Lifting the veil: Using sensing technologies to examine variance in young children’s classroom network experiences

• In Event: Using and studying technology in research involving children and families

Abstract

Preschoolers spend a large portion of their classroom time interacting with peers (Chaparro-Moreno et al., 2019). These interactions impact young children’s development across academic and social domains (Neal et al., 2022). While findings consistently show the influence exerted by children’s interactions with peers in their classroom networks (Chen et al., 2020), gaps exist in our understanding of how peer effects may vary across classroom contexts (Neal et al., 2022). Methods used to study children’s social networks (e.g., Daniel et al., 2016; DeLay et al., 2016) often employ a one a one-size-fits-all approach, using a meta-network and structural-zeroes. While this is an efficient means for descriptively understanding network trends, this method relies on the assumption that peer effects operate the same across all classroom social networks. This may be unrealistic considering the variance found across classroom contexts at levels of the child, teacher, and classroom ecology (Varmuza et al., 2021). For instance, some classroom play locations may facilitate less opportunities for language exchange with peers (Irvin et al., 2021). The aim of the present study is to understand if peer language exchange may vary across different early education classrooms using state-of-the-art technologies.
Studying peer effects and classroom social networks typically relies on expert observation (Altman et al., 2020). However, in-person observations are prohibitively expensive to collect (Pianta & Hamre, 2009) and susceptible to observer bias (Hunter, 2020). Investigating social network phenomena and child development via these methods can lead to divergent scientific findings (Bergelson et al., 2019). Sensing technologies may be a promising alternative. These tools provide continuous, automated data on the experiences of all children in the classroom. The present study uses a newly developed sensing system called Interaction Detection in Early Education Settings (IDEAS) to understand the differences in children’s exposure to peer talk within and between classrooms.
The sample includes 14 classrooms and 178 children. Demographic data were collected from teachers (e.g., education, experience) and caregivers (e.g., mothers’ education, income). IDEAS recorded children’s peer interactions at eight timepoints. Children wore Bluetooth beacons and voice recorders during these sessions. The Bluetooth component is used to examine affiliation through measures of mutual orientation and proximity. Voice recorders are used to provide various measurements related to ingoing and outgoing talk and conversational turns. Data collection is complete and currently being processed via the IDEAS automated pipeline (Figure 1). Processing and subsequent analyses will be complete by January, 2022.
To examine our aim of understanding how peer language exchange may vary across different early education classrooms, first we conduct hypothesis testing using T-tests to examine if any significant differences in average rates of talk exist (Cappella, 2013). We will then compare network descriptives (e.g., degree; density; reciprocity; Burt’s constraint) across classrooms. If significant variance between classrooms is found, we will examine what child- and classroom-level characteristics effect the rates of interaction children experience (Table 1). This work seeks to fundamentally increase our understanding of how we can leverage state-of-the-art technology to understand young children’s classroom networks and how they shape their development.

Group Authors

Logan Pelfrey (The Ohio State University)
Tiffany Foster, PhD (The Ohio State University)
Laura Justice, PhD (The Ohio State University)
Hugo Gonzalez-Villasanti, PhD (University of Michigan)
Dwight Irvin, PhD (Kansas University)
Daniel Messinger, PhD (University of Miami)
Lynn Perry, PhD (University of Miami)

Authors

• G. Logan Pelfrey, The Ohio State University
  Presenting Author

• Tiffany Jamie Foster, Ohio State University

• Laura M. Justice, Ohio State University

• Hugo Gonzalez Villasanti

• Daniel Messinger, University of Miami

• Lynn K Perry, University of Miami

• Dwight Irvin, University of Kansas