Differences in Social Expectations About Robot Signals and Human Signals

Area of Science:

• Human-computer interaction
• Cognitive psychology
• Social robotics

Background:

• Decision-making often occurs within social interactions, influenced by communicative signals.
• Human interaction with artificial agents, like robots, is increasing.
• Understanding how non-human agents' social signals affect human choices is crucial.

Purpose of the Study:

• To investigate the influence of robot non-verbal communicative behavior on human decision-making.
• To examine how robot cues and feedback affect choices in a social context.

Main Methods:

• A two-alternative-choice "Shell Game" task was adapted.
• Participants made decisions with a robot avatar providing pre-decision cues and post-decision feedback.
• Robot cues (pointing) and feedback (thumbs up/none) were manipulated.

Main Results:

• Participants were slower when robot cues were invalid and feedback was positive.
• This performance decrease was attributed to incongruent signals violating expectations.
• Incongruence between pre- and post-decision signals significantly impacted task performance.

Conclusions:

• Robot social signals, particularly when incongruent, significantly influence human decision-making and task performance.
• Understanding user expectations of social robots is vital for effective human-robot interaction.
• This research highlights the importance of consistent non-verbal communication in human-robot collaborations.