Self-Explanation in Math and Physics: Learn from Worked Examples

If worked solutions usually feel clear while you are reading them, but new problems still stop you when you try them yourself, self-explanation is often the missing piece. It turns a worked example from something you can recognize into something you can reuse.

This guide draws on Vegard Gjerde’s university physics teaching and research on how students learn from worked examples in physics and math.

---

On this page: What Is Self-Explanation? · Why It Works · Strong vs. Weak · How to Self-Explain · Full Physics Walkthrough · Short Math Example · Practical Tips · When to Self-Explain · Common Mistakes · FAQ

---

What Is Self-Explanation?

Self-explanation is how you study a worked example when the goal is to learn the reasoning, not just finish the page. Instead of narrating the visible actions, you make the hidden structure explicit.

• For physics or model steps, explain four things: the principle, the conditions, the setup or description, and the goal.
• For mathematical procedure steps, explain three things: the condition that makes the move valid, the action you perform, and the goal of that action.
• Use it on important non-mastered steps. You do not need equal time on trivial steps you can already justify cold.
• Use it whenever you study a worked solution seriously. It becomes especially valuable after you get stuck on a problem, check the solution, and need to turn that solution into something you can use again.

---

Why Self-Explanation Works

When you self-explain, you:

• Expose gaps – You spot where your understanding is incomplete.
• Anchor principles – You link each step to the conditions and concepts that justify it.
• Make the setup visible – You stop skipping past system choice, sign conventions, diagram meaning, and term interpretation.
• Build retrievable solution rules – You organize what you need to solve new problems later: which principle is used, when it applies, how it is set up, and what the step is trying to achieve.
• Strengthen transfer – You learn to see past surface details and map problems to the underlying principles.

Self-explanation builds understanding, but it is not the whole study loop. After you understand a worked solution better, you still need problem solving to convert that retrievable understanding into skill and automaticity.

Research note: This guide leans especially on Vegard Gjerde’s work on effective self-explanations in introductory physics, especially “Problem solving in basic physics: Effective self-explanations based on four elements with support from retrieval practice” and “Integrating effective learning strategies in basic physics lectures: A thematic analysis”.

---

Strong vs. Weak Self-Explanation

Weak self-explanation just narrates actions:

First, set this equal to that. Then plug in the numbers.

Strong self-explanation reveals the logic:

We apply conservation of mechanical energy because only gravity does work on the object here. The height term represents lost gravitational potential energy, the kinetic term represents gained speed, and this step sets up the speed we need before the collision model.

The difference?
Weak = descriptions of procedures without purpose.
Strong = principle, conditions, setup, and goal made explicit.

---

How to Self-Explain a Worked Example

Worked solutions in physics and math usually contain two layers: model steps and procedure steps. They need slightly different kinds of explanation.

For physics or model steps

Your self-explanation should usually cover four elements:

• Principle – What principle, definition, or supporting relation is being used to build the model?
• Conditions – Why does it apply in this situation?
• Setup or description – How is the principle set up on the page: which system, states, sign choices, omitted terms, coordinate directions, or geometric relations are being used?
• Goal – How does this step move you toward the target?

This is where many students stop too early. They name the main principle, but they do not explain the setup choices or the supporting relations that make the equation mean something in this specific problem.

For mathematical procedure steps

For transformations and algebraic procedures, use the Condition–Action–Goal lens:

• Condition – Why this operation is valid.
• Action – What you did.
• Goal – Why you did it (“to isolate acceleration”).

You do not need polished prose. Self-explanation is a process, not a performance. The point is to make the reasoning explicit enough that you can see what you understand and what you do not.

---

Full Physics Walkthrough

Use this as the longer deep-dive example. The worked solution stays structural on purpose: verbal decoding, visual decoding, model, procedures, then explanation. If the worked solution does too much of the explanatory work for you, it can suppress the self-explanation you need to do yourself.

Problem

A $65\,\text{kg}$ ninja in an obstacle-course competition runs at $4.0\,\text{m/s}$ and then slides down a $2.0\,\text{m}$ frictionless ramp. At the bottom of the ramp, the ninja collides with a $35\,\text{kg}$ sack hanging straight down from a $5.0\,\text{m}$ rope. By holding on to the sack, the ninja swings upward. Find the rope tension immediately after the collision.

Step 1: Verbal Decoding

Target: $T$

Given: $m_n, m_s, v_0, h, r$

Helpful derived quantity: $m_{\text{tot}} = m_n + m_s$

Step 2: Visual Decoding

Draw the ramp, the ninja at the top with speed $v_0$, the bottom-of-ramp speed $v_1$, the hanging sack, and the swing immediately after the collision with speed $v_2$. Mark the height drop $h$, the rope length $r$, and the upward radial direction at the bottom of the swing. The point of this drawing is to separate the three phases: ramp, collision, and swing.

Step 3: Physics Model

Step 4: Mathematical Procedures

Notice that the numbers are inserted only after the symbolic expression for $T$ is found. Keeping symbols until the end preserves the physics of the expression; substituting numbers too early turns the work into arithmetic and makes the final result harder to interpret.

