Using GenAI to develop a feedback bank

Well-crafted feed-forward comments that identify areas for improvement and offer advice for student development can be time-consuming to compose. The driver for this activity is educators identifying the feedback points within student work, then using AI to help flesh out the idea into a meaningful and actionable feedback comment.

What actions were taken?

	When marking a single assessment task, identify repeated areas for development that appear across multiple students' submissions. Consider overarching themes that could apply to all, such as use/integration of evidence, quality of evidence, paragraphing, clarity and task structure. Write the idea down as a basic concept. For example, here are a few feedback statements that are commonly used for undergraduate essays: What’s the relevance? Needs more explanation. This is unclear. Reference? Argue your point.
	Write a prompt (below) that takes the feedback comments above and fleshes them out into meaningful and comprehensive statements. Prompt: Take these comments and create a series of short but helpful feedback comments for a written assessment task. Explain the issue and how to correct it in a simple and concise way. 1. What’s the relevance? 2. Needs more explanation. 3. This is unclear. 4. Reference? 4. Argue your point. Create each feedback comment individually. Responses: (Copilot) This point feels disconnected from your main argument. Make sure every idea directly supports your thesis or topic. You've introduced an idea, but it's underdeveloped. Add more detail or examples to clarify your meaning and strengthen your point. The wording here is confusing. Try rephrasing for clarity—consider breaking it into shorter sentences or using simpler language. This claim needs a source, located near the evidence. Include a citation to support your statement and show where the information comes from. You've stated a position, but it lacks justification. Provide evidence or reasoning to persuade the reader and make your argument convincing.
	To make these statements more usable for students, it’s recommended to add suggestions for improvement. You can either focus on small, actionable tasks for the students or leverage the information provided in Learn HQ. It’s recommended that you do this yourself to ensure your own familiarity with the resources on Learn HQ. However, this can also be prompted: Prompt: Find articles in Monash's Learn HQ and add a support link to each idea. (Note: Copilot identified Learn HQ links, but not the most helpful links for the feedback comment) This point feels disconnected from your main argument. Make sure every idea directly supports your thesis or topic. This section on Learn HQ will help you to edit for relevance and coherence. You've introduced an idea, but it's underdeveloped. Add more detail or examples to clarify your meaning and strengthen your point. Take a look at the information on the Learn HQ page ‘Refine your argument’ to connect your ideas more meaningfully. The wording here is confusing. Try rephrasing for clarity—consider breaking it into shorter sentences or using simpler language. This section of Learn HQ will help you write clear sentences. This claim needs a source, located near the evidence. Include a citation to support your statement and show where the information comes from. Take a look at the information on the LearnHQ page ‘Integrate your sources’ to consider the different ways of acknowledging your sources. You've stated a position, but it lacks justification. Provide evidence or reasoning to persuade the reader and make your argument convincing. This section of Learn HQ will help you create a good argument.
	Organise these statements into a single document, categorising them appropriately, ideally in headings that match the rubric criteria. Now we have a simple and meaningful feed-forward comment that can be applied to multiple tasks. While this may take time initially, over time, the workload will reduce as the bank is built up. Ideally, this process should form part of the marking, and the bank should be slowly built up over time to avoid impacting workload too heavily. Alternatively, if the assessment details haven’t changed much from the previous semester, use past feedback to create the bank before marking in the current semester.
	When applying this statement to a specific student’s work, it’s important to substantiate these comments by tailoring them to the student by identifying specific details where relevant. You've introduced an idea, but it's underdeveloped. Add more detail or examples to clarify your meaning and strengthen your point. You have stated the argument without clearly explaining how and why it connects to the population you’ve identified. Take a look at the information on the Learn HQ page ‘Refine your argument’ to connect your ideas more meaningfully.

Benefits of using GenAI to develop feedback:
- Saves time in the construction of feedback
- Offers broader feedback to students, where previously little may have been provided
- Refines the clarity and quality of written feedback

Try it out

This exemplar easy to implement.

Recommended resources and training:

Microsoft Copilot is available to all Monash staff and students, however any free AI tool will provide useful broad feedback. Discipline-specific or content-based feedback may vary in quality.
Access Microsoft Copilot
TeachHQ: Feedback
How-to: Grade using Moodle marking guides

While AI is able to help us form meaningful and simple feedback statements, these statements are based on the educator’s ability to identify strengths and areas for development within the task. Further to this, it’s critical to contextualise the feedback bank comments to the student’s individual task, asserting the educator’s voice in the process.
Be aware that there is a large initial workload, but this leads to increased efficiency during the marking process.
Also, student work should not be uploaded to a third-party AI platform unless explicit consent has been provided by the student. Additionally, ensure that no identifying information is present and that the data will not be used for model training.

Supporting resources

Here are some additional resources that you can browse to help you implement this assessment.

Framework for Effective Feedback

A collaboration between Monash University, Deakin University, and Melbourne University, this empirically based study of feedback designs and conditions aims to guide educators academic developers and instructional designers.

GenAI services currently offered at Monash

Monash encourages the use of AI to innovate the way we work and create efficiency. While AI has widespread usage, recent advancements in generative AI technologies introduce complex and evolving risks.

Responsible use of AI in education at Monash

Monash University supports the responsible and ethical use of AI and Generative AI. This resource provides information about protecting data privacy, copyright and intellectual property, bias, ethics and transparency and accuracy and reliability.

Asynchronous	Any year level
Class size: Any	Individual

Learning and Teaching: Teach HQ

Learning and Teaching: Teach HQ

Using GenAI to develop a feedback bank

Try it out

Supporting resources

Framework for Effective Feedback

GenAI services currently offered at Monash

Responsible use of AI in education at Monash

Using GenAI to develop a feedback bank

About this example

Faculty of Medicine, Nursing and Health Sciences

Try it out

Supporting resources

Framework for Effective Feedback

GenAI services currently offered at Monash

Responsible use of AI in education at Monash