Implementing Deep Learning in Arabic Language Education at Madrasah Aliyah: Enhancing Critical Thinking and Contextual Learning

This study examines the implementation of deep learning in Arabic language education at Madrasah Aliyah (MA) in Indonesia. Employing a descriptive qualitative approach, data were obtained through library research and in-depth interviews with five Arabic teachers from different madrasah. The study identified five key instructional strategies applied by teachers: critical discussion, problem-solving, project-based learning (PjBL), student presentations, and the integration of digital tools and AI-based applications such as Kahoot and Quizizz. Critical discussion was consistently implemented across all madrasah, while the adoption of other strategies varied depending on institutional context and teacher readiness. Student responses toward interactive and technology-supported strategies were generally positive, demonstrating increased engagement and participation. Nevertheless, several challenges hinder optimal implementation, including limited technological infrastructure, rigid curricula, insufficient professional training, and assessment practices dominated by objective tests and lower-order thinking skills (LOTS). Authentic, project-based, and reflective assessments remain underutilized, resulting in a misalignment with the constructive alignment principle and limiting the comprehensive evaluation of learning outcomes. The findings suggest that while deep learning principles have begun to take root in Arabic language teaching at MA, achieving their full potential requires targeted interventions in infrastructure development, teacher capacity building, and the alignment of assessment practices with higher-order learning objectives.