The Four Levels of Learner Agency: A Research-Backed Path to Raising Self-Directed Kids

Your eight-year-old asks for help with every problem. Your friend's same-age child plans their own coding projects. What separates them?

It's not talent. It's agency.

Learner agency is a child's capacity to intentionally shape, monitor, and adjust their own learning. Not just "doing things alone." The research is clear: agency involves intentionality, self-regulation, motivation, decision-making, and reflection (Code, 2020). Children who develop these capacities don't just perform better in school. They become adults who can navigate uncertainty, solve novel problems, and lead.

At Kubrio, we built our Agency Development Framework on this body of research. Four distinct levels. Each grounded in peer-reviewed science. Each designed so parents can guide the progression based on what they observe at home.

No algorithms deciding for your family. You set your child's level during onboarding. You adjust it anytime.

Why Learner Agency Matters More Than Knowledge

The AI era belongs to high-agency kids.

When AI can retrieve any fact in seconds, the differentiator is not what your child knows. It's whether they can set a goal, plan a path, adjust when things go wrong, and reflect on what they learned. That cluster of skills has a name in the literature: learner agency.

A 2020 study in Computers & Education (Taub et al.) found that students with higher agency in game-based learning environments showed stronger problem-solving behaviors, more positive emotions, and deeper learning outcomes. Agency didn't just correlate with better performance. It caused it.

Self-Determination Theory, the most validated framework in motivation science, identifies three basic psychological needs: autonomy, competence, and relatedness. When these needs are met, children develop intrinsic motivation. When they're thwarted, motivation collapses. (Joussemet, Landry, & Koestner, 2008)

The implication for parents is direct. Your child's ability to own their learning is not fixed. It's built. And it's built through specific, researched conditions you can create at home.

What the Research Actually Says About Developing Agency

Before we get to the four levels, here's what the science tells us about how agency develops in children ages 6-13.

Agency Is Not Independence

This is the most common mistake parents and educators make. Autonomy in the research does not mean "leave the child alone." It means volition: the child endorses their actions as self-chosen. It means self-regulation: planning, monitoring, adjusting. And it means competence-in-context: being able to act effectively with appropriate support.

Soenens and Vansteenkiste (2010) drew a critical distinction between autonomy-as-volition (healthy) and autonomy-as-separation (often harmful when premature). Children need to feel their choices are their own. They do not need to be abandoned to figure everything out alone.

Agency Develops Through Scaffolded Handoff

Developmental research shows a consistent pattern:

Age Range	What Agency Looks Like	What Adults Should Do
6-8	Autonomy emerges through supported choice, routines, and play-based self-regulation	Offer bounded choices, provide rationales, use guided play (Skene et al., 2022; Weisberg et al., 2016)
8-10	Autonomy becomes responsibility and skill: planning, follow-through, emotional regulation	Combine autonomy support with structure (Grolnick & Ryan, 1989)
10-13	Autonomy shifts toward identity, values, and independent decision-making	Acknowledge their perspective, reduce control, increase coaching (Beyers, Soenens, & Vansteenkiste, 2024)

This is not a rigid timeline. Children move at different paces. But the pattern is reliable: support first, choice second, self-regulation third, full ownership fourth.

What Kills Agency

Barber (1996) identified psychological control as the single most damaging parenting behavior for agency development. This includes guilt-tripping, love withdrawal, and manipulation of feelings. It's consistently linked to worse adjustment, lower motivation, and reduced self-regulation.

The takeaway: structure is good. Behavioral limits are good. Controlling a child's thoughts and emotions is harmful.

Kubrio's Four Levels of Learner Agency

Based on this research, we designed four levels that map to how agency actually develops. Each level describes what the child can do, what the parent's role is, and how Kubrio adapts its environment to match.

These are not stages a child "passes through" in a fixed sequence. Some children operate at Level 3 in one skill and Level 1 in another. The framework is per-child and per-context. You know your child best.

