🤖 AI Summary
Children with autism spectrum disorder (ASD) often experience vocabulary acquisition deficits that impede social communication; existing technological interventions frequently lack structured turn-taking mechanisms essential for authentic social interaction. This study introduces a dyadic collaborative vocabulary learning game, the first to embed a formalized social turn-taking framework into intervention design—mimicking the rhythmic alternation and role distribution characteristic of natural dialogue. Methodologically, we integrated gamified design, human-computer interaction prototyping, and clinical co-design—including in-depth interviews with 10 speech-language pathologists—to develop a functional prototype. Empirical evaluation demonstrated significantly improved child engagement and task adherence, alongside support for dynamic feedback and cross-contextual generalization. The core contribution is a closed-loop “turn-taking → comprehension → application” model that aligns lexical learning processes with real-world pragmatic communication demands, establishing a scalable, socially grounded technological paradigm for social-pragmatic intervention in ASD.
📝 Abstract
Children with Autism commonly face difficulties in vocabulary acquisition, which can have an impact on their social communication. Using digital tools for vocabulary learning can prove beneficial for these children, as they can provide a predictable environment and effective individualized feedback. While existing work has explored the use of technology-assisted vocabulary learning for children with Autism, no study has incorporated turn-taking to facilitate learning and use of vocabulary similar to that used in real-world social contexts. To address this gap, we propose the design of a cooperative two-player vocabulary learning game, CoVoL. CoVoL allows children to engage in game-based vocabulary learning useful for real-world social communication scenarios. We discuss our first prototype and its evaluation. Additionally, we present planned features which are based on feedback obtained through ten interviews with researchers and therapists, as well as an evaluation plan for the final release of CoVoL.