Designing Educational AI ChatBots for Metacognition & Self-Regulated Learning

The integration of AI chatbots in education is reshaping traditional learning paradigms, offering new opportunities to enhance students' ability to engage in self-regulated learning (SRL). Rooted in the principles of Zimmerman’s SRL framework and Judgment of Learning (JOL), AI technologies like ChatGPT present educators with a dual challenge: leveraging these tools for pedagogical benefit while mitigating potential misuse. The research article "Educational Design Principles of Using AI Chatbot That Supports Self-Regulated Learning in Education: Goal Setting, Feedback, and Personalization" by Daniel H. Chang (2023) and colleagues provides an innovative exploration of how AI chatbots can be used to promote SRL through targeted educational design principles.

Central to the discussion is the need for AI chatbots to actively support students in goal setting, self-assessment, and personalisation—key tenets of SRL. Goal setting, enabled by teaching prompting skills, empowers students to articulate their learning objectives effectively, while self-assessment through feedback mechanisms fosters critical reflection on their progress. Personalisation, driven by learning analytics, tailors the educational experience to individual needs, encouraging deeper engagement and autonomy. These principles not only align with contemporary educational trends but also offer actionable guidelines for integrating AI into classrooms to cultivate self-directed learners.

The article emphasises the importance of collaboration between educators, AI developers, and instructional designers in creating tools that support sustainable and ethical learning practices. By moving beyond concerns of academic dishonesty and focusing on pedagogical innovation, educators can harness the potential of AI to transform learning environments. This approach underscores the broader goal of using emerging technologies to prepare students for a future where adaptability, self-regulation, and technological fluency are essential.

This article concludes by examining the role of AI chatbots in fostering metacognition, with a particular focus on helping students develop metacognitive knowledge. Metacognitive knowledge—the awareness of one’s own learning processes, strengths, and strategies—forms the foundation for effective self-regulation and lifelong learning. By guiding students through reflective practices and prompting them to think critically about how they learn, AI chatbots can facilitate a deeper understanding of themselves as learners. This final section explores how these tools can be leveraged to build metacognitive knowledge, ensuring students are equipped to navigate their educational journeys with greater confidence and insight.

How Might an AI ChatBot Be Used To Help With Self-Regulated Learning?

The authors propose that AI chatbots can assist in self-regulated learning (SRL) by aligning their capabilities with educational principles that promote autonomy, reflection, and personalised learning strategies. Specifically, the integration of chatbots is centred on three pedagogical pillars:

1. Goal Setting and Prompting: Chatbots can support students in setting clear and actionable learning goals, a fundamental aspect of SRL. Students interact with chatbots by formulating prompts, which can be viewed as explicit goals for their tasks. These interactions help students articulate objectives, develop strategic approaches, and sustain focus. Educators can also guide students in crafting effective prompts that encourage thoughtful engagement with the chatbot rather than simple task completion.

   
   

2. Feedback and Self-Assessment: Chatbots are envisioned as tools for facilitating continuous feedback and self-assessment. By integrating reverse prompting—where the chatbot poses reflective questions or encourages deeper analysis—students can evaluate their understanding and identify knowledge gaps. This process not only enhances metacognitive skills but also aligns with the Judgment of Learning (JOL) concept, enabling learners to critically assess their progress and adjust strategies accordingly.
   

3. Personalisation and Adaptation: AI chatbots can use learning analytics to tailor feedback and recommendations to individual learners. By analysing interactions and performance data, chatbots can provide customised guidance, helping students reflect on their learning patterns and adapt their approaches. This personalised feedback fosters deeper engagement and supports the development of self-regulatory behaviours.

   
   

Together, these principles aim to reposition chatbots from mere information providers to active facilitators of learning, enhancing students' ability to independently manage their educational journeys while maintaining academic integrity.

---

Additional Ideas from The Global Metacognition Institute Team

We wanted to put some new ideas out there in order to further the possibilities discussed by Chang et al (2023) here are some of our ideas as to how an AI ChatBot might, in the future, assist students in their self-regulated learning processes:

1. Guided Learning Pathways: An AI chatbot could offer dynamically generated learning paths tailored to a student’s pace and preferences. Based on initial assessments or ongoing performance, the chatbot might suggest specific modules, exercises, or resources that align with the learner’s goals, while adjusting recommendations in real time based on their progress. These pathways could also include “stretch goals” to encourage students to challenge themselves.
   

2. Time Management Coaching: Chatbots could act as time management assistants by helping students create and adhere to study schedules. Through regular check-ins, reminders, and adaptive updates, the chatbot might provide insights into how the student is allocating their time, suggesting ways to improve focus or manage overlapping deadlines effectively.
   

