Abstract
This mixed-methods study investigated the impact of algorithmically driven adaptive learning pathways on metacognitive awareness and academic achievement among middle school students in blended mathematics classrooms. Over a 16-week intervention involving 342 eighth-grade students across two public middle schools, participants were assigned to either an experimental group utilizing an intelligent adaptive mathematics platform or a control group engaging with traditional blended curriculum materials. Quantitative data gathered via pre- and post-intervention administrations of the Metacognitive Awareness Inventory (MAI) and standardized mathematics achievement tests were complemented by semi-structured student interviews, teacher focus groups, and platform telemetry logs. Quantitative findings demonstrated statistically significant gains in both procedural and conceptual mathematics achievement for the experimental cohort, accompanied by substantial improvements in the regulation of cognition subscales, specifically monitoring and evaluation. Qualitative insights revealed that real-time visual progress dashboards and micro-scaffolding prompts fostered learner agency and reduced cognitive overload during problem-solving tasks. However, variations in teacher facilitation strategies played a moderating role in students' strategic regulation habits. The study underscores the necessity of designing adaptive instructional platforms that balance automated content delivery with explicit metacognitive prompts, providing actionable implications for curriculum designers and instructional leaders seeking to optimize blended mathematics environments.
