Nonverbal behaviors toward an audience and a screen for a presentation by a humanoid robot

Hiroko Kamide, Koji Kawabe, Satoshi Shigemi, Tatuo Arai


Objective: We propose a model which predicts patterns of nonverbal behaviors for a successful presentation by a humanoidrobot, especially focusing on two types of the behaviors. One is the nonverbal behaviors of eye-contact with its face and openpostures with arms to keep an attention of an audience who are listening to the presentation. The other is pointing with its handand approaching to a screen with a step to emphasis important points on the screen which is used in the presentation.

Methods: We tested the hypothesis that both types of nonverbal behaviors are effective for ensuring better understanding of thepresentation. We prepared four conditions which show high or low tendency in each type of the nonverbal behaviors. A totalof 139 participants observed a presentation by a humanoid robot in a between-subject design and then completed a surprisedtest. They also evaluated general impressions of the robot based on a psychological scale which is developed for an evaluationof humanoids.

Results: We found that both approaches are related to higher scores regarding the audience’s correct understanding of thepresentation, with higher psychological impressions given in relation to utility of the robot and the clearness of the voice of therobot. Additionally, we found that the behaviors toward a screen is more effective than the behaviors toward an audience in thiscase of the presentation by a humanoid robot.

Conclusions: We concluded that both types of the nonverbal behaviors are important for the audience’s correct understandingand also the behaviors which emphasize the key points in the screen is crucial rather than the behaviors which keep the attentionsof the audience when a humanoid robot gives a presentation. Finally, we discuss the universality of the proposed model for usewith other humanoid robots.

Full Text:




  • There are currently no refbacks.

Artificial Intelligence Research

ISSN 1927-6974 (Print)   ISSN 1927-6982 (Online)

Copyright © Sciedu Press 
To make sure that you can receive messages from us, please add the '' domain to your e-mail 'safe list'. If you do not receive e-mail in your 'inbox', check your 'bulk mail' or 'junk mail' folders.