How to Build a Trustworthy Robot

AI  Communication and Technology 

Robot narratives have been with us for a long time, 155 years to be exact, when Steam Man of The Steam Man of the Prairies first made an appearance in a dime-store novel. Since then, humans have fantasized, advised and prophesized about robots existing among us.

But would people actually trust these fictional robots if they were real? That’s a question that got University of Florida College of Journalism and Communications researchers thinking about how humans respond to social cues generated by “social robots”—robots that possess the capacity to interact with people.

Social cues are non-verbal signals that humans use to communicate with each other, including facial expressions, body language, eye gaze and tone of voice. Humans depend on social cues to regulate interactions with others and to build and strengthen relationships.

Social cues can also affect our interactions with robots. For example, we may be more likely to trust a robot that uses friendly facial expressions and makes eye contact with us. We may also be more likely to follow instructions from a robot that we perceive to be competent.

Media Production, Management, and Technology Assistant Professor in Emerging Media Kun Xu, doctoral student Mo Chen and alumna Leping You, M.A.M.C 2014, Ph.D. 2020, began with the premise that although there has been some research that distinguishes the different effects between voice and text, gestures and movements, and human voices and synthetic voices, there has not been a systematic organization of existing research about different robot responses to social cues from humans.

The researchers conducted two meta-analyses of already existing studies to synthesize the findings. All of the studies suggested that “some cues may be more powerful than others in evoking users’ natural and instinctive responses to social actors.”

The meta-analyses found that social cues have positive but small-sized effects on users’ social responses, meaning that humanlike cues have stronger effects than machinelike cues.  More specifically, facial and kinetic cues have medium-to-large-sized effects on users’ social presence and trust. This means that robots with facial and kinetic cues are more likely to be perceived as real people and trustworthy, which can guide future social robot designers as they forge ahead in this brave, new world of human–robot interaction.

The original article, “The Hitchhiker’s Guide to a Credible and Socially Present Robot: Two Meta-Analyses of the Power of Social Cues in Human–Robot Interaction,” was published online in the International Journal of Social Robotics on Jan. 18, 2023.

Authors: Kun Xu, Mo Chen and Leping You.

This summary was written by Gigi Marino.

Posted: October 2, 2023
Tagged as: Kun Xu, Robots, social cues