Laboratory of Informatics of Grenoble Engineering Human-Computer Interaction Research Group

Engineering Human-Computer Interaction
Research Group

Non-verbal Signals in HRI: Interference in Human Perception

In Proceedings of The International Conference on Social Robotics 2015 (ICSR'15). pages 275-284. 2015.

Wafa Johal, Gaëlle Calvary, Sylvie Pesty

Abstract

Non-verbal cues of communication can influence the human understanding of verbal signals in human-human communication. We present two illustrative experimental studies showing how non-verbal cues can both interfere and facilitate communication when passing a message to a user in HRI. In the first study, participants found that the cues enabling them to discriminate between two conditions : permissive or authoritative robots were mainly verbal. The verbal message was however unchanged between these two conditions and in this case, non-verbal cues of communication (gestures, posture, voice tone and gaze) substituted the neutral verbal message. The second study highlights the fact that verbal and non-verbal communication can facilitate the understanding of messages when combined appropriately. This study is based on a Stroop task of identifying the colour of the LEDs of a robot while the robot says words that are either facilitating, neutral or disturbing for the participant. These two studies put into perspective the importa