Teaching Students AI Strategies to Enhance Metacognitive Processing

Todd Zakrajsek, UNC-Chapel Hill; ITLC-Lilly Conferences on Evidence-Based Teaching and Learning

Artificial intelligence (AI) has become synonymous with efficiency and outcomes in higher education. I have read much about how AI can be used for scheduling, drafting email responses, resource management, and checking references. AI is also often used to create lecture outlines, simulations, problem-solving activities, formative assessment analyses, and content mapping. Without question, there is much to consider with respect to what AI can do and the speed at which it can be done. That said, learning is more about process than outcomes. Yes, we are certainly interested in educational outcomes; but for the most part, when it comes to learning, outcomes are designed to assess the extent to which processes have been learned. One area in which AI can be very helpful is to assist students in engaging in metacognitive processes.

Young man making notes in front of a computer. Courtesy of Unsplash.

The Processes

Metacognition is typically defined simply as knowing what one knows, or the process of thinking about one’s own thinking. It has long been shown to be an important part of learning (Flavell, 1979). To foster deeper learning, students can develop skills in the three main areas of metacognition with respect to learning: planning, monitoring, and reflecting.

Planning is thinking about how one sets up the learning situation. If you are studying for an upcoming exam, planning would include setting goals, setting criteria for success, considering which strategy will be used to study, where you will study, whether you have the resources you need, and how you will keep from being interrupted (by others and by your own phone). If you are studying with someone else, you might set parameters for the study session.

Monitoring pertains to thinking about how well the work is going and knowing whether goals are being met. This may include checking comprehension of new material after reading a block of material, determining the extent to which a new strategy is working as intended, setting aside time in a study group to talk about what is being learned, and describing any struggles experienced in learning the new material.

Reflection is done after the study session or work being completed. Here, students might think about the extent to which the study strategy selected was effective or what might need to change to keep the study group on task. This is also the time when students can determine the extent to which goals and criteria set in the planning portion of metacognition were met. If not,what learning gaps remain? Students can also reflect on how they feel about the study session (affective component).

Developing these metacognitive strategies has been shown time and again to help students become better learners (McGuire, 2015; Tanner, 2012; Zakrajsek, 2022). This is one area in which AI can be used to help students become better learners, regardless of the system used (e.g., ChatGPT, Gemini, Bard, Claude, Copilot, Perplexity, Socratic).

Cartoon silhouette of a man thinking through the planning, monitoring, reflecting process. Created 5/8/25 using OpenAI's ChatGPT, with input and design guidance provided by Todd Zakrajsek.

The Strategies

To assist in metacognitive planning:

1. Use an AI-powered goal-setting app to outline what might be accomplished in a study session and suggest a learning strategy appropriate for that block of material.

2. Explain in a prompt what should be accomplished in the study session. Have AI draft a plan of attack for learning the new material, complete with time stamps for when to take breaks and when certain blocks of material should be completed.

3. Ask AI to list areas where students tend to have problems within a given block of material.

4. Use AI to draft “ground rules” for the study group, and if desired, draw up a group study agreement for participants to sign.

5. Ask AI to generate a pre-test for the material to be learned to set study goals or simply to get an idea of how much foundational knowledge one has for the area to be studied.

To assist in metacognitive monitoring:

1. Use an AI quiz tool (e.g., Quizlet, Socrative, EdApp) to periodically ask one or two questions to check understanding.

2. Use the chat mode to ask ChatGPT if you could explain concepts as they are being learned. Have ChatGPT explain what is correct and what appears to be incorrect.

3. Use an AI dashboard to track progress toward a goal, such as a research paper or complex block of material.

4. Explain what challenges are being experienced in the material and ask an AI for alternative strategies that might work better.

5. Ask an AI for suggestions for what body stretches might be done during a 10-minute study break or use a meditation app like Headspace for a 5-minute guided breathing exercise. (e.g., “I am studying for two hours in the library, and it has been one hour and I need a break, but I don’t want to leave the building. I am physically and mentally exhausted because I didn’t get a lot of sleep last night. List 4 things I might do that would make a good cognitive break from this material.”)

To assist in metacognitive reflection:

1. Explain the block of material and ask AI to provide a typical essay-type question and give the answer a grade. Once given the grade, ask what would have made the response better and any patterns noted in the response.

2. Explain what did not go well during the session. Request possible options for the next session that would result in a better learning experience.

3. Use an AI-based time tracker to analyze time spent on different aspects of the study session and consider how the timing might be adjusted next time.

4. Use chat mode to talk through the emotional response to the study session. Prompt AI to ask probing questions about the experience. (e.g., “I am going to tell you about my study session today. When I pause, ask me to elaborate on what I was explaining or ask me how I feel about what I just said.")

The goal for all exercises is that the learner engages in the learning process. AI is used to support that process, providing assessments of effectiveness and suggesting alternatives based on what was presented. It is important to emphasize to your students that although AI can prompt reflection, genuine introspection comes from actively engaging with these prompts. Passively accepting AI’s responses will not result in actual learning and growth. Another consideration is to avoid the outcome trap. Talk to students about the power of the process. Taking a quick quiz as a metacognitive monitoring process is supposed to gauge the extent to which the material is being learned, not just count correct answers.

Learning requires effort. AI should not be used as a crutch to do the cognitive heavy lifting. The brain, like a muscle, operates on a “use it or lose it” principle. Just as jogging builds aerobic capacity because it is physically demanding, engaging in metacognitive processes builds cognitive endurance because it is challenging. It might be tempting to let AI do the work, but that approach does not build the skills needed for long-term success.

Conclusion

As AI becomes more integrated into education, we must use it to augment the learning process rather than race to achieve outcomes. By anchoring intentional, critical AI use in the metacognitive practices of planning, monitoring, and reflecting, educators can help students develop into thoughtful, self-directed learners.

Discussion Questions

1. Which of the three metacognitive strategies (planning, monitoring, reflecting) do you think is best aided by AI.? Which area holds the most concern that students might just have AI do the work and not develop desirable processes?

2. Describe one metacognitive strategy that students might use to enhance metacognition that is not listed in this piece. Be sure to explain why the strategy suggested would be helpful.

3. How might you ensure that AI enhances rather than replaces student reflection? 

References

Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist, 34(10), 906–911.DOI: 10.1037/0003-066X.34.10.906

McGuire, S. Y., & McGuire, S. (2015). Teaching students how to learn: Strategies you can incorporate into any course to improve student metacognition, study skills, and motivation. Stylus Publishing.

Tanner, K. D. (2012). Promoting student metacognition. CBE—Life Sciences Education, 11(2), 113–120. https://doi.org/10.1187/cbe.12-03-0033

Zakrajsek, T. D. (2022). The new science of learning: How to learn in harmony with your brain (3rd ed.). Stylus Publishing.

AI Use Statement

For this blog I used the chat function of ChatGPT. I explained what ideas I had for the article and asked for areas that might be important that I had not noted. I also used ChatGPT a few times to suggest additional AI apps for certain functions, such as creating quiz questions. At one point, after explaining the main areas I anticipated for the piece, I asked ChatGPT to create an outline, which I then almost completely ignored. I list that because although I didn’t use it, the outline provided likely gave me some ideas I did not have before.  

About the Author