Step 5: Reflection

• Dimensional analysis: In 4.7, the term $m_{\text{tot}}g$ has force units. The factor $m_n^2/m_{\text{tot}}$ has units of mass, and $\left(v_0^2 + 2gh\right)/r$ has units of acceleration, so the second term also has units of newtons. The left and right sides therefore match as force.
• Size evaluation: The combined weight is about $100 \cdot 9.8 = 980\,\text{N}$, so the tension should be larger than that because the rope must also provide centripetal acceleration. A reported result of $1.4 \times 10^3\,\text{N}$ is plausible.
• Physics evaluation: The final expression says tension increases with total weight and with the centripetal requirement. Greater incoming speed, greater height drop, or a shorter rope increase the tension, which matches the physics.

How to self-explain this model

This is what you should try to explain while studying the solution. In this classical mechanics problem, the main physics learning is in the model. The algebra matters, but it is mostly routine transformation after the model is correct. If algebra is your bottleneck, self-explain the procedure steps too; the short math example below shows what that looks like.

3.1 Energy on the ramp

$\frac{1}{2}m_n v_0^2 + m_n g h = \frac{1}{2}m_n v_1^2$

• Primary principle: Conservation of Mechanical Energy
• Supporting relations: Translational Kinetic Energy, Gravitational Potential Energy (Near Surface)
• Conditions: Friction is neglected on the ramp, air drag is neglected, and the height drop is small enough that $g$ is effectively constant.
• Setup: The system is the ninja on the ramp. The initial state is the run before the drop; the final state is the bottom of the ramp. The left side contains initial kinetic energy plus gravitational potential energy, and the right side contains the later kinetic energy.
• Relevance: This is the cleanest way to get the speed right before the collision.
• Goal: Solve for $v_1$ so the collision model has the correct incoming speed.

3.2 Collision with the sack

$m_n v_1 = m_{\text{tot}} v_2$

• Primary principle: Conservation of Linear Momentum
• Supporting relation: Linear Momentum (Definition)
• Conditions: The collision is brief, the sack starts at rest, and the external horizontal impulse on the ninja+sack system is negligible during impact.
• Setup: The system boundary includes both the ninja and the sack during the collision. The equation is written only in the horizontal direction because the rope force and weight do not supply the relevant horizontal impulse during the short impact.
• Relevance: Tension right after the collision depends on the speed right after the collision, not on the speed at the bottom of the ramp.
• Goal: Convert the pre-collision speed $v_1$ into the post-collision shared speed $v_2$.

3.3 Radial force at the bottom of the swing

$\sum F_r = T - m_{\text{tot}} g = m_{\text{tot}} \frac{v_2^2}{r}$

• Primary principle: Newton’s Second Law
• Supporting relations: Weight (Near Surface), Centripetal Acceleration
• Conditions: We are analyzing the combined mass immediately after the collision at the bottom of the circular path, where the acceleration is purely radial toward the pivot.
• Setup: Choose the radial direction toward the pivot as positive. Then tension points in the positive radial direction, weight points opposite it, and the radial acceleration is $v_2^2/r$.
• Relevance: This is the model step that directly contains the target quantity $T$.
• Goal: Express the rope tension in terms of the known weight term and the centripetal requirement.

Why the procedures are not the main target here

The algebra matters, but most of the physics learning sits in the model steps above. In this problem, the procedures are mostly rearrangement and substitution. That is why self-explanation should focus first on the model: the principle, the conditions, the setup, and the goal of each model equation.

Short Math Example

This one is deliberately short. The point is to contrast a math example where the main explanation target is a model step plus a couple of transformations.

Here the model uses a representational principle, while the later equation steps use transformational principles that preserve equality.

Problem

A gym charges 15 euros per month plus a 30-euro sign-up fee. After $m$ months the total paid is 105 euros. Find $m$.

Short structured solution

Verbal decoding: Target $m$. Given $T, b, r$.

Model: $T = b + rm$

Transformations:

$rm = T - b$

$m = \frac{T - b}{r}$

$m = \frac{105 - 30}{15} = 5$

How to self-explain it

• Model step: The representational principle is the Linear Model. The setup is fixed fee plus monthly rate times months. The goal is to turn the words into one solvable equation.
• Procedure step 1: The transformational principle is additive equality. Condition: equality is preserved when you add or subtract the same expression on both sides. Action: subtract $b$. Goal: isolate the rate-times-months term $rm$.
• Procedure step 2: The transformational principle is multiplicative equality. Condition: equality is preserved when you divide both sides by the same nonzero quantity, so here $r \neq 0$. Action: divide by $r$. Goal: isolate $m$.

This is the contrast with physics: in the gym example, the main self-explanation target after the model is the validity and goal of each transformation. In the ninja example, the richest explanation target is the model itself.