Level 1: Scaffolded

Research basis: Guided play theory (Skene et al., 2022; Weisberg et al., 2016); SDT autonomy support + structure (Joussemet et al., 2008; Grolnick & Ryan, 1989)

The core idea: The child builds foundational skills with structured support. They make choices within clear boundaries. The parent provides direction while preserving the child's sense of volition.

What it looks like in your child:

• Asks "What should I do next?" and benefits from clear options
• Needs frequent check-ins and encouragement
• Builds confidence through small, visible wins
• Learns how learning works: what it feels like to try, fail, adjust, and succeed

The parent's role: Active Guide Research on autonomy-supportive parenting (Joussemet et al., 2008) defines five specific behaviors that support agency even at this early stage:

1. Take the child's perspective
2. Offer meaningful choices within boundaries ("Do you want to start with the characters or the setting?")
3. Give rationales for limits ("We're starting with a shorter project so you can finish it and feel great about it")
4. Acknowledge negative feelings without caving ("I get that this part is frustrating. Let's figure out what's tricky about it")
5. Use non-controlling language (invite, not command)

How Kubrio supports Scaffolded learners:

• Activities broken into small, completable steps that build self-efficacy
• Triple-angle AI feedback (Krea, Tek, Brio) calibrated for encouragement and guided reflection
• Parent coaching prompts that model autonomy-supportive language
• Notifications and reminders that keep parents engaged as active guides
• Quick-win project design so kids taste success early and often

Example: Luca, age 7, starts a storytelling activity on Kubrio. He picks between two story themes (bounded choice). Each step has a clear prompt. After uploading his work, Krea sparks a creative idea ("What if your character had a secret?"), Tek stretches his thinking ("What happens if the setting changes?"), and Brio asks a reflective question ("What part did you enjoy making most?"). His parent gets a coaching prompt: "Ask Luca what surprised him about his story."

That's guided play translated into a digital learning environment. The child leads. The structure supports.

---

Level 2: Choice-Driven

Research basis: Self-Determination Theory autonomy support (Soenens & Vansteenkiste, 2010); agentic engagement (Patall et al., 2019); autonomy-supportive teaching (Reeve & Jang, 2006)

The core idea: The child drives their learning through meaningful choices. They select topics, difficulty levels, and project directions. The parent shifts from directing to coaching.

What it looks like in your child:

• Chooses activities based on genuine interest, not just compliance
• Says "I want to try that" and dives in without needing every step laid out
• Treats mistakes as information, not failure
• Asks thoughtful questions about how things work
• Begins expressing preferences about how they want to learn

This is where agentic engagement (Patall et al., 2019) starts to show. The child doesn't just respond to learning environments. They begin to shape them: asking questions, requesting different resources, expressing preferences.

The parent's role: Coach Your job shifts. Instead of providing direction, you provide options. Instead of checking their work, you ask about their process. The key SDT insight here: autonomy support is not the same as permissiveness. You still provide structure. But the structure serves the child's growing capacity to choose.

How Kubrio supports Choice-Driven learners:

• Open-ended project options across 30+ skills
• Choice in difficulty levels and project themes within each activity
• AI feedback that asks "What would you change?" before telling them what to change
• Parent coaching prompts that shift from "encourage" to "ask good questions"
• Skill analytics showing which interests light up your child's curiosity

Example: Amara, age 9, discovers game design activities on Kubrio. She picks her own game concept. She chooses the difficulty level. When she hits a design problem, Tek offers a stretch goal instead of a solution. She tries three different approaches before finding one that works. Her parent sees in the dashboard that she spent 40 minutes on a single mechanic. The coaching prompt says: "Ask Amara what approaches she tried before the one that worked."

---

Level 3: Self-Regulated

Research basis: Agency for Learning framework (Code, 2020); self-regulation and self-directed learning (Gupta et al., 2024); goal-setting and reflection in primary education (Ponomariovienė & Jakavonytė-Staškuvienė, 2025)

The core idea: The child sets their own goals, monitors their progress, and adjusts their approach when something isn't working. This is the heart of what researchers call self-regulated learning: the capacity to plan, execute, evaluate, and revise.

