There are theoretical models to measure EFL students' use of AI tools in self-assessment or feedback. the Self-Regulated Learning (SRL) model by Zimmerman is commonly used, focusing on planning, monitoring, and reflecting. Also, the Technological Acceptance Model (TAM) measures how students use AI, like Grammarly, to check their work. For example, in SRL, student use Grammarly to check their mistakes, replace suitable words with the feedback, and reflect on their progress. in TAM, a questionnaire asks if students find AI helpful and easy for assessment tasks, like improving essays.

@ Noorhani Laksmi, thank you very much for your valuable response. Could you share link to these work, pleae?

[2505.08672] Cómo los estudiantes usan la retroalimentación de la IA importa: evidencia experimental sobre el rendimiento y la autonomía de la física

[2302.09319] MAILS -- Meta AI Literacy Scale: Desarrollo y prueba de un cuestionario de alfabetización de IA basado en modelos de competencia bien fundamentados y cambio psicológico y metacompetencias

La IA es un recurso poderoso para la auto-retroalimentación, siempre que se integre bajo un enfoque instruccional sólido, con respeto a la diversidad de estilos de aprendizaje y considerando principios del conectivismo. Además, su aplicación debe enmarcarse en competencias docentes alineadas con EC0366 y EC0362, asegurando así un uso responsable, formativo y con respaldo institucional. Saludos.

In my view, the most useful “model” is a hybrid: TAM/UTAUT to capture adoption (usefulness, effort, support), layered onto a Self-Regulated Learning cycle to track when and how students use AI for self-assessment and feedback. I’d add a small SDT/metacognition block (autonomy, strategy awareness) and explicit AI-literacy/ethics items (transparency, attribution). Practically, pair the survey with prompt/usage logs and revision gains, so perceived use aligns with demonstrated improvement.