---

Practical Tips

• Start with the first non-obvious step. You do not need equal effort on lines you already understand cold.
• Use drawings and labels. Good self-explanation usually gets better when you connect the equation to a sketch, free-body diagram, or coordinate choice.
• Ask the same small set of questions repeatedly. What principle is this? Why does it apply? How is it set up? What is the goal of the step?
• Explain back after getting help. If an AI, peer, instructor, or textbook helps you, restate the step in your own words before moving on.
• Cut background audio when the explanation starts to blur. If your verbal explanation collapses into vague narration, switch to lower-interference sound or silence. The rule of thumb from Should You Listen to Music While Studying Math or Physics? applies well here because verbal working memory is often the bottleneck.

---

When to Self-Explain

• Whenever you study a worked example with reasoning you do not already own.
• Immediately after getting stuck and checking a solution. This is one of the highest-value moments because the gap is still visible.
• When a step involves principle choice, setup, sign convention, coordinate direction, omitted terms, or a non-obvious transformation.
• Lightly or not at all on trivial steps you already master. The point is not to perform explanation theater.
• Before doing the follow-up problem. Self-explanation helps turn the worked example into something you can use when the next problem is slightly different.

What to Do If You Can’t Explain a Step

Do not stop at “I guess this is what they did.” Explain as much as you can, locate the missing part, then get targeted help.

• Ask for help from the right source. Use AI, a peer, an instructor, or a textbook/reference to focus on the missing principle, setup, or transformation.
• Use AI as a cueing tool, not just a solver. Ask it to question your current explanation, point out what is missing, or explain the gap so you can explain it back.
• Explain the step back in your own words after help. If you cannot restate it, you probably still do not own it.
• Patch the underlying gap. If the problem is missing principle knowledge or weak recall, use elaborative encoding, retrieval practice, or targeted review before coming back to the example.

---

Common Mistakes

• Narrating actions instead of explaining reasoning. “Then I move this term over” is weaker than naming the condition, action, and goal.
• Skipping the setup. In physics especially, students often name the principle but never explain the system, signs, terms, or diagram that make the equation meaningful.
• Treating self-explanation as a last-resort rescue move only. Use it broadly on worked examples, not only in panic mode.
• Replacing follow-up problem solving with explanation alone. Self-explanation reveals gaps and builds understanding, but you still need problem solving to build fluent skill.
• Writing essays for trivial steps while ignoring the important unclear ones. Exhaust the meaningful gap; do not perform explanation for its own sake.

---

Why This Matters

Self-explanation matters because it turns worked solutions into knowledge you can reuse when the surface story changes. It is one of the clearest ways to feel the difference between “I can follow this” and “I can do this.”

It is also domain-general. The same habit helps when studying proofs, code, chess lines, or any other worked reasoning: make the hidden structure explicit, then try to use it yourself. In physics and math, that usually means better principle selection, better setup, and fewer fake-understanding moments.

---

Continue Learning

Become a better problem solver: For the broader solve-from-scratch workflow, see Five-Step Strategy.

Reinforce the principles and forms: Use retrieval practice for names, equations, and conditions to strengthen principle names, equations or forms, and conditions. Use self-explanation and problem solving to retrieve the richer step structure.

Build the weekly study loop: For the bigger study system around self-explanation, retrieval, elaboration, and problem solving, see How to Self-Study Math and Physics Effectively.

Go deeper into the research base: See Learning Literature or the broader framework in Masterful Learning.

---

FAQ

What is self-explanation?

It’s the habit of explaining the hidden reasoning in a worked solution instead of just reading the visible actions. For important steps, that means naming the principle, checking the conditions, describing the setup, and stating the goal.

Do I have to self-explain every step?

No. You do not need full explanation for trivial steps you already understand cold. Focus on the steps where the principle, setup, condition, or transformation is not obvious to you yet.

How long should self-explanation take?

As long as the important unclear part needs. Do not write essays for trivial steps, but do exhaust the explainable content of an important model step or non-obvious transformation if that is where your gap is.

What should I do if I can’t explain a step?

Explain as much as you can first, then get targeted help. Ask AI, a peer, an instructor, or a textbook/reference to focus on the missing step, and then explain that step back in your own words. If the gap is deeper, patch the underlying principle or math with elaborative encoding, retrieval practice, or targeted review.

When should I use self-explanation?

Whenever you study a worked example seriously, especially after getting stuck and checking a solution. It is one of the best ways to stop shallow recognition from masquerading as understanding.

Is self-explanation only for physics?

No. The method is domain-general. But it is especially powerful in physics and math because so much of the difficulty is hidden in principle choice, setup, representation, and non-obvious transformations.

---

How This Fits in Unisium

In the Unisium Study System, self-explanation is about worked-solution steps: you explain why a step is valid, how the setup works, and what the step is trying to do before moving on. The goal is not an AI tutor that solves the problem for you. The goal is structured work on worked solutions so the reasoning becomes yours.

If you want the broader learning framework, see Masterful Learning. If you want the product overview first, start with What Is Unisium?.