Lessons from Real-World Settings: What Makes It Uniquely Difficult to Design Cognitive Training Programs for Children with Autism Spectrum Disorder and Other Developmental Disabilities

Research Background and Issues

• Problem or Challenge: Despite the high prevalence of Autism Spectrum Disorder (ASD) and other developmental disorders (DD), many affected children face barriers—economic, geographical, or psychological—that prevent them from accessing necessary health and educational services. Moreover, existing software solutions developed in research and practice often lack validation in real-world environments and detailed information about their development processes, making it challenging to effectively design technologies that support children with ASD and DD.
• Significance: Children with autism require specialized support in their cognitive and social skill development, yet existing technologies fail to adequately meet these complex needs. Addressing these issues can contribute to the development of higher-quality and more accessible educational technology solutions.
• Research Motivation and Related Work: Most ASD software solutions remain at the prototype stage, with their concepts validated only under experimental conditions, lacking in-depth analysis and adjustments for real-world usage scenarios. Additionally, similar studies have failed to provide specific guidance on technical details and design decisions, hindering the dissemination of innovative practices.

Solution

• Proposed Approach:
  • This study uses the mobile cognitive training program Dubupang, developed by the South Korean startup Dubu Inc., as a case study to analyze its design process, technical foundations (e.g., discrete trial training, applied behavior analysis, zone of proximal development theory), key design issues, and challenges in real-world environments.
  • Through field observations and interviews, the study explores whether the technological design adequately supports the needs of children with ASD and examines the role of caregivers in the technology usage process.
• Innovations:
  • The research not only focuses on technological design but also discusses the importance of caregiver involvement, emphasizing the synergy between technology and human intervention.
  • It proposes design recommendations for improving reward systems and activity personalization using Generative Artificial Intelligence (Generative AI).
  • Adaptive systems are introduced into the design framework to dynamically adjust user interfaces and task difficulty.
• Implementation Steps:
  • Conducted two months of field observations to understand the interaction patterns between children with ASD and Dubupang.
  • User research included two rounds of semi-structured interviews with caregivers and expert interviews, analyzing design features, challenges, and lessons learned from real-world deployment.
  • Qualitative analysis of the data was performed using open coding to summarize themes and extract design recommendations.

Research Outcomes

• Specific Findings:
  • Identified four key areas influencing design: attention and engagement, personalization and adaptability, instructional design and feedback, and interaction design.
  • Found that caregivers' strategic interventions (e.g., physical guidance, behavioral rewards, and verbal repetition) are indispensable in real-world scenarios, as technological design alone cannot fully address the nuanced and dynamic needs of children.
• Advantages:
  • Integrated human factors and real-world deployment experiences into the foundation of technological design, addressing the lack of real-world validation in previous studies.
  • Proposed the use of artificial intelligence and modular design to enhance the feasibility and personalization capabilities of existing solutions.
• Experimental or Evaluation Results:
  • Fixed task lengths failed to accommodate variations in children's attention across different environments, leading to user interest in dynamic task length adjustments.
  • Simple language was easily understood by children but sometimes failed to address individual vocabulary differences, requiring real-time language support from caregivers.
  • Multimodal input was effective, but competition between visual and auditory elements distracted attention, highlighting the need for balanced multisensory design.
• Limitations and Future Directions:
  • This study was primarily conducted in South Korea, and the results may be limited by cultural and socioeconomic contexts. Further validation in cross-national and other social environments is crucial.
  • The balance between caregiver and technological support remains an area requiring deeper research and optimization.
  • The long-term effects of Dubupang were not evaluated. Future research could explore functional improvements and long-term impact assessments of cognitive training devices.
  • Some design recommendations (e.g., the use of Generative AI) have not yet been field-tested, necessitating experimental validation to assess their practical effectiveness.

Through this analysis, the researchers provide practical design improvement recommendations and highlight the central role of caregivers in children's learning applications. This offers significant empirical evidence and theoretical guidance for the future development of educational technologies for children with ASD and DD.