Feedback With AI

A look at feedback 

Feedback involves a communication of information, generally from the teacher to the learner, to modify or evaluate the learner’s thinking or behaviour (Meyer et al., 2024; Shute, 2008).  

Overall, a student is striving towards a learning goal. Feedback allows students to evaluate where their level of knowledge or performance is currently, and where they want it to be (Hattie & Timperley, 2007).  

Feedback, then, helps with the motivation, emotional response, and learning improvement of the student (Meyer et al., 2024).  

To provide effective feedback, teachers and students should ask themselves these questions:  

1. “Where am I going?”  

2. “How am I going?” 

3. “Where to next?” 

These three questions help to resolve the gap between a learner’s current level of knowledge/performance, and the target, or goal, level (Hattie & Timperley, 2007). The target level of knowledge or performance can be effectively outlined in the learning intentions and success criteria of a class or topic. When these are made familiar and obvious to the students, they have a greater understanding of what is expected of them, and what they may need to do to reach that (Hattie & Timperley, 2007).  

Question 1: Where am I going? 
This question helps to identify any learning goals related with the task or topic, which gives students a clear sense of direction and helps motivates them to reach their goals.  

Question 2: How am I going? 
The second question helps to determine what progress has already been made and identify any gaps in the student’s knowledge. Once answered, this provides a springboard on how to proceed, which can be encouraging to the learner.  

Question 3: Where to next? 
The final question looks to the future and leads to a continued forward momentum in learning. This gently opens up the space for further possibilities, without overloading students with all the ‘more’ that they are yet to learn.  

Ultimately, providing feedback helps students to close the gap between their current level of understanding and their goal performance.  

Using Artificial Intelligence to provide feedback

So, what does Artificial Intelligence have to do with feedback? 

Firstly, what is Artificial Intelligence? 

According to John McCarthy from Stanford University, Artificial Intelligence is, “the science and engineering of making intelligent machines, especially intelligent computer programs” (McCarthy, 2007).  

Some forms of Artificial Intelligence that you may know of, or have used, include Claude and ChatGPT. These are examples of generative AI. The ‘generative’ bit means that they can produce a written response, similar to that of a human, based upon the commands or requests provided (Lambert & Stevens, 2023).  

As well as being generative AI, they are what is known as Large Language Models (LLMs), which is a specific type of AI that is designed to sound human. This means that they have been trained on large amounts of texts, data, and information, which shapes their intelligence and forms human-like responses (Lambert & Stevens, 2023).  

LLMs are a useful tool for providing feedback to students about their work, as the teacher does not have to manually write general feedback from scratch. LLMs can automatically adapt responses to match students’ strengths and weaknesses, which can help to build and encourage their performance and productivity in the drafting and revision process.  

How to use AI to give the best feedback 

When using LLMs to provide feedback, it is important that the prompt is clear and direct. To do this, your prompt should provide instructions for the feedback to: 

Linking back to what we know about feedback… 

You may be able to see that some of the prompts relate to the questions that help provide effective feedback.  

• Prompt (a) responds to the question of ‘how am I going’ to assist in identifying any current gaps in students’ knowledge.  
• Prompt (c) addresses ‘where to next?’ in an attempt to guide the students in how best to improve their writing.  

Using Artificial Intelligence in this way can easily establish clear learning opportunities, without increasing – and potentially even decreasing – teachers’ workload. Taking advantage of AI to provide more generalised feedback can create space for teachers to focus on providing personal, individualised, feedback.  

It is important to recognise that AI-generated feedback has a greater effect on motivation than on performance. A recent study explored the effects of providing feedback generated by LLMs on students’ revision of their writing, compared to the control group who was simply told to review their work. They found that perhaps feedback is viewed as a reward, acting as extrinsic motivation, and hence increasing students’ motivation to revise their work (Meyer et al., 2024).  

To summarise 

It can be difficult to navigate the many different pathways and opinions regarding the use of Artificial Intelligence in education, particularly in providing feedback to students. However, if we take what we know about feedback, its usefulness and importance, then we can use AI to our advantage and to the benefit of our students.  

To summarise:

• Feedback aims to resolve the gap between a learner’s current and desired performance levels.  
• Effective feedback addresses the questions: “Where am I going?”, “How am I going?”, and “Where to next?”.  
• AI, particularly Large Language Models such as ChatGPT, can generate personalised feedback for students based on the right prompts.  
• This can relieve some pressures for teachers and instead create more space for them to focus on other aspects of feedback and learning.  

All the best in your AI-provided feedback endeavours! 

References 

• Hattie, J., & Timperley, H. (2007). The Power of Feedback. Review of Educational Research, 77(1), 81-112. https://doi.org/10.3102/003465430298487  
• Lambert, J., & Stevens, M. (2023). ChatGPT and Generative AI Technology: A Mixed Bag of Concerns and New Opportunities. Computers in the Schools, 1-25. https://doi.org/10.1080/07380569.2023.2256710  
• McCarthy, J. (2007). What is artificial intelligence.  
• Meyer, J., Jansen, T., Schiller, R., Liebenow, L. W., Steinbach, M., Horbach, A., & Fleckenstein, J. (2024). Using LLMs to bring evidence-based feedback into the classroom: AI-generated feedback increases secondary students’ text revision, motivation, and positive emotions. Computers and Education: Artificial Intelligence, 6, 100199. https://doi.org/https://doi.org/10.1016/j.caeai.2023.100199  
• Shute, V. J. (2008). Focus on Formative Feedback. Review of Educational Research, 78(1), 153-189. https://doi.org/10.3102/0034654307313795  
• Wang, T., & Lajoie, S. P. (2023). How Does Cognitive Load Interact with Self-Regulated Learning? A Dynamic and Integrative Model [Review]. Educational Psychology Review, 35(3), Article 69. https://doi.org/10.1007/s10648-023-09794-6