A methodology for AI-supported cross-level professional collaboration in computer science education using the AI-Ustaz platform

Abstract

Contemporary educational environments require educators to participate in professional networks that extend across institutional levels. This study presents a methodology for AI-supported cross-level professional collaboration involving pre-service teachers, secondary-school teachers, and college instructors, with the AI-Ustaz platform serving as the shared technological environment. The methodology comprised three sequential stages: theoretical–orientational, technological, and project–networked. A cluster quasi-experimental pretest–posttest design was implemented during the 2024–2025 academic year with 347 participants across nine educational institutions in Kazakhstan. Participants were allocated to an experimental group (n = 178) and a control group (n = 169). Cognitive, performative, and reflective components of networked-collaboration competence were assessed. Descriptive analyses indicated greater pretest-to-posttest improvements in the experimental group across the cognitive, performative, and reflective components. At posttest, the experimental group required approximately 61% less time to prepare instructional materials than the control group and received higher expert ratings for content quality. The findings suggest that AI-supported cross-level professional collaboration may provide a promising model for educational practice. However, the results should be interpreted cautiously because participants were nested within institutional clusters and cluster-adjusted analyses could not be performed..