Scaffolding Academic Success: The Role of Instructor-Guided GPTs in Managing Task Complexity for Pre-Service Teachers

Vollständiger Abstract

Worum geht es in dieser Arbeit?

This study evaluates the instructional impact of Teacher’s Bag, an instructor-designed, context-aware custom GPT system, on pre-service teachers’ academic achievement. Utilizing a course-embedded quasi-experimental design at a Turkish public university, the research analyzed performance-based assessments in Curriculum Development and Educational Research Methods courses. The system provided pedagogically constrained, task-specific scaffolding aligned with national curricular objectives. Results revealed tasksensitive outcomes. In Curriculum Development, the AI functioned as a stabilizing scaffold, maintaining performance despite significantly increased final task complexity. In Educational Research Methods, the system enabled students to navigate a major transition from quantitative to mixed-methods designs, though accompanied by lower mean scores reflecting heightened cognitive demand. These findings challenge the “AI as a shortcut” narrative, demonstrating that pedagogically designed AI supports meaningful engagement with complex tasks without artificial grade inflation.