back · main · about · writing · notes · d3 · contact


Five methods to learn effectively
14 Jul 2018 · 2111 words

There’s a world of difference between effective and ineffective learning methods.

Effective learning methods ensure your knowledge is robust, intuitive and long-lasting. It’s knowledge you can use to build on. Ineffective learning methods will achieve the opposite.

What separates the two? The successful methods place an emphasis on

By contrast, ineffective methods emphasise

It’s easier to use an ineffective method than an effective one. Rereading is easier than self-quizzing, but self-quizzing is much more effective.

If we’re not consciously choosing an effective study method, we’re likely to use easier alternatives - which tend to be ineffective. We must be deliberate about the ways that we learn.

Effective learning methods tend to be recall-based methods, meaning they challenge you to generate answers rather than recognise them. By using recall based methods you’ll remember more, be able to apply your knowledge better and have a more intuitive feel for the material.

Here are five of these effective recall-based methods.

1. The textbook method

**Source:** Cal Newport

In one sentence: Write your own textbook chapter on a topic.

Details: With this method, you’ll be creating your own version of a textbook chapter on the thing that you’re learning. It’s more like a mini-textbook chapter, really, since you’ll need to condense the original material.

Here’s how it works:

  1. Identify the scope of the topic, or what exactly you’ll be writing about. Err on the side of making the scope larger., since it’s better to make the mini-textbook chapters bigger rather than smaller. A good scope could be a textbook chapter, or a group of a few lectures.

  2. Write your own version of the textbook chapter, trying as much as possible to go from memory.  Some things that you might like to include:

    1. A list of definitions, or perhaps theorems and lemmas
    2. High-level description of concepts and how they interlink
    3. Good, succinct descriptions of the big ideas, frameworks and theories
    4. A discussion on how other ideas connect to the current material, and how they compare and contrast
  3. Create a separate document containing an outline of the chapter. The outline contains the headings of your chapter, like each definition, concept, big idea or theory.

  4. To review, rewrite the textbook chapter using the outline as a memory prompt. Don’t refer to the original version until you’re finished. Identify what you couldn’t reproduce and quiz/review those topics. The aim isn’t to replicate the wording of the original chapter exactly, but rather to see where you have trouble.

The textbook method can be quite time consuming, so if you’re time-poor perhaps save it for topics that give you the most trouble.

Why is it useful? Writing a textbook chapter isn’t easy. You’re forced to be concise, accurate and to create a mental structure for how the concepts fit together. You’ll need to develop your knowledge chunks and organise them to understand things at a deeper level.

You also have to write about everything in your scope. No hiding behind knowledge gaps.

Further reading

Original source

Why it works

Experimentation with a LaTeX template

Taking notes by textbook method

2. The notebook method 

**Source:** Cal Newport

In one sentence: Go somewhere quiet and write out your thinking.

Details: This method is used to make progress on a difficult problem. Here are the steps:

  1. Get a notebook and a pen. Go somewhere quiet and distraction free, and don’t bring your phone.
  2. Spend a good chunk of time (1-2 hours) working out your thinking on the problem. You could draw diagrams, flesh out links between concepts, write out definitions, sketch out different approaches, or just write whatever comes into your head.
  3. Spend the last 20 minutes summarising your thinking on a new page.

You could use this method to think understand the similarities between two algorithms, figure out the flow of a research paper, make progress on a problem set or even design a machine learning project. You could use this method to organise your knowledge on a topic, like Richard Feynman did.

Why is it useful? By drawing links between concepts, creating analogies and metaphors, sketching out diagrams and organising your thinking, your understanding becomes resilient to context switches. Avoiding fragile knowledge needs to be one of the primary aims of learning.

Further reading

Original source

Feynman’s organisation of knowledge

The thoughts of Terence Tao.

3. The ADEPT method

Source: Kalid Azad

In one sentence: Build intuition through analogies, diagrams, examples and simple explanations.

What is it? The ADEPT method is built around the idea that when learning a concept, intuition isn’t optional.