3. Scenario-Based Learning Simulations: A chatbot could simulate real-world problem-solving scenarios relevant to the student’s subject area. These interactive scenarios would encourage students to apply their knowledge, experiment with different approaches, and reflect on their decisions, thereby fostering critical thinking and self-regulation.
   

4. Emotional Regulation Support: Incorporating emotional recognition, a chatbot might identify when students express frustration, fatigue, or overconfidence during interactions. It could then provide calming strategies, motivational feedback, or suggest breaks, helping students maintain a productive emotional state conducive to learning.
   

5. Gamification of SRL Processes: To make SRL more engaging, the chatbot could gamify tasks such as goal setting, progress monitoring, and self-reflection. By awarding points, badges, or unlocking new features for successfully managing these tasks, the chatbot could motivate students to take ownership of their learning in a fun and interactive way.
   

6. Community and Peer Support Integration: The chatbot could connect students with peers working on similar topics or challenges, fostering collaborative learning. Through moderated group chats or forums facilitated by the chatbot, students could share insights, discuss strategies, and reflect collectively, enriching their self-regulation process.
   

7. Adaptive Mindset Training: The chatbot might provide micro-lessons or exercises to cultivate a growth mindset, encouraging students to view challenges as opportunities to learn. This could include daily affirmations, problem-solving tips, or stories of successful learning journeys.
   

8. Cognitive Load Balancing: By analysing the complexity and volume of tasks the student is working on, the chatbot could suggest splitting tasks into manageable chunks or recommend alternating between easier and more difficult topics to optimise cognitive engagement.
   

9. Reflection Journal Integration: The chatbot could serve as a digital journal, prompting students to log their reflections on daily learning activities. It might ask targeted questions, such as, “What did you find challenging today?” or “What strategies worked well?” These reflections could later be summarised to show growth over time.
   

10. Ethical Decision-Making Scenarios: For subjects involving ethics or decision-making, the chatbot might present dilemmas or questions where students must weigh options, justify their reasoning, and evaluate the outcomes. This encourages both metacognitive awareness and a deeper engagement with the material.
    

11. Clarifying Motivations and Aligning Content: Drawing from the Motivational Affective Model of Self-Regulated Learning (MASRL), an AI chatbot could engage students in conversations to identify their intrinsic and extrinsic motivations for learning. For instance, it could ask questions such as, “What excites you about this subject?” or “What is your goal for mastering this material?” Based on the student’s responses, the chatbot could tailor the learning content to align more closely with their motivations. For example, a student motivated by career aspirations might receive real-world applications of the topic, while a student driven by curiosity could be provided with exploratory or advanced material.

    
    

12. Addressing Emotional Issues and Supporting Regulation: Building on the affective aspect of MASRL, the chatbot could use sentiment analysis to detect signs of stress, frustration, or disengagement during interactions. For example, if a student expresses phrases like “This is too hard” or “I can’t do this,” the chatbot could acknowledge their feelings and provide emotional support. It might offer encouraging messages, suggest a brief mindfulness exercise, or recommend strategies for tackling the challenging task. By helping students recognise and manage their emotions, the chatbot fosters a balanced emotional state that supports sustained learning and effective self-regulation.
    

These approaches provide fresh avenues for chatbots to enhance self-regulated learning, fostering independence, engagement, and adaptability in learners.

AI Assistance Focussed On The Self-Regulated Learning Cycle

The self-regulated learning (SRL) cycle is a dynamic process in which learners actively manage their cognitive, emotional, and behavioural engagement with a task. This cycle consists of four interrelated phases: planning, monitoring, evaluating, and regulating. During planning, students set specific goals and devise strategies to achieve them. In the monitoring phase, they track their progress and assess their understanding. The evaluating phase involves reflecting on outcomes to determine the effectiveness of their strategies and the quality of their work. Finally, in the regulating phase, learners make adjustments to improve their approach, whether by refining their strategies, seeking help, or recalibrating their goals. By iteratively engaging in this cycle, students develop metacognitive skills, maintain motivation, and take greater ownership of their learning.

AI chatbots, with their ability to provide personalised guidance and real-time feedback, are uniquely positioned to support each phase of the SRL cycle. Below, we explore how chatbots can facilitate planning, monitoring, evaluating, and regulating, with examples of specific prompt questions that can help students engage deeply with each phase.

Planning

AI chatbots can play a pivotal role in helping students with the planning phase of the SRL cycle. They can guide students to define clear, actionable learning objectives and identify effective strategies for achieving them. For instance, a chatbot might prompt students to break down a large assignment into manageable steps or suggest tools and resources aligned with the student’s learning style. By encouraging students to articulate their goals and consider the steps required to meet them, the chatbot fosters a structured approach to learning.

