Making Student Thinking Visible
If there is one idea I keep returning to in conversations about AI and education, it's this: We need better ways to see student thinking.
Over the past few years, I've spent a lot of time talking with faculty about AI, assessment, academic integrity, instructional design, and student learning.
The conversations often sound different on the surface, but they frequently lead back to the same underlying question: How do we know learning actually happened?
For many years, higher education has relied heavily on final products (artifacts) as evidence of learning.
The paper.
The discussion post.
The project.
The presentation.
The exam answer.
And to be clear, those artifacts still matter. The problem isn't that they are useless. The problem is that they are often incomplete. A final product tells us what a student submitted. It doesn't always tell us how they got there.
That distinction matters more today than it did just a few years ago. Because AI has exposed a weakness that may have existed in many assessments all along. A polished final product does not necessarily tell us how much thinking occurred. Students can sometimes arrive at a correct answer without fully understanding the underlying concepts. Students can sometimes complete assignments through memorization, formula-following, or pattern matching. And now, increasingly, AI can help generate products that appear sophisticated, organized, and academically complete.
The challenge is that an artifact is not the same thing as learning. Learning happens inside the process. The reasoning. The decision-making. The evaluation. The connections students make. The revisions. The moments of confusion. The moments of realization.
The problem is that much of that thinking remains invisible. And when thinking is invisible, assessment becomes more difficult. Feedback becomes less precise. Students have fewer opportunities to examine their own understanding.
This idea isn't new. Researchers associated with Harvard's Project Zero have spent decades exploring the concept of Visible Thinking, arguing that learning improves when thinking is externalized and made observable. Their work suggests that thinking should not remain hidden inside a student's head. Instead, educators should create opportunities for students to reveal their reasoning, assumptions, questions, and decision-making processes.
When that happens, instructors gain a clearer picture of student understanding. But perhaps more importantly, students gain a clearer picture of their own understanding.
That second point may be even more important in an AI-rich world. Because visible thinking isn't only about assessment. It's also about learning. One of the most consistent findings in educational research is the importance of metacognition, the ability to think about our own thinking.
Students who monitor their understanding, evaluate their learning strategies, and reflect on their decision-making tend to become more effective learners. Yet many students spend very little time examining how they arrived at an answer.
Instead, they focus on whether the answer was correct. That tendency isn't surprising. Educational systems have often rewarded performance more visibly than process. Students learn to focus on grades, completion, and outcomes. AI can make that tendency even stronger. When powerful tools can summarize information, generate explanations, suggest solutions, and produce polished work in seconds, it becomes easier than ever to focus on the outcome while paying less attention to the thinking that produced it.
But learning doesn't occur because an answer exists. Learning occurs because a learner engages with ideas. That's why I increasingly find myself asking a simple instructional design question: How can I make student thinking more visible?
Sometimes the answer is surprisingly simple.
A reflection question.
A justification prompt.
A process checkpoint.
A request to explain why one solution was selected over another.
A comparison of competing ideas.
A discussion focused on reasoning rather than answers.
These small additions often reveal far more about student understanding than the final artifact alone. In my own technical courses, I see this regularly. A student may successfully solve a problem. But the most valuable learning often appears when I ask them to explain how they approached it. Why did they choose that troubleshooting step? What evidence led them in that direction? What alternatives did they consider? What changed when their first solution didn't work?
Those conversations reveal understanding in ways that a completed answer never could. And they often reveal misunderstandings as well. That is one of the greatest benefits of visible thinking. It helps instructors see where students are struggling. But it also helps students recognize where they are struggling.
Learning becomes easier to support because it becomes easier to see. This is why I increasingly view visible thinking as one of the most important instructional design principles in an AI-rich classroom. Not because it prevents AI use. Not because it creates AI-proof assignments. But because it shifts our focus toward the part of learning that has always mattered most.
The thinking.
The judgment.
The reasoning.
The sense-making.
The growth.
Those things have always been at the center of education. AI didn't change that. If anything, it made it easier to see. And once we begin making thinking visible to instructors, another challenge emerges: How do we help students make their own thinking visible to themselves?
That's where reflection enters the conversation.
Series 1: AI Is Exposing Existing Problems ✓ Completed
Series 2: What We Do About It
Current Post (3 of 8): Making Student Thinking Visible
Next Up: Building Reflection into Learning