Generative AI as a Grading Partner
As noted in the Faculty Handbook, faculty are responsible for “evaluating students’ work, including providing adequate and timely feedback” (Faculty Handbook, 2021, p. 10). Large language models (LLMs) can be valuable tools for enhancing grading consistency, efficiency, and rationale behind evaluation. Thoughtful and ethical incorporation of AI into grading practices preserves high standards of academic integrity, the centrality of instructor expertise, and the essential human element of teaching and learning.
Important Reminders
- Protect student privacy: Use only University approved tools. For further guidance, please see the Northwestern IT Guidance on the Use of Generative AI for a description of data approved for use with generative AI.
- Verify AI-generated feedback: Check all output for inaccuracies and fabrications and ensure the feedback accurately reflects the work and represents its achievements or areas for improvement.
- Maintain transparency with students: Communicate to students how AI is being used in the grading process and how it supports professional judgment.
Innovative Uses
- Augmenting feedback: Combine AI tools with professional observations and comments to provide a well-rounded evaluation.
- Grade norming: Leverage AI tools to standardize grades across various sections or assignments, ensuring uniformity in grading criteria.
- Crafting growth-focused comments: Formulate constructive and growth-oriented feedback for students, such as identifying strengths to build upon or offering actionable steps for further development.
Sample Prompts
- Write assignment instructions and a rubric: Use these prompts to write an assignment and a rubric that can be used for feedback and scoring.
- Create a low-stakes quiz: This prompt will create a multiple-choice quiz on a subject of an instructor's choice. With some extra instructions, an LLM can create the questions in a format that can be imported into Canvas.