Hong Kong Kindergarten Children’s Play with Educational Robotics: Child Engagement and Its Association with Computational Thinking

• In Event: Computational Playgrounds: Foundations and Futures of AI, Play, and Pedagogy in Early Childhood

Abstract

Xiaoyi Hu, The Education University of Hong Kong, hxiaoyi@eduhk.hk, Presenting author
Weipeng Yang, The Education University of Hong Kong, wyang@eduhk.hk, Presenting author
Hui Li, The Education University of Hong Kong, huili@eduhk.hk, Presenting author
The integration of educational robotics in early childhood settings has emerged as a promising approach to foster foundational skills in young learners (Garvis & Keane, 2023). In Hong Kong, the adoption of coding robots in kindergartens provides a unique opportunity to explore how children are engaged in free play with educational robotics (Yang et al., 2022). This study investigates the dynamics of young children’s engagement with a coding robot and examines how this engagement relates to computational thinking (CT). By focusing on individual engagement levels, this research aims to shed light on the potential of robotics to support early childhood education, with the following guiding questions: (1) How do kindergarten children engage in free play with a coding robot? (2) Is child engagement in playing with educational robotics associated with age, gender and socioeconomic status (SES)? and (3) Is children’s CT associated with their engagement levels when playing with educational robotics, after controlling for their demographic characteristics (i.e., age, gender and SES)?
This cross-sectional study collected data from 313 children (49.8% girls, M = 62.88 months, SD = 4.29) of nine Hong Kong kindergartens. Children’s engagement with a coding robot called Tale-Bot was assessed with the Individual Child Engagement Record (Kishida et al., 2008) and Child-Coding Robot Interaction Observational Scale (CCRIOS; Xiang et al., 2025). Their CT was assessed with Xiang et al.’s (2025) Computational Thinking Observational Scale (CTOS).
Frequency analysis showed that 8.0% of children were of Highly Active Engagement (Highly AE) level, 29.1% of children were of Moderate AE level, 44.1% of children were of Lower AE Level and 13.7% of children were of Passive (Non-)Engagement (PN) level. And 5.1% missing data. Children's overall engagement measured by CCRIOS ranged from 11 to 55 (M=41.66, SD=10.8). CTOS Scores ranged from 15 to 71 out of 75(M=43.87, SD=12.98).
Demographic associations with engagement are summarized in Table 5-1. Only SES showed a significant relationship with continuous engagement (CCRIOS scores), with higher SES weakly predicting greater engagement (ρ = .18, p = .012), accounting for 3.2% of variance. No other demographic variables significantly related to either engagement measure, with all remaining p-values > .05. As shown in Table 5-2, ICER Engagement level significantly predicted computational thinking (CTOS) beyond demographic factors. While demographics alone explained negligible variance (R² = .007, p = .19), adding ICER engagement levels caused a large significant increase (ΔR² = .329, p < .001), with all active engagement levels showing positive effects. Highly and moderately engaged children demonstrated particularly strong CT advantages compared to passive engagement. Table 5-3 shows similar results for continuous engagement: CCRIOS scores caused a substantial predictive improvement (ΔR² = .419, p < .001), with each 1-point increase corresponding to 0.79 higher CTOS points (B = 0.79, SE = 0.054). In both models, demographic variables became non-significant after accounting for engagement.
The study highlights that greater engagement during free play with coding robots is associated with stronger CT skills in kindergarten children, while controlling for factors such as age, gender, and SES. Despite children from higher socioeconomic backgrounds demonstrating marginally higher levels of engagement, neither age nor gender exhibited a significant impact on engagement levels. It is important to note that children who displayed a greater degree of engagement, particularly those in the moderate to highly range, showed a significantly stronger development of CT skills in comparison to their passively engaged peers. The findings emphasize the pivotal role of active, sustained interaction with educational robotics in facilitating early CT development. These findings underscore the potential of educational robotics to enhance learning and engagement in the early years, emphasizing the importance of inclusive pedagogical approaches specifically designed to maximize and sustain children's active, meaningful involvement with educational robotics.

Authors

• Xiaoyi Hu, The Education University of Hong Kong

• Weipeng Yang, The Education University of Hong Kong

• Hui Li, The Education University of Hong Kong