Sketch it for the Robot! How Child-Like Robots' Joint Attention Affects Humans' Drawing Strategies

摘要

The work proposes investigating drawing activities in interactive contexts to shed light on the links between socio-cognitive and visual representation mechanisms. In our study, 53 adult participants were instructed to draw some object categories (e.g., a duck, an ambulance, …) for a child-like robot, which first was just shown in a picture (individual condition) and then was co-present and engaged in joint attention behaviors with them (robot condition). The data collection was carried out in two sites, in Italy (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{N}=26$</tex>) and Slovakia (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{N}=27$</tex>), with two different robots, iCub and Nico. Participants significantly changed their drawing strategy in the presence of the robots by enlarging their sketches while speeding up their drawing. The phenomenon was more evident the more the individuals perceived the robot as closer to them, according to the IOS scale. The results were highly consistent between the two sites and showed that participants put more effort into drawing understandably when a robot actively attends to their behavior. Higher clarity is obtained with increased figure size rather than simplifying the drawing. Counter-intuitively, participants did not slow down their tracing to be more comprehensible. Instead, they became faster in front of the robot, potentially induced by the pressure of being observed by it. These findings are discussed in the framework of the motionese literature.