Dissertation Defense Announcement Ph.D. in Education Program: Changyun Liu “Personalized Interactive Cognitive Scaffolding (PICS) for High School Students Struggling with Mathematics in Computer-Based Assessment”

2:00 pm - 3:30 pm

This dissertation investigates the effectiveness of an AI-based platform, the Personalized Interactive Cognitive Scaffolding (PICS) system, in supporting high school students’ performance on algebra word problem–solving tasks, particularly for students with mathematics learning difficulties (MD). Algebra word problem solving (WPS) presents significant challenges due to its cognitive and linguistic demands. In this study, PICS enabled teachers to generate individualized computer-based assessments with embedded testing accommodations and AI-generated scaffolding tools, including linguistic simplification, contextual–semantic modification, diagram generation, and step-by-step problem breakdown.

A mixed-methods research design was employed. Study 1 examined teachers’ evaluations of AI-generated items using a structured rubric to assess the quality of the four scaffolding tools. Results indicated that linguistic supports and problem breakdown were rated most favorably, while diagrams showed greater variability due to concerns related to accuracy and clarity. Study 2 used a within-subject, quasi-experimental design to compare student performance under conditions with and without PICS scaffolding. Results demonstrated that students’ problem-solving accuracy improved significantly when PICS scaffolding was available. PICS functioned as an access support within the assessment rather than as an instructional or tutoring system.

These findings provide empirical evidence that AI-generated scaffolding can enhance accessibility and support student performance without altering mathematical demands.

To attend this event virtually and for more information, please contact academic.services@gse.rutgers.edu.