Teaching Students to Analyze AI-Generated Writing
Context
After beginning the semester with foundational writing concepts, students move into analyzing AI-generated writing. Before completing this activity, they read research articles from scholarly journals and educational websites and annotate them to learn the required components of a peer-reviewed research paper.
As a class, students then examine a comparison paper generated by ChatGPT. Together, the class annotates the paper, discusses questions, and evaluates its strengths and weaknesses. Students also complete a proofreading questionnaire and keep the annotated paper as a reference for future writing analysis activities. The attachment for this activity is titled Compare Practice Paper.
To introduce their first major writing assignment, students write a comparison paper on a topic of their choice using the AI platform of their choice. Two additional attachments provide examples of student work from this first assignment.
Students have already practiced prompt writing before beginning the assignment. During this first major paper, they discover that effective prompting requires precision and multiple layers of instruction. They are often impressed by the speed with which AI produces a complete paper.
Although no specific prompt is included in this submission, future prompts could be displayed using semantic markup such as:
Write a comparison paper on [Topic] using APA style...
As students evaluate the AI-generated papers, they begin to recognize recurring weaknesses. They observe that introductions are often generic, thesis statements may lack clarity, and the body of the paper does not always align with the thesis.
One of the most significant issues students identify is that AI-generated papers frequently lack validated, specific evidence to support their claims. They also discover that quotations and citations may be inaccurate, fabricated, or difficult to verify, reinforcing the importance of evaluating AI-generated content critically.