What it looks like in your child:

• Sets personal learning goals ("I want to make a 3-minute film by the end of the month")
• Plans steps independently before asking for help
• Notices when they're stuck and tries different strategies
• Reflects on what worked and what didn't after completing a project
• Tracks their own progress and takes pride in growth over time

Code (2020) describes this as the integration of four capacities: intentionality, motivation, self-efficacy, and self-regulation. The child isn't just doing activities. They're thinking about their thinking. That's metacognition, and it's the strongest predictor of long-term learning success.

The parent's role: Advisor You're no longer coaching day-to-day decisions. You're available for strategic conversations. "How's your project going?" becomes a real question, not a check-up. You help them reflect, not redirect.

Research on goal-setting with primary students (Ponomariovienė & Jakavonytė-Staškuvienė, 2025) shows that children at this level still benefit from coaching around goal quality. Kids tend to set vague goals ("Get better at coding") rather than specific ones ("Build a game with three levels by Friday"). Your role is to help them sharpen the target.

How Kubrio supports Self-Regulated learners:

• Goal-setting tools built into the learning flow
• Progress tracking visible to both child and parent
• AI feedback that prompts self-evaluation ("Before I share my thoughts, what do you think went well?")
• Reflection prompts after each completed project
• Skill portfolio that makes growth tangible and visible over time
• Parent dashboard showing not just completion, but patterns of self-direction

Example: Jonas, age 11, decides he wants to learn data science. He sets a goal on Kubrio: complete five data visualization projects in three weeks. After the second project, he realizes his charts aren't telling clear stories. He goes back to an earlier activity on storytelling, applies what he learns, and his third visualization is stronger. His reflection note says: "I learned that data needs a story, not just numbers." His parent's dashboard shows the self-correction loop.

---

Level 4: High-Agency

Research basis: Agentic engagement and environmental influence (Patall et al., 2019; Reeve & Cheon, 2021); manifestation of agency in learners (Spence et al., 2025); student agency as sustainable competency (Rupnik & Avsec, 2025)

The core idea: The child doesn't just regulate their own learning. They influence their learning environment. They propose project directions, mentor others, create original work, and reflect on their trajectory as learners. This is the full expression of what researchers call high-agency learning.

Spence et al. (2025) found that agency manifests most clearly when learners "overcome and own challenges," transforming obstacles into growth opportunities rather than waiting for someone to remove the obstacle.

What it looks like in your child:

• Sets long-term learning trajectories, not just project goals
• Creates original work that goes beyond any template or prompt
• Helps other kids: teaches, shares feedback, mentors
• Reflects on their own learning process and adjusts their approach across projects
• Sees themselves as a creator and leader, not a student following instructions
• Asks for different resources, proposes new project directions, shapes their environment

This maps directly to what Patall et al. (2019) call agentic engagement: the degree to which learners proactively contribute to the flow of instruction. High-agency learners don't just participate. They co-create the experience.

The parent's role: Thinking Partner You're no longer advising. You're discussing. Your child's learning becomes a conversation between two people who respect each other's perspective. You stay informed through the dashboard. You celebrate their independence. And occasionally, they teach you something.

How Kubrio supports High-Agency learners:

• Open creation tools with minimal constraints
• Mentorship and collaboration features where they can guide younger learners
• Living skill portfolio showcasing their full learning journey
• AI feedback that treats them as a peer creator, not a student
• Leadership opportunities in collaborative projects
• Parent coaching prompts that focus on rich conversation, not oversight

Example: Sofia, age 12, creates a tutorial series teaching other kids her photography techniques. She designs the curriculum herself, decides on the progression, and provides feedback to learners who try her activities. Her skill portfolio shows 18 months of growth across photography, storytelling, and design thinking. Her parent's coaching prompt: "Ask Sofia what she's learning about teaching that surprises her."

