AI in Feedback and Marking
Information Guide:
Thinking about Feedback with AI
Feedback is crucial for student learning, but delivering quality feedback is challenging. Here, we outline eight principles for effective feedback (adapted from Henderson and Philips, 2014) and show how AI can enhance instructor-provided feedback, such as written comments, making the process more efficient and personalised for students.
Information Guide:
Things to consider when using AI for Feedback
When using AI for feedback, educators must actively participate and monitor the process to ensure quality. AI can improve feedback by editing, generating ideas, acting as a peer, or providing immediate responses. Ensure student communication, consent, data privacy, quality assurance, and fairness when integrating AI into feedback practices.
Examples
Using AI to develop a feedback bank
A feedback bank is a collection of reusable, pre-written comments for student work. While feedback should be unique and tailored, similarities in student work benefit from consistent comments. AI can save time and effort, crafting scaffolded, clear, and consistent feedback. Educators identify feedback points, and AI fleshes them out. This approach saves time, offers broader feedback, and refines quality. Consider the initial workload, which leads to increased efficiency.
Scaffolded assessment feedback from (LLM) AI - basic
Students can upload work via a secure portal using an API, ensuring data security and non-retention for training. This method suits well-represented content, offering feedback on broader elements for specialised domains. Benefits include developing evaluative judgement and early feedback. Collaboration with educational designers is recommended for creating a seamless web interface.
Leverage GenAI & Voice Type to Craft Detailed Personalised Feedback Fast
Struggling to provide detailed, personalised feedback while managing time constraints? AI offers a solution. I’ve developed and tested a step-by-step method to leverage AI and voice typing for crafting feedback quickly. This approach protects student privacy without uploading their work to AI, ensuring high-quality feedback in less time. Ready-to-use resources are also included.
Providing feedback on written assignments with generative AI
Using generative AI with ChatGPT Plus helps draft feedback on student writing. Custom GPT offers control over AI responses and data usage. While uploading student work can guide your feedback, consider potential bias. Use AI after forming your opinion. Benefits include increased accuracy, reduced bias, and time savings. Ensure student consent to comply with data privacy laws.
Simulated job/grant application - ATLAS
With AI screening platforms rising, students must learn to decode job descriptions, follow steps, and meet formatting requirements. Performing arts job seekers face writing challenges. AI drafting isn’t always ideal, but using AI simulators for quick feedback and human insights can help students develop effective application processes and improve expertise.
Using AI in Peer-to-Peer Feedback
Using AI in peer feedback fosters self-regulation and active learning, challenging traditional teacher-led methods. AI-generated frameworks help students provide constructive comments, saving time and enhancing feedback literacy. Considerations include the need for moderation and data protection. Suitable for written assignments or video presentations with generative AI like ChatGPT.
Contributors: Dana Bui , Michael Crocco, Tim Fawns, Joel Moore, Tridib Saha, Ari Seligmann, Zachari Swiecki, Thao Vu, Pauline Wong