This is probably what everyone has been waiting for, but do remember that this alone doesn’t work without the other two. To repeat the 3 practical insights from this article:
- You can only encode what you can pay attention to.
- You can only retrieve what you’ve encoded.
- The rate at which you learn is directly determined by your prior knowledge.
Did you notice that in the last lesson, I told you that deconstructing words is probably the most important in your playbook? That’s because when you know the meaning of suffixes, prefixes, and root words, it gets easier to decode even unknown jargon. They’re kinda like “non-domain specific prior knowledge,” or “knowledge decoders,” if you like that term.
Having a lot of prior knowledge means having a lot of “decoders”
To quote Anderson & Bradshaw in their paper, Elaborative encoding as an explanation of levels of processing:
A person is best at elaborating that type of information with which he or she has had the most experience. This is because one’s elaborative abilities are a function of what one knows about that domain. Semantic processing tends to produce better memory than surface processing because one tends to have a richer knowledge base about the semantic properties of the concepts than the surface properties of the words.
Basically, your ability to make sense of new information is increased when you have more prior knowledge. The more you know, the more you’ll be able to know. It’s a virtuous cycle — a learning flywheel, “learning momentum,” “compound effect,” whatever.
Yet, there is a crumbling path to this virtuous cycle of knowledge
You may have experienced this already: the more knowledge gaps you have, the slower you learn the next topics.
And what causes these knowledge gaps? It’s called forgetting. More appropriately, the natural process of forgetting. The interesting about this is, well, it’s natural. Everybody forgets. Let me explain.
Any biological system is subject to decay — including its parts. Since our brain is part of this biological system, and our memories (synapses) being part of the brain, this means that memories are NO exception to this rule. Forgetting, then, is a natural process.
Sure, building upon old knowledge and/or producing coherence creates more robust memory traces. But this isn’t enough to get you to long-term retention, for two practical reasons:
First, while retention is also a function of processing depth, it alone doesn’t guarantee long-term retention.
A shallow level of processing may lead to a faster rate of forgetting, and a deeper level of processing may lead to a slower rate of forgetting. But even though memory psychology wasn’t really able to define what “depth” really means, one fact still remains: regardless of depth, you’re still undergoing the natural process of forgetting.
It’s the main reason why you shouldn’t resort to “deep understanding” alone to remember stuff in the long term. No matter how vivid you understand something today, it will still fade unless you do something to maintain it.
So all we have to do is to process, process, process, right?
Wrong. This brings me to the second reason.
Second, not every learning opportunity is an opportunity to go deeper into what you already learned.
In University, for example, there is a long gap before the knowledge of Cell Biology gets “built upon” by new knowledge. If you’re an Engineering student, there may also be a long gap before your Physics knowledge gets “built upon” by major Engineering subjects.
So before we could even build upon our understanding of one topic, the natural process of forgetting would have already beaten us to it.
Before we could ever have the chance to reinforce the hard-earned connections we made during Processing, those connections would’ve already crumbled to a state of inaccessibility.
(Fun fact: Psychologists believe that encoded memories don’t really get lost — they just become harder to access, which explains why relearning is easier.)
Therefore, instead of just relying on understanding alone for long-term retention, we have to use tactics alongside it that are more practical for that specific goal — namely, rehearsal strategies.
Rehearsal strategies are the core of Permanence
And the purpose of Permanence in your study system is to maintain the hard-earned understanding you got from deeply focusing on and mentally processing your material.
What happens, then, is that you can start using this knowledge to process a more complex material, and after that, you could move on to the next one, and the next one, and the next one.
What’s more, with rehearsal strategies, you don’t need to wait for an opportunity to build up your knowledge just to maintain what you already know. You only have to formulate your prompts a few times, then use them for months or years on end.
The ultimate result is that you won’t have to go back to what you have already understood before moving on to the next topic.
But, like any other strategy, rehearsal strategies also easy to mess up and overthink
- “Active recall + spaced repetition doesn’t work for me”
- “It’s taking so much time to make my Anki flashcards!”
- “What’s the best algorithm for retention??”
If you’re thinking this way, here’s how you do it in an efficient way that doesn’t need overthinking:
- Formulate prompts from your insights
- Put the questions and answers directly into your spaced repetition software
- Review on a spaced schedule, then reformulate to retain usefulness
Let’s expand on each.
Step 1. Formulate prompts from your insights
The main problem with the common advice to “do active recall + spaced repetition” is that they completely neglect coherence and promote these techniques as some kind of magic panacea.
So, let’s break down the instructions to make it clear.
“Formulate prompts…” means that you’re making questions, not statements.
“…from your insights” means that you’re formulating based on the insights you got from mental processing. (Did you keep your insights from the last lesson?)
