Social Approach in Autism: Modeling Objective Measurements of Child Movement in a Naturalistic Environment

Chitra Banarjee,Yudong Tao,Regina M. Fasano,Laura Vitale,Chaoming Song,Mei-Ling Shyu,Lynn K. Perry,Daniel S. Messinger

DOI: https://doi.org/10.21203/rs.3.rs-90297/v1

2020-01-01

Abstract:Background. Atypicalities in social approach are thought to be characteristic of children with autism spectrum disorder (ASD), but few studies have quantified the social movement of children with ASD using objective measures. The purpose of this paper is to introduce a new method—computational modeling of radio frequency identification (RFID) child tracking—for studying children with ASD in a naturalistic setting. We present the use of RFID measurements to investigate the velocity and social approach of children with ASD and typically developing (TD) children interacting together in preschool inclusion classrooms during repeated multi-hour observations. Methods. Observations of 14 preschoolers with ASD and 16 TD preschoolers in two inclusion classrooms on a total of 10 days yielded approximately 10 hours of data per child. Objective measurements of position and orientation were collected using four corner-mounted Ubisense ultra-wide sensors, which tracked a right and left tag worn by each child (in a vest) and teacher in the classroom. We calculate angular velocity, velocity, and social approach, and compare ASD and TD children on these parameters using multilevel statistical models. Results. In this initial exploration of the ASD phenotype in situ, children with ASD did not differ from TD children in angular velocity or velocity of movement in the classroom. Rather, pairs of TD children moved toward and away from each other at higher velocities than both pairs of children with ASD and pairs in which one child had ASD and the other child was TD. Children with ASD, however, moved toward and away from teachers at higher velocities than TD children. Limitations. Illustrative data from repeated observations of 30 children in two classrooms are reported. Results are preliminary. Conclusions. Multi-hour, objective measurements in a preschool inclusion classroom indicated that children with ASD did not move through space or turn at higher velocities than other children. Instead, ASD differences were evident in social approach. Children with ASD were slower in approaching peers but quicker in approaching teachers than were TD children. The results suggest the potential of modeling RFID measurements to produce a quantitative understanding of the ASD phenotype in naturalistic social contexts.