Prior studies of children with Autism Spectrum Disorder (ASD) with low verbal skills have demonstrated an increase of social behaviors in the presence of a robot as measured by imitation, touch, proximity, and gaze (Michaud et al. 2003, Robins et al. 2004, 2006). Studies have demonstrated a clear preference for plain, non-human features of robots over more human-like features by children with ASD. The ASD children's preference for and increased social behavior with robots arises due to their social simplicity, predictability, and responsiveness (Robins et al. 2003).
Objectives:
Our aim in this study was to examine: (1) ASD children’s reactions to various mechanical toys, including teleoperated robots and animatronic animals, to determine which features lead to increased social interaction between the mechanical toy and the child; and (2) ASD children’s interaction with typically developing (TD) peers in the presence of the various mechanical toys. We sought to extend our trials from a single child interacting with the toys to interactions between two children while engaging with the toys. Shared play was observed in the context of a variety of different toys in order to understand what characteristics led to increased social-engagement behaviors.
Methods:
Trials were conducted with ten pairs of children aged 2 to 8, each pair consisting of one child with ASD and one TD child. Seven pairs included ASD subjects with very low-verbal ability. Each pair of children was led into a child-centered playroom that contained one of four mechanical toys that represented a range in complexity of features. Two of the toys were teleoperated robots: the simplest, “Keepon,” can bounce up and down, turn and nod its head, and lean side-to-side on a stationary platform. "Cosmo” had similar complexity but could also move on a wheeled platform. These two robots were human-controlled in response to social-interactive child behaviors (e.g., nearness, gaze, smile, touch). The other two toys were animatronic animals, which behaved autonomously but whose sounds and movements were not always predictable (e.g., a dog that barked and moved). These toys also had much more complex features as they were meant to look lifelike.
Results:
Seven of the ten children with ASD demonstrated greater interest in the teleoperated robots than in the animatronic toys. The majority of these subjects also demonstrated greater imitative behaviors with their peers in the context of the robots. Seven of the ten children with ASD turned or walked away from mechanical animals. A number of different types of social interactions were observed with the two robots (smiles, vocalizations, gestures, gaze). As verbal abilities increased, engagement with the animatronic animals increased.
Conclusions:
Results indicate that the lower the child’s verbal abilities, the simpler their preferred robotic toy. In other words, higher-functioning children preferred the animatronic toys, whereas the more nonverbal children showed a definite preference for the robots. The next steps will be to determine which features lead to greater shared and social responses between children with ASD and their peers and expand on these findings regarding the response hierarchy.