---

How This Framework Differs From Other Approaches

It's Family-Driven, Not School-Based

Most autonomy frameworks in education are designed for classrooms with teachers, learning guides, or institutional rubrics. Kubrio's framework is designed for families. The parent is the decision-maker. The AI is the creative partner. The child is the leader.

You know your child across different contexts: at home, with friends, during hard moments and easy ones. No classroom rubric captures that. You set the level. You adjust it. The platform adapts.

It's Grounded in Specific Research, Not General Philosophy

Each level maps to a specific body of peer-reviewed research:

Level	Research Foundation	Key Citations
Scaffolded	Guided play theory, SDT autonomy support + structure	Skene et al., 2022; Grolnick & Ryan, 1989
Choice-Driven	SDT autonomy, agentic engagement	Patall et al., 2019; Soenens & Vansteenkiste, 2010
Self-Regulated	Agency for Learning, self-directed learning, goal-setting	Code, 2020; Gupta et al., 2024
High-Agency	Agentic engagement, agency manifestation	Spence et al., 2025; Reeve & Cheon, 2021

It's Per-Child and Per-Context

A child might be Self-Regulated in game design (a passion area) and Scaffolded in writing (a newer skill). That's normal. The framework doesn't force a single label. It gives you a lens for understanding where your child is in each domain, so you can provide the right support.

How to Identify Your Child's Current Level

Use these behavioral indicators. Look for patterns, not one-off moments.

Scaffolded indicators:

• Asks "Is this right?" or "What should I do next?" frequently
• Hesitates to start without clear instructions
• Gives up quickly when stuck or frustrated
• Looks to you for validation at each step
• Completes tasks but rarely initiates new ones

Choice-Driven indicators:

• Chooses activities based on genuine interest
• Says "I want to try that" without prompting
• Tries different approaches when stuck before asking for help
• Shares work proudly and asks for specific feedback
• Asks "how" and "why" questions about how things work

Self-Regulated indicators:

• Sets personal learning goals without being asked
• Plans steps before diving in
• Notices when something isn't working and adjusts strategy
• Reflects on completed work ("I think next time I'd do X differently")
• Tracks their own progress and takes satisfaction in growth

High-Agency indicators:

• Sets long-term learning trajectories across multiple projects
• Creates original work that goes beyond templates or prompts
• Helps siblings, friends, or peers with their projects
• Reflects on their own learning process across time
• Proposes new project ideas and shapes their learning environment

The Parent's Evolving Role: A Research-Backed Guide

One of the clearest findings in the autonomy literature is that what works changes as the child develops. The parent who provides exactly the right support at Level 1 needs to evolve their approach as the child grows.

Level	Parent Role	Key Behavior	Research Basis
Scaffolded	Active Guide	Offer bounded choices, give rationales, celebrate effort	Joussemet et al., 2008
Choice-Driven	Coach	Ask questions, provide options not directions, acknowledge feelings	Soenens & Vansteenkiste, 2010
Self-Regulated	Advisor	Help sharpen goals, prompt reflection, discuss strategy	Ponomariovienė & Jakavonytė-Staškuvienė, 2025
High-Agency	Thinking Partner	Discuss as peers, celebrate independence, learn from them	Reeve & Cheon, 2021

The hardest transition for most parents is from Guide to Coach. It means saying "What do you think?" instead of "Here's what you should do." The research on autonomy-supportive parenting (Joussemet et al., 2008) shows this shift is learnable. Interventions that trained parents in autonomy-supportive behaviors showed measurable improvements in children's self-regulation and executive function.

Why Most Education Gets Agency Wrong

Traditional schools keep children at Level 1 permanently. Raise your hand. Wait for permission. Follow the worksheet. The environment is structurally designed to suppress agency.

Even well-intentioned progressive approaches often make the opposite mistake: jumping children to Level 4 without building Levels 2 and 3. The result is overwhelmed children who shut down. Pisani and Haw (2023) found that autonomy-supportive, choice-rich tasks require scaffolding to prevent overwhelm. Freedom without structure isn't agency. It's chaos.

