in Human-Robot Interaction (HRI), humans perceive less agency over the negative outcomes of their actions, raising the Diffusion of responsibility (DOR) phenomenon. In the present study, we examined the effect of anthropomorphism on the reduction of agency when interacting with robots. To this end, young adults played a risk-taking task, in which they were asked to stop an inflating balloon before it reaches a pin and burst. However, every time they acted to stop the inflation of the balloon they were losing points from a starting score. Participants play the task alone, or together with a co-agent. Within-task, we manipulated the co-agent: a human, a robotic arm, and a humanoid robot. Results showed lower agency ratings reported when participants shared the task with the co-agent compared to when they performed the task alone. Interestingly, such a reduction was comparable across the three co-agents. This suggests that DOR in HRI occurs similarly to when interacting with a human being regardless of the level of anthropomorphism of the robotic partner.
Tuvo, E., Ricciardelli, P., Ciardo, F. (2023). The Effect of Anthropomorphism on Diffusion or Responsibility in HRI. In F. Cavallo, J. Cabibihan, L. Fiorini, A. Sorrentino, H. He, X. Liu, et al. (a cura di), Social Robotics. 14th International Conference, ICSR 2022, Florence, Italy, December 13–16, 2022, Proceedings, Part II (pp. 488-497). Springer, Cham [10.1007/978-3-031-24670-8_43].
The Effect of Anthropomorphism on Diffusion or Responsibility in HRI
Tuvo, E;Ricciardelli, P;Ciardo, F
2023
Abstract
Recent studies showed that in Human-Robot Interaction (HRI), humans perceive less agency over the negative outcomes of their actions, raising the Diffusion of responsibility (DOR) phenomenon. In the present study, we examined the effect of anthropomorphism on the reduction of agency when interacting with robots. To this end, young adults played a risk-taking task, in which they were asked to stop an inflating balloon before it reaches a pin and burst. However, every time they acted to stop the inflation of the balloon they were losing points from a starting score. Participants play the task alone, or together with a co-agent. Within-task, we manipulated the co-agent: a human, a robotic arm, and a humanoid robot. Results showed lower agency ratings reported when participants shared the task with the co-agent compared to when they performed the task alone. Interestingly, such a reduction was comparable across the three co-agents. This suggests that DOR in HRI occurs similarly to when interacting with a human being regardless of the level of anthropomorphism of the robotic partner.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
