Scaffolding Writing Skills Using Automated Essay Generation Systems
Summary of the project
The innovative initiative "Using Generative AI in the Teaching of Academic Writing: Scaffolding Writing Skills Using Automated Essay Generation Systems" aims to enhance the writing skills of engineering students at TU/e by integrating generative AI tools into the curriculum while retaining constructive alignment. Recognizing the challenges students face in developing writing skills, this project leverages advanced AI tools to provide personalized, automated, and timely feedback, thereby improving the overall learning experience.
Background Information: Writing is a critical skill for engineering students, essential for building scientific knowledge and reporting research findings. However, many students struggle with various aspects of writing, such as formulating ideas, structuring arguments, and critical reflection. Traditional methods of providing feedback and grading are time-consuming and often inconsistent. Automated essay generation tools and generative AI like ChatGPT have shown promise in assisting students with their writing tasks. However, the use of large language models may also undermine the principle of constructive alignment between learning objectives and assessment methods, impacting students and teachers alike. Therefore, this project aims to explore how generative AI can be effectively used to assist student learning, improve writing skills, and enhance understanding of course content.
Aim of the project
The primary aim of the project is to develop and implement a pedagogical activity that utilizes generative AI tools to effectively support students in their writing.
Project Objectives:
- Impact of generative AI on higher education: The project aims to provide an overview of the scientific literature as well as case studies at TU/e on the effects of generative AI tools on writing instruction in higher education.
- Enhance Writing Skills: The project aims to improve students' writing skills by providing structured, scaffolded support through pedagogical activities combined with generative AI tools, helping them to develop better arguments and clearer writing.
- “AI-proof” assessment methods: The project aims to create AI-proof assessment methods that allow teachers to measure the extent to which students have mastered the ability to write clearly and effectively, either in collaboration with relevant AI systems or on their own.
- “Future-oriented” learning objectives: The project aims to identify and articulate future-oriented learning objectives to be achieved in writing-based university courses.
The project will be implemented in several courses across different levels and departments, including Philosophy and Ethics of AI, Thinking and Deciding, Data Science Ethics, and E3 Challenge 2 on Data and Health-Mobility-Energy. The evaluation will involve a combination of track-changes analysis, rubric-based assessment, automated linguistic analysis, and student surveys and interviews. The ultimate goal is to provide clear recommendations for using generative AI in higher education, contributing to the broader educational objectives of TU/e.
Results and learnings
The project is still ongoing. We are currently finalizing our literature review and implementing pilot studies with generative AI in our courses. An overview of the pilot studies, as well as tips for prompt engineering, guidelines, and the usage of AI in teaching can be found at https://tueindhoven.ai/. A full report can be found below as well as an infographic focused on recommendations for educators derived from the report.
For more information, please contact:
