In this article, the authors explore how students' argumentation skills, particularly strategic decision-making and persuasion, can be improved using machine learning. They address the challenge of providing scalable and personalized feedback to enhance these skills. The study uses machine learning to offer scalable, immediate feedback, which is important for educational innovation.

A dynamic ML-based system was developed and tested in three empirical studies, comparing it with traditional script-based and adaptive supports. This approach allows for the evaluation of the effectiveness of dynamic modeling in improving argumentation skills.

The results show that the dynamic system significantly enhances students' argumentation skills across various tasks and skill levels, outperforming traditional methods. It is effective for both complex and simple argumentation tasks, providing robust support tailored to individual needs.

This research contributes new insights by demonstrating the effectiveness of dynamic ML-based modeling in areas like persuasive writing, beyond traditional STEM fields. It highlights the potential of adaptive learning technologies to transform educational practice and policy by promoting critical thinking and communication skills.

The paper can be accessed at: https://pubsonline.informs.org/doi/10.1287/isre.2021.0615