This method is a thorough one designed to create lasting results. It might not feel quick while you’re doing it, but over the long term you’ll save considerable time as you don’t need to relearn your embedded concepts.

Here’s the ADEPT method:

You can also this method to understand material from the textbook, or from lectures – you just need to do ADEPT in reverse.

Why is it useful? When you return to material years later, odds are you won’t remember much of it (unless you build other knowledge on it). The things you’ve rote learned or memorised will have gone. Knowledge that was partially understood needs to be relearned. The knowledge acquired with the ADEPT method, that which you have an intuition about, that which you know how to use - that will stay.

Further reading Original source

ADEPT method to learn machine learning

The Feynman Technique

Source: Many

In one sentence: Explain a topic as if you are talking to a five year old.

Details: Ever tried to give an explanation to someone and found it’s not as easy as you thought?  Did you notice that when giving the explanation, you got stuck when you didn’t know something that well? The Feynman technique was born around this idea; generating explanations is a good way to learn.

You must make your explanations as simple as possible, as if a five year old child could understand it. Five year olds can’t grasp complexity as well as you can, and can’t understand many technical words either, so you have to find plain english explanations for any jargon or technical concepts.

If you stumble over your words, it’s a sign you don’t know the material as well as you think you do. Review what gave you trouble and try again.

If your explanation is wordy, not clear or feels like you’re faffing, try to simplify it by stripping the explanation down to its core arguments. You could also create a metaphor or analogy to convey an intuitive understanding of the topic. It’s not just kids who love a good analogy; it’s how you learn best too.

Bonus step: write down your story to refine it even further, or even publish it on a blog somewhere!

Why is it useful? Generating simple explanations for complicated topics helps with the knowledge chunking process. The explanations also help you see how concepts relate and lets you understand the material in a broader context. You can also identify your weak areas and work on strengthening them.

As Feynman said: “What I cannot create, I do not understand

Further reading

Explanation 1

Explanation 2

Explanation 3

The quiz and recall method

Source: Many, Cal Newport

In one sentence: Quiz yourself while keeping the textbook closed

Details: Here’s how the method works:

  1. Obtain a list of quiz questions. Where the questions come from is quite flexible. You could take them from a textbook or lecture course, or while you are learning material write down quiz questions covering  the key concepts and ideas.
  2. Organise these questions into groups, sorting by topic. These are your quizzes.
  3. Let some time pass and then attempt to do the quizzes. Don’t look at the answer until you’ve answered with everything that you remember.
  4. Put a star next to all the questions that you didn’t know. Review the material and try again.

For technical subjects you can add technical explanation questions: specialised questions that ask you to go deeper into a technical concept. Some examples could be listing the steps to do a type of practice problem, reproducing a proof, or discussing the situations a mathematical technique applies in.

This method only works well if the questions are comprehensive enough. It’s hard to get the quantity right; if there’s not enough questions you’ll miss areas, but if there’s too many of them you might get overwhelmed and skip over some.

  1. Write a collection of short questions from your notes on the subject, comprehensive enough to cover everything that you need. The questions should be answerable in about a sentence. For example
    • Evidence Nero started the fire of Rome
    • Emperor that came before Caesar
    • Where Cleopatra and Mark Antony met
  2. Arrange the questions by topic (e.g. “Early Roman emperors”). Have one page per topic cluster. Keep the answers somewhere where you can’t see them when answering the questions.
  3. Add some background questions on the key concepts that need more detailed longer answers
  4. When reviewing the questions, the short questions can be answered very quickly (a few words for each is fine). Answer the conceptual questions in more detail, giving your answers out loud.

Why is it useful? Besides the benefits similar to the other recall-based techniques above, this method is also great for spaced repetition. After the quiz question sets have been created, it’s easy to review them every few weeks or so, minimising the chance you’ll forget the material.

Further reading

Wiki entry

Technical explanation questions

Focused cluster - original source

Focused cluster - memorisation

Focused cluster - note streamlining


back · main · about · writing · notes · d3 · contact