Deep vs Surface Learning: How to Study for Understanding

Deep vs surface learning describes two ways students approach the same material. With a surface approach, you aim to reproduce it: memorize what you expect to be asked and repeat it on the exam. With a deep approach, you aim to understand it: what it means, why it's true, and how it connects to what you already know. That network of existing knowledge is what psychologists call a schema.

The distinction comes from research by Ference Marton and Roger Säljö at the University of Gothenburg in the 1970s. The most useful lesson from this research is easy to miss: the approach isn't a personality type. The same student can take a deep approach in one course and a surface approach in another, depending on the task and what the exam rewards.

This guide covers the difference between deep and surface learning, where the idea comes from, why students slide into memorizing, how to tell which approach you're using, and a week-long plan to study for understanding.

What is the difference between deep and surface learning?

The difference is in what you're trying to do when you study, which then shapes how you do it.

Here's the same economics paragraph studied both ways:

"A price ceiling is a legal maximum price. When it is set below the equilibrium price, quantity demanded exceeds quantity supplied, causing a shortage."

• Surface: highlight it, make a flashcard ("What is a price ceiling?" / "A legal maximum price"), and repeat it until it sticks.
• Deep: sketch the supply and demand graph and mark the ceiling. Ask why buyers want more and sellers offer less at the lower price. Think of a real case, like rent control. Then ask what happens if the ceiling is set above the equilibrium price. (Nothing: it doesn't bind.)

The surface student can define a price ceiling. The deep student can answer an exam question about a ceiling on a product they've never seen.

Where the idea comes from: Marton and Säljö

In their 1976 studies, Marton and Säljö asked university students to read an academic article and then interviewed them about what they remembered and how they had gone about reading it. They borrowed the language of "levels of processing" from memory research of the time.

Students fell into two broad groups:

• Surface-level processing: students focused on the text itself, trying to remember facts and details, often aiming at what they expected to be asked.
• Deep-level processing: students focused on what the author meant, the main point and the argument behind it.

Students who read for meaning were the ones who understood the author's main point. Those who read to remember the text often recalled pieces without the point that tied them together.

A second study looked at whether the questions themselves change how students read. Students who were repeatedly asked detailed factual questions about each section drifted toward a surface approach. Trying to force a deep approach with "what is the main point?" questions didn't work as intended: many students simply learned to produce answers to the expected questions without changing how they read. The authors called this "technifying" the task.

Two lessons stand out. First, students adapt their approach to what they think the task demands. Second, you can't produce deep learning just by adding deep-looking questions. The student has to actually intend to understand.

One caution: the 1976 experiments were about reading a single article. The broader framework of "approaches to learning" was built over the following decades by researchers such as Noel Entwistle, Paul Ramsden, and John Biggs, using interviews and questionnaires about how students study across whole courses.

Is deep processing the same as "levels of processing"?

Not quite. Craik and Lockhart's 1972 "levels of processing" idea is a memory theory about single items: in the experiments that followed, words people thought about for their meaning were remembered better than words they judged only by sound or by how they were printed. Marton and Säljö's deep and surface approaches describe a student's intention toward a whole task, such as a chapter or a course. The two point the same way, but if you want the memory research and the classic experiments, see our guide to encoding strategies for memory.

The strategic approach (and why it isn't a bad word)

Later researchers described a third pattern: the strategic, or "achieving," approach. Strategic students organize their effort around the best possible grade. They manage time, study what's weighted most, and pay close attention to what each exam rewards.

Strategic isn't the opposite of deep. A strategic student in a course that rewards understanding will take a deep approach because that's what scores well. In a course that rewards recall, the same student may memorize. John Biggs's revised Study Process Questionnaire, published in 2001 and designed for teachers to use, dropped its achieving scale and measures only deep and surface approaches.

The best combination for most students is deep and strategic: aim to understand, and be organized about where your hours go.

Why students slip into surface learning

Few students choose surface learning on purpose. It usually happens because of the situation:

• The exam rewards recall. If past papers are all definitions and lists, memorizing is a rational response. That's an assessment problem, not a student problem.
• Too much content, too little time. When a syllabus is overloaded, students fall back on memorizing to get through it. Our guide to cognitive load theory explains why overload pushes out understanding.
• Missing background. You can't connect new ideas to knowledge you don't have yet. Fill the gaps first.
• Anxiety and low interest. When you're stressed or bored, reproducing what's on the page feels safer than wrestling with meaning.
• Familiarity feels like understanding. After rereading, the text looks familiar, so you assume you know it. Highlighting has the same problem, as covered in why highlighting doesn't work.

Is surface learning always bad?

No. Some things you just need to know: vocabulary in a new language, drug names, the periodic table, formulas. Memorizing them well is useful, and having them in memory frees up attention for understanding. The problem is using memorization for material you'll be asked to explain, apply, or argue about. Knowledge filed under surface features also tends to stay stuck in one context, as the post on why knowledge fails to transfer shows.

How to tell if you're learning deeply

Run this check on a topic you think you've learned. Do it without notes.

1. Explain it in your own words in two or three sentences, as if to a friend who missed the lecture.
2. Give an example that wasn't in the textbook or slides.
3. Say why, not just what. Why does the price ceiling cause a shortage? Why does the enzyme stop working at high temperature?
4. Link it to something else. Which earlier topic does it depend on, or contradict?
5. Answer a reworded question. Ask a friend or an AI tool to rephrase a practice question and see if you still get it.
6. Spot the error. Can you tell what's wrong with a flawed explanation of the idea?

If you fail items 2 to 6 but pass item 1 only by reciting a memorized definition, you've learned the surface. The fix is not more rereading but the strategies below.

Deep processing strategies that build understanding

These are what a deep approach looks like in practice. Each has its own guide if you want the details.

• Ask why and how. For every key claim, ask why it's true and answer from what you know. This is elaborative interrogation, rated moderately useful in Dunlosky and colleagues' 2013 review of study techniques.
• Explain each step to yourself. When reading a worked solution or a mechanism, say why each step follows from the last. See self-explanation.
• Make your own examples and non-examples. Rent control is a price ceiling. Is a minimum wage one? (No, it's a price floor.) More in using concrete examples.
• Compare similar ideas. Mitosis vs meiosis, mean vs median, a price ceiling vs a price floor. Write how they differ and why.
• Explain it out loud. Teaching an imaginary student exposes the gaps that rereading hides, as in the Feynman technique.
• Test yourself with "why" and new-case questions. Retrieval practice isn't only for facts. In a 2011 study in Science, students who practiced recalling a text did better on comprehension and inference questions a week later than students who built detailed concept maps. Drawing a map looks deep; pulling the ideas out of memory did more.

That last finding is the key. Deep learning isn't about choosing activities that look sophisticated. It's about doing the mental work of making sense of the material and then retrieving it.

A 7-day plan to study for understanding

Use one course and 30 to 45 minutes a day.

If your exam mixes recall and application, keep a set of fact flashcards alongside this plan and review them on a spaced repetition schedule. Deep and surface aren't either/or: know the facts, and understand what they mean.