- DistillED
- Posts
- Poor Proxies for Learning
Poor Proxies for Learning
Design Lessons for Thinking, Not Just Doing [FREE GUIDE]
Jamie Clark
April 18, 2025 • Estimated Reading Time: 5 minutes
⚗️ Hello, and thanks for opening this edition of DistillED!
This time, we’re exploring a powerful idea from Professor Rob Coe that continues to challenge the way we interpret what’s really going on in our classrooms.
What are poor proxies for learning?
About the time of the influential 2014 report What Makes Great Teaching? (Sutton Trust), Professor Rob Coe introduced the idea of poor proxies for learning—those visible classroom behaviours we often assume signal learning: students are on task, busy, well behaved, giving correct answers, or receiving teacher feedback.
These signs are tempting to trust. But here’s the problem: they don’t guarantee that learning has occurred. A student might look engaged or get an answer right in the moment, yet forget it by tomorrow—or never have understood it in the first place.
Why do poor proxies matter?
Poor proxies create the illusion of learning. They cause us to overestimate how much students truly understand and overlook the moments when genuine thinking isn’t happening. As research reminds us, learning is not about what students appear to be doing in the moment—it’s about changes to long-term memory and schema building. Or, as Kirschner, Sweller, and Clark (2006) put it:
"The aim of all instruction is to alter long-term memory. If nothing has changed in long-term memory, nothing has been learned.”
When we rely on misleading indicators, we risk designing lessons that look effective on the surface—but don’t generate real, durable learning.
This is why thinking matters so much. Learning occurs when new information is actively processed in working memory and successfully transferred to long-term memory, where it connects with prior knowledge. Without this cognitive effort—without thinking—there is no real learning. The diagram below illustrates this journey: without active encoding and cognitive effort, information never makes the leap, and true learning doesn’t happen.
Diagram adapted from InnerDrive’s ‘Working Memory vs Long-Term Memory’.
How do I design lessons for thinking, not just doing?
When we design lessons, the question shouldn’t be “What activity should students do?” but “What should students be thinking about?” Learning that sticks requires cognitive effort, not just engagement. As Daniel Willingham puts it:
"Students remember what they think about.”
With Willingham’s words in mind, here are five common poor proxies for learning—and what to do instead:
Poor Proxy | Description | Strategy |
|---|---|---|
Proxy 1 | Students are busy: lots of work is done (especially written work) | Use retrieval practice, elaboration, or self-explanation tasks that require students to recall, connect, and explain ideas, not just complete worksheets. |
Proxy 2 | Students are engaged, interested in learning and are motivated. | Ask: “What are students thinking about?” Design activities that direct attention to core ideas—e.g., concept mapping, “think hard” prompts, or guided questioning. |
Proxy 3 | Students are getting attention: feedback, explanations. | Build in response to feedback—use "actionable next steps," redrafting, or follow-up questions. Ask students to explain what they changed and why. |
Proxy 4 | Classroom is ordered, calm, under control. | Use cold call, whiteboards, or choral response to check for understanding across the room and surface misconceptions in real time. |
Proxy 5 | A few students answering correctly can give a false sense of whole-class learning. | Use whole-class retrieval, hinge questions, or low-stakes quizzes to gather responses from every student—not just the confident few. |
Where can I find out more?
One-Page Guide
Download my FREE high-quality one-page guide: Poor Proxies for Learning: Design Lessons for Thinking, Not Just Doing
|
Report: What Makes Great Teaching?
Read the 2014 report titled ‘What Makes Great Teaching? Review of the Underpinning Research’ by Rob Coe et al (it’s still as relevant and powerful as ever).
Until next time — stay curious, stay clear!
Jamie
P.S. If this issue sparked something useful, feel free to forward it to a colleague