Implementation of a metacognition-based deep learning approach to improve metacognitional awareness and understanding of geometry concepts of mathematics education students

This study aims to apply and test the effectiveness of a Metacognition-Based Deep Learning approach to improve metacognitive awareness and understanding of geometry concepts among Mathematics Education students at Universitas Halu Oleo. This study used a quasi-experimental design with a One-Group Pretest-Posttest Design. The research subjects were 108 second-semester students of the Mathematics Education Study Program divided into three parallel classes (Class A, B, and C, each with 36 students). The research instruments were a metacognitive awareness questionnaire adapting the Metacognitive Awareness Inventory (MAI) from Schraw & Dennison and a geometry concept comprehension test based on the revised Bloom's taxonomy. The results show: (1) The application of Metacognition-Based Deep Learning approach, which integrates problem orientation, collaborative exploration, thinking strategy reflection, concept connection, and self-evaluation, ran with very high validity based on expert assessment (score = 4.23/5.00); (2) Metacognitive awareness increased significantly with N-Gain = 0.41 (high category), from an average of 87.57 to 92.62, confirmed by paired t-test (p < 0.05); (3) Geometry concept understanding improved with N-Gain = 0.25 (moderate category); (4) Mediation analysis shows metacognitive awareness acts as a strong mediator with a total influence of 74%. This study makes a real contribution to the application of metacognition-based mathematics pedagogy at the university level