ASSESSING INDEPENDENT LEARNING IN AI-SUPPORTED PHILOLOGY TASKS: A PROCESS-EVIDENCE MODEL UDK

Abstract

Abstract. Independent learning has traditionally been assessed indirectly through essays, translations, presentations, and other academic products that students prepare outside the classroom. The widespread use of generative artificial intelligence in everyday learning has weakened the connection between the quality of a finished text and the learning process through which it was produced. A fluent paragraph, coherent argument, or polished translation can now be generated without the planning, verification, and revision that such academic work has traditionally required. This challenge is particularly relevant to philology programmes, where the assessed product is predominantly textual and therefore can be convincingly generated by artificial intelligence systems. This article develops a conceptual model for assessing the independent learning of philology students in AI-supported learning environments. Drawing on self-regulated learning theory and research on learner autonomy, the article identifies four interrelated components of independent learning that can be assessed through process evidence rather than product quality alone: goal setting and planning, strategic action, monitoring and self-evaluation, and source and tool literacy. The proposed model emphasises the assessment of how students formulate learning goals, select and use resources and AI tools, verify information, revise their work, respond to feedback, and evaluate the effectiveness of their own learning strategies. Keywords: independent learning, self-regulated learning, assessment methodology, artificial intelligence tools, philology education, learner autonomy, process evidence, feedback literacy, academic integrity, higher education.