The research is consistent: pair choice with structure. Give children the mental work of planning, deciding, and reflecting while adults provide guardrails and rationales. This is exactly what Kubrio's four-level system does.

You Control the Framework

During Kubrio onboarding, you set your child's starting level based on what you observe. This is not a test. It's your informed decision as a parent.

How it works:

• You choose Level 1, 2, 3, or 4 based on your child's current behavior
• Kubrio tailors all activities, AI feedback tone, and coaching prompts to match
• You can change the level anytime in your parent settings
• If patterns suggest a different level might fit better, we'll send a gentle notification. But you always decide.

Why this matters: Jääskelä et al. (2020) developed student agency analytics that combine surveys with learning data to profile learner autonomy. Kubrio does something similar, but with a crucial difference: the parent, not an algorithm, interprets the data and makes the call.

Your parent dashboard shows not just what your child learned, but how their agency is developing. Which skills they're self-directing. Where they still need scaffolding. How their goal-setting and reflection patterns are evolving.

Because in the AI era, how they learn matters more than what they learn.

Frequently Asked Questions

What if my child is at different levels for different skills? That's expected and healthy. A child passionate about game design might be Self-Regulated there but Scaffolded in writing. Kubrio lets you set levels per context, and the AI adapts accordingly.

How long does it take to move between levels? There's no fixed timeline. Some children move from Scaffolded to Choice-Driven in weeks. Others spend months deepening their self-regulation skills at Level 3. The pace depends on the child, the skill, and the quality of support they receive. Research on goal-setting with primary students (Ponomariovienė & Jakavonytė-Staškuvienė, 2025) shows that even young children can develop agency when the environment is intentionally designed for it.

Can my child move backward? Yes, and that's fine. A child tackling a brand-new, challenging skill might benefit from Scaffolded support even if they're High-Agency in familiar areas. Regression in a new context is not failure. It's smart learning.

How is this different from Montessori? Montessori shares the principle that children should lead their learning. The difference is implementation. Kubrio is AI-powered and family-driven. The parent (not a classroom teacher) guides the progression. The AI provides instant, triple-angle feedback. And the digital portfolio tracks growth across 30+ modern skills. Research comparing Montessori and traditional approaches (Rupnik & Avsec, 2025) confirms that student agency is an enabler of sustainable competency, and that structured autonomy support outperforms both rigid control and unconstrained freedom.

What does the research say about using AI in agency development? Taub et al. (2020) studied student agency in game-based learning environments and found that agency positively impacted learning, emotions, and problem-solving behaviors. The key is that AI should expand choices and provide feedback, not make decisions for the child. Kubrio's AI mentors (Krea, Tek, Brio) are designed to prompt thinking, not replace it.

My child is 6. Is Level 1 the right starting point? For most 6-year-olds, yes. The guided play research (Skene et al., 2022) shows that younger children build agency most effectively through supported choice and structured play. But if your 6-year-old already shows Choice-Driven behaviors (picks projects independently, experiments without fear of failure), start at Level 2. You know your child.

What if I disagree with the level Kubrio suggests? You override it. Always. The suggestion is based on behavioral patterns in the platform. Your knowledge of your child across all contexts is more complete than any algorithm. Parent control is a design principle, not a feature.

---

References

1. Barber, B. K. (1996). Parental psychological control: Revisiting a neglected construct. Child Development, 67(6), 3296-3319. https://doi.org/10.2307/1131928

2. Beyers, W., Soenens, B., & Vansteenkiste, M. (2024). Autonomy in adolescence: conceptual, developmental, cross-cultural perspective. Annual Review of Developmental Psychology. https://doi.org/10.1146/annurev-devpsych-010823-023928

3. Code, J. (2020). Agency for Learning: Intention, Motivation, Self-Efficacy and Self-Regulation. Frontiers in Education, 5, 19. https://doi.org/10.3389/feduc.2020.00019

