You spend three hours reading a calculus textbook. You highlight equations. You nod along. Then you close the book and cannot explain why the chain rule works. That is not learning. That is pattern matching.
The Feynman Technique is a structured method to expose what you do not understand. It works because it replaces passive consumption with active explanation. Here is how to apply it to any STEM subject in 2026 — and where most students get it wrong.
Step 1: Pick a Concept and Write Everything You Know on a Blank Sheet
Grab a blank sheet of paper. Write the name of the concept at the top. Then write everything you know about it in plain English. No jargon. No formulas unless you can explain what each symbol means.
For example, take “gradient descent” in machine learning. Do not write “optimization algorithm that minimizes cost function.” Write: “Gradient descent is like walking downhill blindfolded. You take a step in the direction that feels steepest downward. Then you repeat until you hit the bottom.”
Common mistake: Most students skip this step. They think they already know the material. They do not. Writing forces you to see the gaps.
Specific target: If you are studying linear algebra, pick “eigenvectors.” If you are studying thermodynamics, pick “entropy.” Do not pick a whole chapter. Pick one atomic concept.
Step 2: Identify the Gaps — Where Your Explanation Breaks Down
This is where the real work happens. Read through what you wrote. Mark every sentence that feels vague, uses jargon you cannot define, or relies on a formula you memorized without understanding.
Example: You wrote “entropy is the measure of disorder in a system.” But when someone asks “disorder of what?” you freeze. That is a gap. Go back to the textbook, a lecture video, or a reliable source like Khan Academy or 3Blue1Brown. Find the explanation that fills that gap.
Data point: A 2014 study in Science showed that students who practiced retrieval (explaining without notes) scored 50% higher on conceptual tests than those who reread the material. The Feynman Technique is retrieval practice with a built-in diagnostic.
Real product reference: For physics concepts, the “Feynman Lectures on Physics” (Addison-Wesley, three volumes, ~$150 for the set) remain the gold standard for clear explanations. For calculus, “Calculus Made Easy” by Silvanus P. Thompson (St. Martin’s Press, ~$15) uses the same plain-language approach.
Step 3: Simplify and Use Analogies — Teach It to a Child
Now rewrite your explanation. Use only the 1,000 most common English words. Imagine you are explaining it to a 12-year-old who has never taken a STEM class.
For “eigenvectors”: “Imagine a rubber band stretched in one direction. If you push it exactly along that stretched direction, it only stretches more. It does not twist. That special direction is an eigenvector. The amount it stretches is the eigenvalue.”
If you cannot explain it simply, you have not understood it well enough. This is the core of the Feynman Technique. Richard Feynman himself said, “What I cannot create, I do not understand.”
Table of common STEM concepts and their child-friendly analogies:
| Concept | Jargon Explanation | Child-Friendly Analogy |
|---|---|---|
| Derivative | Rate of change of a function | Your speedometer shows your speed at each instant. That speed is the derivative of your position. |
| Logistic Regression | Statistical model for binary classification | Like asking “will it rain?” and the model says “70% chance yes” based on clouds and wind. |
| Ohm’s Law | V = IR, voltage equals current times resistance | Water flowing through a pipe. Water pressure = voltage. Flow rate = current. Narrow pipe = resistance. |
| Big O Notation | Upper bound of algorithm runtime | If sorting 10 items takes 1 second, sorting 100 items takes about 100 seconds. That is O(n²) — slow for big lists. |
Warning: Simplification is not dumbing down. It is distillation. A good analogy captures the essence without misleading. Test your analogy on a friend who does not study STEM. If they nod, you are on the right track.
Step 4: Repeat and Refine — The Iteration Loop
This is the step most people skip. They do Step 1, feel clever, and move on. The technique requires iteration. Go back to Step 2. Find new gaps. Refine your analogy. Add detail where your simplified version oversimplified.
Example iteration cycle for “Markov chains”:
First pass: “A Markov chain is a random process where the next step depends only on the current state.”
Gap identified: “Why only the current state? What about memory?”
Refined: “Imagine a frog hopping between lily pads. Where it hops next depends only on which pad it is on right now. It does not remember the pad before that. That is the Markov property — no memory of the past.”
Second gap: “What about weather prediction? Weather has memory.”
Second refinement: “You can still use Markov chains if you define ‘state’ as the weather for the last two days. Then the next day depends on the last two days. That is a higher-order Markov chain.”
Each iteration deepens understanding. Aim for three passes per concept. One pass is not enough.
When the Feynman Technique Fails — And What to Do Instead
The technique works for most STEM topics, but it has limits. Here are three situations where it struggles and what to do.
1. Highly procedural topics (e.g., solving differential equations): The technique is great for conceptual understanding but weak for procedural fluency. If you need to solve 50 ODEs for an exam, you need practice problems, not analogies. Use the Feynman Technique to understand why the method works, then drill the procedure separately.
2. Topics with no intuitive anchor (e.g., quantum mechanics, abstract algebra): Some concepts resist everyday analogies. “Spin” in quantum mechanics does not behave like a spinning top. For these topics, the Feynman Technique still works — but your analogies will be imperfect. Accept that. The goal is functional understanding, not perfect poetry.
3. When you are too tired to think: The technique requires active mental effort. If you have been studying for six hours, your explanations will be sloppy, and you will not identify real gaps. Stop. Sleep. Come back fresh. A 2019 study in Nature Communications showed that sleep consolidates procedural and declarative memory — your brain literally fills gaps while you rest.
Alternative method for procedural topics: Use the “Blurting Method” — write everything you remember about a procedure from memory, then check your textbook for missing steps. This is faster than the Feynman Technique for pure procedure.
Alternative method for abstract topics: Use the “Concrete-Representational-Abstract” (CRA) framework. Start with a concrete example (e.g., a coin flip for probability), then a visual representation (tree diagram), then the abstract formula (Bayes’ theorem). The Feynman Technique fits best at the concrete stage.
Final recommendation: If you are studying for a conceptual exam (physics, computer science theory, biology mechanisms), the Feynman Technique is your best tool. Use it for one concept per day, three iterations each. If you are studying for a procedural exam (calculus, linear algebra computations, statistics calculations), use the Feynman Technique once per week for conceptual understanding, and spend the other six days on practice problems. That combination yields the highest retention per hour studied.
