<![CDATA[The adoption of artificial intelligence (AI) in education offers new opportunities for assessment while challenging traditional approaches focused on final products and standardized tests. This study aimed to develop an AI-enhanced framework for assessing multidimensional learning evidence in progressive classrooms. A Design-Based Research (DBR) methodology was employed through iterative cycles of analysis, design, implementation, evaluation, reflection, and redesign. The framework integrates authentic learning tasks, product and process evidence, performance, interaction, reflection, AI analytics, teacher judgment, and formative feedback. Data were collected through classroom observations, interviews, student learning artifacts, assessment records, AI-generated analytics, expert validation, and participant feedback. The findings indicate that the framework extends assessment beyond final products by integrating diverse evidence generated throughout the learning process. Iterative implementation demonstrated improvements in assessment quality, usability, feedback usefulness, teacher satisfaction, and student engagement. Expert evaluation indicated high perceived validity, authenticity, usability, and feasibility, while AI-generated assessment insights showed strong alignment with teacher judgments. The study positions AI as a decision-support mechanism rather than an autonomous grading tool, preserving teachers’ professional and contextual judgment. The proposed framework contributes an integrated, human-centered approach to authentic assessment and provides design principles for AI-supported assessment in progressive educational contexts.]]>
Copyrights © 2026
