Understanding Student Engagement: The Four Modes

Rebecca Winthrop, Director of the Center for Universal Education at Brookings, discussed her work on transforming education and the challenges and opportunities presented by AI in learning. Her expertise is sought by organizations like the UN, the White House, and Fortune 500 companies, as well as parent networks and school districts. Her work focuses on evidence-based strategies that unite families, educators, policymakers, and companies to help children reach their full potential.

Understanding Student Engagement: The Four Modes

Winthrop, along with co-author Jenny Anderson, explored why many children dislike school despite being naturally curious learners. Their research, detailed in the book The Disengaged Teen, identified four modes of student engagement: Passenger, Achiever, Resistor, and Explorer.

Passenger Mode

Students in Passenger mode do the bare minimum to get by. They might enjoy school for social reasons or extracurricular activities but are largely uninterested in academics. Many students in this mode achieve good grades but are bored and unchallenged. A common complaint from these students is that school is "boring" and they don't see the relevance of what they're learning to real life. Approximately 50% of middle and high school students reported regularly spending time in Passenger mode.

Achiever Mode

Achievers are goal-oriented, organized, and ambitious, striving for top grades and recognition. While these are positive traits, the downside is that they often focus solely on outcomes rather than the learning process. They become frustrated when teachers don't provide explicit instructions for achieving an A. Many achievers experience high levels of anxiety and depression due to the pressure to be perfect, leading to a lack of resilience when faced with less-than-perfect results. They are cognitively engaged but tend to be followers rather than proactive learners.

Resistor Mode

Resistors actively avoid learning, often by not completing homework, skipping classes, or being disruptive. While challenging for parents and teachers, these students often possess significant agency and creativity, which they direct towards avoiding school. If the learning environment can be pivoted to excite them, they can transition to Explorer mode more quickly than Achievers.

Explorer Mode

This is the ideal mode for students, though it's not expected to be maintained 24/7. Explorers are genuinely interested in the learning process and take initiative to make their learning more engaging. This can involve asking teachers for specific project topics, collaborating with peers, or independently researching subjects online. They are not solely focused on outcomes and, surprisingly, often achieve the best academic results. A key indicator of a child in Explorer mode is their enthusiastic sharing of what they've learned, demonstrating deep engagement with the subject matter.

Winthrop noted that less than 4% of middle and high school students regularly spend time in Explorer mode within school settings. This is largely attributed to the high-stakes, ranking-and-sorting design of traditional education systems, which prioritize grades and college admissions.

Shifting from Compliance to Partnership

The relationship between schools and families often gets stuck in a "nagging-procrastination loop," particularly with students in Passenger mode. When schools report a child isn't doing homework, parents often nag, which neuroscience shows can shut down the problem-solving part of an adolescent's brain.

To counter this, Winthrop suggests a "learning-to-learn loop" where parents help children plan their homework, execute it, and then reflect on the process. This builds metacognitive skills and independent learning abilities. The goal is not just to complete homework but to teach children how to learn. This approach requires lower stakes, allowing children to experience failure and learn from it without severe consequences.

AI in Education: Risks and Promises

Winthrop led a Brookings task force on AI and education, which produced a report titled A New Direction for Students in an AI World: Prosper, Prepare, Protect. The report's main question was whether the current deployment of generative AI (GenAI) in education is on the right track.

Risks of GenAI

The task force found that the risks of GenAI currently overshadow its benefits, largely because most students experience GenAI through un-scaffolded, open-ended interactions with frontier models like ChatGPT, which are not designed for learning.

1. Cognitive Stunting/Offloading: Students use AI to skip steps in learning, missing opportunities to practice and build essential skills. Teachers report that students who use AI to write essays often cannot recall the content or arguments they supposedly produced.
2. Social-Emotional Development: AI companions, while offering a sense of constant agreement, can hinder a child's ability to take feedback and develop resilience. Studies show that one in three children prefer talking to an AI friend over a human, raising concerns about their capacity for real-world social interaction.
3. Degradation of Trust: AI use erodes trust between students and teachers. Teachers struggle to verify the authenticity of student work, while students may feel their work isn't genuinely graded if they suspect AI is involved. Parents, too, may challenge grades based on AI chatbot assessments. This breakdown of trust undermines the fundamental teaching-learning relationship and could have long-term implications for civic engagement and democratic values.

Promises of GenAI

Despite the risks, Winthrop acknowledges the immense potential of AI to personalize and individualize learning.

1. Personalized Learning: AI embedded in vetted content, such as digital textbooks, can explain concepts in different ways, which is particularly beneficial for multilingual learners or those with learning differences.
2. Student Empowerment: Students in Explorer mode are already leveraging AI tools to create personalized learning experiences, such as converting notes into podcasts or flashcards.
3. Teacher Support: AI can save teachers time, help them differentiate learning materials, and provide tailored support for neurodivergent students. For example, AI can help children with aphasia communicate by developing synthetic voices. AI could also revolutionize assessment by tracking learning journeys across a wide range of skills.

The "Prosper, Prepare, Protect" Framework

The Brookings report's recommendations are categorized into three areas:

1. Prosper: Focuses on shifting the teaching and learning experience to help students thrive in an AI world. This includes:
   • AI-aware pedagogy: Designing assignments that cannot be easily completed by AI chatbots.
   • AI-assisted pedagogy: Using AI to support diverse learning needs and levels within a classroom.
   • AI-resistant pedagogy: Ensuring time for essential human experiences without technology, such as face-to-face conversations, outdoor activities, and embodied learning, which are crucial for development.
2. Prepare: Emphasizes holistic AI literacy, advocating for universal computer science education to help students understand how the online world works and how to leverage it effectively, beyond just prompt engineering.
3. Protect: Addresses the critical need for child safety in AI products, highlighting the problematic track records of some companies in this regard.

The order of these categories—Prosper, Prepare, Protect—was deliberately chosen to emphasize that while protection is necessary, the ultimate goal is to create a better education system where AI strategically helps all children learn and grow.

The Purpose of Education

Winthrop highlighted that the purpose of education is often the most neglected aspect of reform efforts. She draws parallels with the healthcare system, which focuses on treating illness rather than promoting wellness. Similarly, education systems often prioritize sorting and ranking students.

To achieve meaningful transformation, societies need a shared vision for the purpose of education. Initiatives like "portraits of a graduate," where communities define the competencies and characteristics of successful students, are crucial. By tackling these underlying mindsets and shared purposes, other reforms—such as shifting power dynamics, measurement, and resource allocation—can be more effectively implemented.

Ultimately, with AI being a task-completion technology, and much of education revolving around task completion, it's imperative to rethink the fundamental purpose of schools and what we want children to achieve. Supporting educators to be in "Explorer mode" themselves is central to this transformation, enabling them to creatively leverage new tools and guide students effectively.