Techniques for using artificial intelligence to foster active learning

5. Automated real-time feedback

Feedback is one of the most powerful elements of learning and teaching in higher education, as it provides learners with insight into the quality of their understanding, reasoning, and application of knowledge. In traditional settings, learners often receive feedback only after completing the learning activity, which limits opportunities for learning while cognitive activity is still ongoing.

Automated real-time feedback enables learners to immediately recognise errors, receive guidance, and adjust their approach to solving tasks during the learning process, thus encouraging active analysis, reflection, and the development of self-assessment skills.

AI-based systems provide learners with immediate, informative, and personalised feedback during the task-solving process. This approach encourages active learning because the learner:

• Immediately notices the error and its cause,
• Questions their own thinking,
• Decides how to adjust their approach,
• Adopts educational content through iteration and real-time reflection.

The instructor acts as an expert moderator who interprets the data collected by the system and encourages discussion about errors and different problem-solving strategies.

Example of an AI tool:

Gradescope is an AI-based system widely used in higher education for grading and analysing student assignments, including essays, math problems, and computer code. The system allows for digital scanning of paperwork or uploading assignments in PDF format, and then automatically analyses and groups similar answers. It recognises correct and incorrect solution patterns and provides personalised feedback in a short time.

The quality of the feedback requires pedagogical supervision by the instructor. Implementing advanced functionalities requires an institutional subscription.

Example in higher education:

In the course “Linear Algebra”, students solve tasks involving calculating determinants and solving systems of linear equations. After submitting solutions via Gradescope, the system automatically analyses the procedures and identifies the most common mathematical errors, such as incorrectly performed elementary transformations or mistakes in calculating determinants.

Students receive immediate feedback that points to the specific step where the error occurred, along with a brief guideline encouraging them to check the logic of their procedure. After correcting their solutions, students compare their initial and improved approaches and explain what additional thinking led them to the correct solution.

The teacher uses the overview of the most common errors to initiate a discussion about different strategies for solving systems of equations and the importance of checking procedures. Such discussion fosters a deeper understanding of mathematical concepts and encourages students to reflect metacognitively on their own learning.