Why Can’t Teachers Just Tell Students the Answer?

This may be the most important question in the science of learning:

How do you help a student learn something without telling them the answer?

At first glance, the question seems absurd. If you want students to know something, why not simply tell them? Why make learning harder than it needs to be?

In fact, that’s exactly how much of schooling works. Teachers explain. Students listen. Then students repeat the information on a test. Schools are rewarded when test scores rise, teachers are rewarded when students get the right answers, and students are relieved when someone finally tells them what they need to know.

Everybody wins.

Except learning.

The uncomfortable truth is that people learn remarkably little from being told things. Real learning happens when learners grapple with ideas, make connections, test explanations, and gradually construct understanding for themselves. The process is slower, messier, and more frustrating than simply being given the answer. But it works.

When the class is easy, students often don’t learn very much.

Among the scientists who study learning, there’s no disagreement. We don’t have different factions arguing about this. The science is very clear. In my podcast The Science of Learning, you won’t find a debate episode, or a point-counterpoint episode, with one scientist advocating “let’s tell the student the answers.” That’s because there aren’t any such scientists.

We all agree about what we need to do. One of the most important researchers to show us the way is Dr. Joseph Krajcik of Michigan State University, my guest on this episode of The Science of Learning:

Research shows that effective learning is effortful. If it’s easy, then you’re not learning. But no one likes to work harder than they have to, and thinking is hard work. I know from my own college teaching experience that when students have to work harder for their grades, your end of semester course evaluations are lower. I’m a learning scientist and I know the research and I practice what I preach. I give assignments that make my students think. I don’t lecture in class; I give hands-on activities and my students stay busy. I don’t simply tell them the answer. My anonymous end-of-semester written comments say things like “This course was harder than my other courses.” My course isn’t hard because I assign a lot of busy work; I don’t. It’s hard because the work that I assign requires thinking and effort, and even more, because there’s always uncertainty and ambiguity. When you don’t know what the answer is, then by definition, you’re uncertain of what the answer is. And uncertainty is frustrating.

When you don’t know the answer, you face ambiguity and frustration. When you have to seek and find the answer for yourself, it requires effort, thought, and time.

When someone tells you the answer, you often get an illusion of learning, but you don’t truly learn it. You might remember it just long enough to get it right on a test, but you’ll forget it not long after the test. You won’t understand what you’ve learned. You won’t be able to use it in the real world. You won’t be able to connect it to anything else you’ve learned in the past, or anything else that you’re learning the same semester. We heard about this in episode 4 of the podcast, in my interview with Dr. Henry Roediger.

For Dr. Krajcik, it starts with this simple fact: people learn more when they figure something out for themselves than when someone tells them the answer. Maybe this sounds like common sense, like something you’ve heard people say since you were twelve. Of course we remember discoveries we make on our own. But if it’s common sense, then why is it that so much of modern schooling still operates according to the opposite assumption? Students are given information, expected to absorb it, and then tested on how much they can recall. The fancy term for it is the “transmission-and-acquisition” model of teaching and learning. The simple term is “the football” model: you pass, they catch.

Krajcik has spent much of his career challenging that view. He’s one of the leading figures in project-based learning and science education, and his research shows that for deep learning, the teacher’s role isn’t to deliver the answers; it’s to create the conditions in which students can gradually build knowledge for themselves.

A good teacher creates the conditions where students can build their own knowledge.

It’s the holy grail of the learning sciences: How do you help a student learn something without telling them the answer? You create the conditions so that they can learn it themselves. It’s not easy to do it and it’s not easy to describe it. But Krajcik knows how. He’s been doing research on it, and developing curricular units, for decades. Watch this episode to learn a lot about how to do it.

A revealing moment came early in our conversation when Krajcik described the teacher as a cognitive partner. This term avoids two common misconceptions. On one hand, it rejects the image of the teacher as a lecturer whose primary task is to tell students what they need to know. On the other hand, it rejects the caricature of “discovery learning” in which students are just turned loose and expected to learn everything on their own. In Krajcik’s view, neither extreme works very well. Active learning requires parameters and constraints, a fairly structured environment. Effective teachers design structured environments that help learners make connections. Then, they give learners the tools they need to build their own knowledge. I teach this way myself, and the way I do it, it’s a lot more work than lecturing.

The art of teaching lies in providing just enough support to make progress possible while gradually removing that support as competence develops. Educational psychologists often refer to this process as scaffolding. The teacher’s task is to design those situations and provide the necessary support for students to learn from them. Students learn from their experiences in these situations, not from the teacher talking to them and giving them the answers.

Good teachers help students reorganize their thinking.

In deep learning, students connect new ideas to prior experiences. We learn most effectively when abstract concepts are attached to something we’ve seen, done, or wondered about. This insight is one of the foundational ideas of the learning sciences. Knowledge is not a collection of isolated facts stored somewhere in memory. It’s a network of conceptual connections that grows over time. The stronger those connections become, the more easily we can use what we know in new situations.

When students learn deeply, they develop models that connect facts and knowledge, rather than simply memorizing facts and knowledge. When you hear the word “model” you might remember the balls you used to connect together with sticks to make a molecule in your chemistry or biology class. Those were models. We’ve all seen the model of the solar system with the planets rotating around the sun. A model is a simple representation of something in the physical world. A diagram or a sketch is also a model. But scientific models can be much deeper than a 2D or 3D visual representation. They’re not always simply pictures of reality. Models are tools for explanation and prediction. They help us understand why something happens and they help us anticipate what might happen next. Most students encounter models as a fixed thing that exists, like the solar system. That’s something unchanging, something you memorize for the test. But in scientific practice, models are tentative hypotheses, and they can change. In March 2020, how could we predict how Covid-19 was going to flow around the world? Scientists built models to predict how the infections would move, and they didn’t know which models were better or worse. Krajcik wants students to learn how to build models themselves. The process of constructing a model, drawing on the data the students are gathering in class, forces them to organize evidence, connect ideas, and test whether their explanations actually work.

Building a model is practice for building knowledge.

As I write this, in June 2026, I’m at the annual conference of the International Society of the Learning Sciences (ISLS) at the University of California, Irvine. Researchers from around the world are here to share new findings about how people learn in schools, workplaces, museums, online communities, and other settings. Joe Krajcik is well-known in the learning sciences; he wrote one of the best-known chapters in The Cambridge Handbook of the Learning Sciences, “Project Based Learning.” Many of the ideas Krajcik discussed in our podcast conversation are central themes within the field:

• The importance of prior knowledge

• The role of scaffolding

• Learning through inquiry

• Building models

• Developing questions

These ideas have shaped decades of research and continue to influence educational practice around the world. Attending this conference has reminded me how much the field has grown since its beginnings and how much excitement there is about bringing research-based ideas into classrooms.

This conference is when I officially launched my new podcast, The Science of Learning. If you’re a subscriber to this newsletter, you’ve already been reading about the first six episodes. My goal is to bring the insights of learning scientists to a broader audience of educators, students, policymakers, and lifelong learners. Joe Krajcik is exactly the kind of scholar I hoped to feature when I began planning this project. His work demonstrates that learning is not about the passive reception of information. It’s about helping people build knowledge, make connections, and figure things out for themselves.

Below, I show how Krajcik uses project based learning to teach students curiosity, inquiry, and how to ask good questions.