The chatbot could also personalise planning advice based on past interactions or the student’s performance data. For example, if a student has struggled with time management in previous tasks, the chatbot might recommend using a study planner or setting interim deadlines. By tailoring guidance to the learner’s needs, the chatbot helps students establish a realistic and effective plan of action.

Example Prompts for Planning:

• "What is your main goal for this study session? Can you break it into smaller tasks?"

• "What resources or tools will you need to achieve your goal?"

• "How much time will you allocate to each task? Let me help you create a timeline."

• "What strategies have worked well for you in the past when tackling similar tasks?"

Monitoring

During the monitoring phase, the chatbot can assist students in tracking their progress and evaluating their understanding in real time. By asking reflective questions, the chatbot encourages students to pause and assess whether they are on track to meet their goals. For instance, the chatbot might prompt the student to summarise what they have learned so far or identify areas where they feel uncertain. These prompts help students maintain focus and detect potential obstacles early.

The chatbot can also provide direct feedback or suggest self-check exercises. For example, after completing a section of study, the chatbot might offer a short quiz or checklist to help the student evaluate their grasp of key concepts. This ongoing interaction keeps students engaged and ensures they remain aware of their progress throughout the task.

Example Prompts for Monitoring:

• "How confident do you feel about the material you just reviewed? Can you summarise it in your own words?"

• "Are you sticking to your planned timeline? Do you need to make adjustments?"

• "What part of the task are you finding most challenging right now?"

• "Would you like me to provide a quick self-check quiz to see how you’re doing?"

Evaluating

The evaluating phase involves reflecting on the outcomes of a learning task to determine whether the goals were met and the strategies used were effective. AI chatbots can facilitate this reflection by prompting students to review their work and consider what went well or poorly. For instance, after completing an assignment, the chatbot might ask the student to rate their satisfaction with the result or identify areas where they excelled and struggled.

By helping students analyse the outcomes of their efforts, the chatbot fosters critical thinking and supports the development of metacognitive awareness. Additionally, it can provide feedback based on predefined criteria or ask questions that guide the student in evaluating the quality of their work independently.

Example Prompts for Evaluating:

• "How satisfied are you with the outcome of your work? What do you think you did well?"

• "What challenges did you face, and how did you address them?"

• "If you were to do this task again, what would you do differently?"

• "Do you think the strategies you used were effective? Why or why not?"

Regulating

In the regulating phase, students make adjustments based on their evaluation to optimise future learning efforts. The chatbot can assist by suggesting alternative strategies, recommending additional resources, or helping students set new goals. For example, if a student identifies a specific weakness, the chatbot might provide exercises or tutorials to address it. If a strategy proved ineffective, the chatbot could suggest exploring a different approach.

This phase also involves addressing motivational or emotional barriers. A chatbot can offer encouragement, recommend breaks, or provide stress-management tips to help students stay on track. By supporting students in refining their learning processes, the chatbot ensures that each cycle of learning becomes more effective and efficient.

Example Prompts for Regulating:

• "What will you do differently next time to improve your results?"

• "Is there a specific skill or topic you’d like to focus on improving?"

• "Would you like suggestions for resources to help with areas you found challenging?"

• "How are you feeling about your progress? Do you need strategies to stay motivated?"

By actively supporting each phase of the self-regulated learning cycle, AI chatbots can empower students to take greater control over their learning journey, fostering independence, metacognition, and resilience.

---

Fostering Metacognition with AI ChatBots: Building Metacognitive Knowledge

Metacognition, often described as “thinking about thinking,” refers to the processes through which individuals understand and regulate their own cognitive activities. It is a cornerstone of effective learning, encompassing two primary components: metacognitive awareness and metacognitive knowledge. Metacognitive awareness involves recognising when and how to use cognitive strategies, while metacognitive knowledge is a deeper, more structured understanding of oneself as a learner, including awareness of personal strengths, weaknesses, and preferences, as well as knowledge of tasks and strategies. While awareness is a dynamic, real-time process, metacognitive knowledge provides a stable foundation for long-term growth and adaptability in learning.

AI chatbots can play a significant role in helping students cultivate metacognitive knowledge by guiding them through reflective exercises that encourage self-awareness and deeper insights into their learning processes. Through well-designed reflection questions, a chatbot can prompt students to consider critical aspects of their learning experiences, such as the strategies they employ, the conditions under which they learn best, and how their emotions influence their academic performance. These reflections help students develop a more nuanced understanding of themselves as learners, fostering growth not only in academic contexts but in lifelong learning.

