AI Tools for Providing Student Feedback

Introduction

Providing timely, constructive feedback is one of the most important yet time-consuming aspects of teaching. This guide explores how AI tools can help engineering educators provide more effective feedback to students while reducing the administrative burden of feedback generation.

Benefits of AI-Enhanced Feedback

• Timeliness: Students receive feedback faster, when it's most relevant
• Consistency: Maintain similar quality and depth across all student submissions
• Scalability: Provide detailed feedback even in large classes
• Personalization: Tailor feedback to individual student needs
• Multilayered: Address technical content, communication, and metacognitive aspects

AI Tools for Engineering Feedback

Large Language Models (LLMs)

Tools: ChatGPT, Claude, Bard

Applications:

• Generating initial feedback drafts for assignments
• Suggesting improvements to technical reports
• Providing explanations for common misconceptions
• Offering multiple perspectives on design choices

Specialized Feedback Tools

Tools: Gradescope AI, FeedbackFruits, Turnitin

Applications:

• Automated grading of problem sets
• Identifying common errors across submissions
• Suggesting targeted resources based on error patterns
• Checking code functionality and style

Implementation Strategies

1. Feedback Templates with AI Enhancement

Create feedback templates for common assignment types, then use AI to:

• Customize template sections based on specific student work
• Generate examples or counterexamples relevant to student submissions
• Suggest additional resources tailored to student needs

2. Two-Stage Feedback Process

1. AI-Generated Initial Feedback: Use AI to provide immediate, basic feedback
2. Instructor Enhancement: Review and enhance AI feedback with discipline-specific insights

3. Student-Directed AI Feedback

Guide students to:

• Submit their work to AI tools with specific feedback prompts
• Reflect on AI feedback before instructor review
• Compare AI feedback with instructor feedback to develop critical evaluation skills

Best Practices

Designing Effective AI Feedback Prompts

Example Prompt Structure:

I am an engineering professor providing feedback on a student's [assignment type].
The assignment asked students to [assignment goals].
Here is the student's work: [student submission]
Please provide:
1. 2-3 specific strengths of this submission
2. 2-3 areas for improvement with specific suggestions
3. Questions that would prompt deeper thinking
4. Connections to relevant course concepts

Maintaining the Human Element

• Always review AI-generated feedback before sharing with students
• Add personal observations and connections to class discussions
• Include encouragement and recognition of individual progress
• Be transparent with students about AI use in feedback generation

Ethical Considerations

• Ensure feedback addresses the student's actual work, not assumed work
• Be mindful of potential biases in AI-generated feedback
• Maintain appropriate privacy and data security
• Use AI as an enhancement to, not replacement for, instructor expertise

Case Examples

Example 1: Circuit Analysis Homework

An instructor uses ChatGPT to generate detailed explanations of solution approaches for common errors, then customizes these explanations for individual students.

Example 2: Engineering Design Reports

Students submit draft reports to Claude for initial feedback on structure and clarity, then the instructor provides domain-specific technical feedback.

Example 3: Programming Assignments

GitHub Copilot is used to generate code improvement suggestions that the instructor reviews and incorporates into personalized feedback.

Assessment of AI-Enhanced Feedback

Evaluate the impact of AI-enhanced feedback by measuring:

• Student perception of feedback quality and usefulness
• Time from submission to feedback delivery
• Changes in subsequent assignment performance
• Student self-efficacy and motivation

Conclusion

AI tools offer significant potential to enhance the feedback process in engineering education. By thoughtfully integrating these tools into your feedback workflow, you can provide more timely, detailed, and personalized feedback while managing your workload more effectively. The key is to maintain the human connection that makes feedback meaningful while leveraging AI to handle routine aspects of feedback generation.