Active Peer Pressure in Human-Robot Interaction
Computer Science and Engineering
StatisticsView Usage Statistics
This work investigates how conformity in human-robot groups can be manipulated by the robots’ ability to actively apply coordinated peer pressure via voicing their agreement or disagreement with the human’s selections in a visual task. Through a combination of kinesics and vocalics, we created an environment where the human subject became aware of the fact that they were not just actively being observed (by employing synchronized robot body and camera gaze motion), but were also under the active judgment and criticism of their robot peers. Our experiments consisted of two studies, a 2×2 and a 2×1 that considered a combination of possible conditions such as the sequencing of expressing opinions in the group and the number of robot peers, as well as the difference between unambiguous, ambiguous, and even duplicate selections in the used visual indicators. Using statistical analysis we show that the application of the new idea of “active peer pressure” manages to achieve increased conformity across many of the considered conditions.