By encouraging structured reflection, chatbots help students uncover patterns and preferences they may not have previously recognised. For instance, a chatbot might ask a student to reflect on why a particular strategy worked well or why they felt more focused during certain study sessions. This process nurtures self-awareness, helping students align their learning behaviours with their goals and adapt to challenges more effectively. Through such interactions, AI chatbots act as facilitators of metacognitive growth, equipping students with the tools to understand themselves better and become more self-regulated, independent learners.

Illustration Dialogue

The following dialogue illustrates the potential applications of this approach. It is a hhypothetical dialogue between a nine-yearold boy and an AI Chatbot that has been designed to bring a student towards a deeper understanding of themselves as a learner and how they can maximise their learning power.

Chatbot: Hi there! I’m here to help you learn better and understand what works best for you. Can I ask you a few questions about how you like to learn?

Student: Sure!

Chatbot: Great! Let’s start with this: when you’re doing homework or studying, what kind of place helps you focus best? Is it somewhere quiet, or do you like a little background noise?

Student: I think I like it quiet. I get distracted if there’s too much noise.

Chatbot: That’s helpful to know! Now, when you’re learning something new, do you like to read about it, watch videos, or have someone explain it to you?

Student: I like watching videos. They make it easier to understand.

Chatbot: Got it! And when you’re working on a tricky problem or question, do you like to figure it out on your own first, or do you like asking for help right away?

Student: I try it on my own for a bit, but if it’s too hard, I ask for help.

Chatbot: That’s a smart way to handle tough problems. Now, think about the last time you did really well on a school project or test. What do you think helped you do so well?

Student: Umm, I practised a lot and asked my teacher questions when I didn’t get it.

Chatbot: That’s excellent! Practising and asking questions are both great strategies. One more question: when you don’t feel like studying or doing homework, what helps you get started?

Student: Sometimes, I tell myself I’ll do a little bit, and then it’s not so bad once I start.

Chatbot: That’s a clever trick! Doing a little bit at a time makes big tasks feel less scary.

Chatbot: Thanks for answering all my questions! Based on what you told me, here’s what I’ve learned about you as a learner:

• You focus best in a quiet space, so try to find a quiet spot for studying or homework.

• You like learning through videos, so watching tutorials or educational videos might help when you’re stuck.

• You’re good at balancing trying things on your own and knowing when to ask for help.

• You work hard and practise a lot when you want to do well.

• Starting with small steps helps you get going when you’re not feeling motivated.
  

Chatbot: To maximise your learning power, here are some tips for you:

1. Find or create a quiet place for study time.

2. Use videos to explore new topics and help with tricky subjects.

3. Keep practising and don’t be afraid to ask for help when you need it—teachers and adults love helping you learn.

4. When a task feels big, break it into smaller steps, just like you already do!
   

Chatbot: You’re already doing lots of things right! Keep up the great work, and I’m here to help if you need ideas or want to try something new. What do you think about these tips?

Student: They sound good! I’ll try to do more of those things.

Chatbot: Awesome! I believe in you—you’re going to be a great learner. Let’s chat again soon and see how it’s going!

---

AI ChatBots for Metacognition & Self-Regulated Learning: Final Thoughts

The integration of AI chatbots into education is no longer a matter of if but when. As these technologies become increasingly accessible and sophisticated, their presence in classrooms and learning environments is inevitable. AI chatbots offer a unique opportunity to transform traditional educational models by providing personalised, adaptive support that empowers students to take greater control of their learning. However, for this potential to be fully realised, educators and developers must move beyond using chatbots as mere information dispensers. Instead, these tools should be designed with a clear focus on fostering metacognitive and self-regulated learning (SRL) skills, helping students develop the critical thinking, self-awareness, and adaptability necessary for success in a rapidly changing world.

Incorporating metacognitive and SRL pedagogy into chatbot functionality ensures that these tools become not just facilitators of knowledge acquisition, but partners in the learning journey. By guiding students through reflective practices, goal setting, progress monitoring, and strategy refinement, chatbots can nurture lifelong learning habits and a deeper understanding of oneself as a learner. This alignment between cutting-edge technology and proven educational frameworks bridges the gap between innovation and pedagogy. As education continues to evolve alongside advancements in AI, the thoughtful integration of metacognitive and SRL principles will ensure that these tools not only enhance learning outcomes but also contribute meaningfully to the development of independent, reflective, and resilient learners.

---

References

Chang, D. H., Lin, M. P. -C., Hajian, S., & Wang, Q. Q. (2023). Educational Design Principles of Using AI Chatbot That Supports Self-Regulated Learning in Education: Goal Setting, Feedback, and Personalization. Sustainability, 15(17), 12921. https://doi.org/10.3390/su151712921