Using a Socially Assistive Robot to Improve Attention in Children with Autism During Academic Instruction

The current study examined how a socially assistive robot (SAR) impacts attending for children with autism spectrum disorder (ASD). We used a multiple-baseline single-case design with an extended intervention timeline to evaluate how robotic-delivered instruction influenced the duration and frequency of eye contact in four children with severe autism symptoms. The current study presents the first evaluation of a commercially available SAR paired with an educational curriculum, in a real-world clinical setting. Participants included four children with severe symptoms of ASD. During baseline, participants completed tablet-based learning tasks with a human instructor; during intervention, the same tasks were led by the robot. Data analysis included both descriptive and Tau-U methods, alongside qualitative social validity assessments from staff and children. Results indicated that children made more frequent and sustained eye contact with the robot than with the human instructor. All four children preferred the robot-led sessions, and all six instructors viewed the intervention as positive and meaningful for the students. Strengths of this study include the use of a rigorously controlled design and focus on a minimally studied population. However, the small sample size and narrow age range limit generalizability. The findings suggest SARs may provide an effective, engaging format for children with significant autism challenges, warranting further study.