Poster #116 - Quantification of homophily in naturalistic, early childhood environments

• In Event: Poster Session 17

Abstract

There is evidence that children preferentially associate with similar children, a phenomenon known as homophily. This phenomenon has been observed in the classroom context in the domains of gender, disability, and academic performance. However, this evidence primarily comes from subjective ratings from teachers. In this study, we aim to use objective measures of position and orientation in the naturalistic setting of preschool inclusion classrooms to quantify homophily in children with developmental disabilities (DD) and typical development (TD).
Participants included 77 children (33 female, M=48.26 months, SD=7.47), including 47 children with DD, and 30 TD children from 8 preschool inclusion classrooms. Children were Hispanic (66) and non-Hispanic (11), as well as White (74), Black (2), and multiracial (1). Each child was observed on 3.52 mean occasions (SD=1.32) for a duration of mean 138.2 minutes (SD=63.6). Objective measurements of position and orientation were collected using ultra-wide radio frequency identification (Ubisense) sensors, which tracked a right and left tag worn by each child in a specially designed vest. Social contact was defined as fulfilling the criteria of distance between 0.2 and 2 meters and relative orientation within 45° and computed as a proportion of available time with each peer. Social approach was computed as the distance each child moved toward the initial position of each peer, weighted by distance, orientation, and angle of movement. Linear mixed effects models were utilized to analyze social approach and time in social contact. Fixed effects included a term for grouping of one child in the group (DD or TD) and homophily status of the pair (concordant or discordant). Observations (Level-1) were nested in children (Level-2) and children were nested in classroom (Level-3).
Children in concordant dyads (TD-TD, DD-DD) move toward and away from one another at higher velocities (ps<0.001, Figure 1) than discordant dyads (TD-DD, DD-TD). Similarly, children in concordant dyads (TD-TD, DD-DD) exhibited greater time (p<0.001, Figure 2) in social contact than discordant dyads (DD-TD). Additionally, DD-DD dyads were comparable to TD-TD dyads in their social approach velocities but spent less time in social contact than TD-TD dyads. In addition, children with DD were approached by other children in the classroom at lower velocities than TD children, but did not differ from TD children in social contact, a social preference effect evident in the movement, but not colocation, of children in the classroom.
Patterns of homophilic social interaction in approach velocities and time in social contact indicate that children with TD and DD face barriers interacting with one another in inclusion classrooms. Greater velocity of approach and increased time in social contact characterizes a preference for particular peers—that children’s velocity of approach and time in social contact indexes their motivation to be close to specific other children. Children’s preference for similar peers in their movement and proximity to one another appears to be a pervasive feature of their naturalistic interactions. The results suggest the potential of a new generation of objective measures to quantify strengths and deficits in social interaction in naturalistic contexts.

Authors

• Chitra Banarjee, University of Central Florida
  Presenting Author

• Yudong Tao, University of Miami

• Regina Fasano, University of Miami

• Chaoming Song, University of Miami

• Laura Vitale, University of Miami

• Jue Wang, University of Science and Technology of China

• Mei-Ling Shyu, University of Missouri - Kansas City

• Lynn K Perry, University of Miami

• Daniel Messinger, University of Miami