4. Grolnick, W. S., & Ryan, R. M. (1989). Parent styles associated with children's self-regulation and competence in school. Journal of Educational Psychology, 81(2), 143-154. https://doi.org/10.1037/0012-1649.25.2.143

5. Gupta, N., Ali, K., Jiang, D., Fink, T., & Du, X. (2024). Beyond autonomy: unpacking self-regulated and self-directed learning through the lens of learner agency. BMC Medical Education, 24. https://doi.org/10.1186/s12909-024-06476-x

6. Jääskelä, P., Heilala, V., Kärkkäinen, T., & Häkkinen, P. (2020). Student agency analytics: learning analytics as a tool for analysing student agency in higher education. Behaviour & Information Technology, 40, 790-808. https://doi.org/10.1080/0144929x.2020.1725130

7. Joussemet, M., Landry, R., & Koestner, R. (2008). A Self-Determination Theory Perspective on Parenting. Canadian Psychology, 49(3), 194-200. https://doi.org/10.1037/a0012754

8. Patall, E., Pituch, K., Steingut, R., Vasquez, A., Yates, N., & Kennedy, A. (2019). Agency and high school science students' motivation, engagement, and classroom support experiences. Journal of Applied Developmental Psychology. https://doi.org/10.1016/j.appdev.2019.01.004

9. Pisani, S., & Haw, M. (2023). Learner agency in a chemical engineering curriculum: Perceptions and critical thinking. Education for Chemical Engineers. https://doi.org/10.1016/j.ece.2023.06.003

10. Ponomariovienė, J., & Jakavonytė-Staškuvienė, D. (2025). Manifestation of Learner Agency in Primary Education: Goal Setting, Implementation, and Reflection. Behavioral Sciences, 15. https://doi.org/10.3390/bs15081116

11. Reeve, J., & Cheon, S. H. (2021). Autonomy-supportive teaching: malleability, benefits, interventions. Educational Review. https://doi.org/10.1016/j.edurev.2021.100453

12. Reeve, J., & Jang, H. (2006). What teachers say and do to support students' autonomy during a learning activity. Journal of Educational Psychology, 98(1), 209-218. https://doi.org/10.1037/0022-0663.98.1.209

13. Rupnik, D., & Avsec, S. (2025). Student Agency as an Enabler in Cultivating Sustainable Competencies. Education Sciences. https://doi.org/10.3390/educsci15040469

14. Skene, K., O'Farrelly, C., Byrne, E. M., Kirby, N., Stevens, E. C., & Ramchandani, P. G. (2022). Can guidance during play enhance children's learning and development in educational contexts? A systematic review and meta-analysis. Developmental Psychology. https://doi.org/10.1037/dev0001515

15. Soenens, B., & Vansteenkiste, M. (2010). A theoretical upgrade of the concept of parental psychological control: Proposing new insights on the basis of self-determination theory. Clinical Psychology Review, 30(1), 74-94. https://doi.org/10.1016/j.cpr.2009.11.001

16. Spence, A., Carr, D., Kajamaa, A., & Gormley, G. (2025). Overcoming and owning challenges: A qualitative study exploring the manifestation of agency in learners. Medical Education, 59, 972-982. https://doi.org/10.1111/medu.15631

17. Taub, M., Sawyer, R., Smith, A., Rowe, J., Azevedo, R., & Lester, J. (2020). The agency effect: The impact of student agency on learning, emotions, and problem-solving behaviors in a game-based learning environment. Computers & Education, 147, 103781. https://doi.org/10.1016/j.compedu.2019.103781

18. Weisberg, D. S., Hirsh-Pasek, K., Golinkoff, R. M., Kittredge, A. K., & Klahr, D. (2016). Guided Play: Principles and Practices. Current Directions in Psychological Science, 25(3), 177-182. https://doi.org/10.1177/0963721416645512

---

What if learning looked like building?

Kubrio uses quest-based learning: real challenges with AI guidance, not passive videos or worksheets. Explore problem-solving activities and resources for parents.