Example of not formulating from your insights:
Q: First law of thermodynamics
A: Energy is neither created nor destroyed; it can only be transformed into other forms
First, this prompt lacks clarity. Just look at the damn thing — you still have to figure out what’s being asked! And only after that would you (hopefully) be able to search your memory for the answer. What a drag!
Second, this question is clearly just taken out of the textbook definition.” You’re not testing your understanding; you’re just testing how well you regurgitate.
To be fair, this can also be treated as “just a fact,” but if you do it that way, you’ll likely won’t be able to apply this knowledge to related topics. Specifically, the example is useful for potential & kinetic energy problems, but if you just memorize that, it likely won’t be in your “toolbox” when solving problems.
Now here’s an example of formulating:
Q: Which law of thermodynamics states that energy is only transformed into other forms rather than being created or destroyed?
A: First law of Thermodynamics
Q: What does the first law of thermodynamics state about energy?
A: It is neither created or destroyed
It’s not necessary to refactor a single question into multiple ones right away, but as a rule of thumb:
Avoid making questions that obviously have the same form as the raw material. It’s a recipe for “recognizing your notes,” feeling like an impostor, and useless retention. (I’ve learned this from painful experience…)
From the last lesson, I showed you how to turn raw information into insight using a lesson from the study system course.
Now let’s continue this example so you can see how I make questions. From the succeeding paragraphs in the Physiology textbook, I ended up with this:
Now here are the questions I’d make for that:
REMINDER: Follow my mental process and try to adopt it for yourself. You don’t need to explicitly create the same table. What matters here is the mental process. So don’t feel the need to use the “perfect note-taking method.” Observe how I approach the thinking process rather falling into the trap of obsessing over the methods I’m using. Again, the mental process is more important.
Notice that I’ve rephrased my insights after formulating the question; I didn’t just ask “what is gross anatomy” or “what is microscopic anatomy.” You should try to do the same when formulating your questions.
Unlike Bento Box Thinking, I haven’t made specific steps about creating effective prompts because they could vary per insight. That’s why, instead of having clear cut rules, I wanted you to adopt my thinking process. That way, you can develop the question-making skills that are specific to your own.
So, feel free to experiment with your questions. We’ll talk more about “why” later, but part of the reason for this laid back approach is that when you do Processing the right way, even simple questions would help you remember what you studied.
Overall, what matters is the fact that you’re doing both Processing and Permanence. You don’t need to create perfect questions so long as you prompt information that’s already made coherent.
Step 2. Put the questions and answers directly into your spaced repetition software
We all know that spaced repetition is the “gold standard” for long-term retention, yet scheduling it is tedious. That is the only reason why we use spaced repetition systems like Anki. These tools merely schedule your reviews individually using a pre-set algorithm — adjusting them based on your actual recall rate. (In Anki, these are: again, easy, good, hard)
Said another way, we use Anki NOT because it’s a “magic BEST TOOL for remembering anything 100X faster.”
We use it because it makes scheduling convenient — just put in the questions, and let it do the scheduling. Same goes for SuperMemo, Quizlet, Mnemosyne or any other SRS.
But regardless of the spaced repetition software you use, the principle is the same. So long as you test yourself of your insights at a spaced interval, it’s enough to retain it in a useful way. (i.e. in a way that allows you to use it for future Processing)
And it really is that simple.
Otherwise, if you don’t want to use any spaced repetition tool because you’re studying only for the short-term, then take a sheet of paper to formulate your questions. Then, simply use your notes to check the right answer.
For the revision intervals, I recommend a simple approach:
- First session: The same day you studied it
- Second session: 3 days after last session
- Third session: 1 week after last session
- Fourth and next sessions: 2 weeks after last session (Note: This is now a uniformly spaced interval.)
It doesn’t have to be strictly 2 weeks, but for manual scheduling, this gives you the benefits without going insane about “optimal intervals.”
NOTE: I was about to recommend an expanding review interval, but in a recent meta-analysis, researchers found no significant difference between an expanding vs. a uniform interval for spaced repetition. Of course, for better implementation, a simple approach works best, thus my recommendation of “every 2 weeks.”
Step 3. Review on a spaced schedule, then reformulate to retain usefulness
Of course, you don’t maintain your knowledge just by making questions. You maintain your knowledge through retrieval practice. If you’re using a spaced repetition software, this process becomes intuitive and self-explanatory. Just review every day.
If you’re going the analog route, all you need to do are 4 steps:
- Get your question and answers sheet
- Hide the answers
- Try to answer a question from the top
- Then after answering each question reveal the answer for feedback
When you get a question wrong, simply mark it with a pencil, and then try to answer the question again after you finish everything. Of course, do it without looking at the correct answers.
After a few days of successful recall, you’ll find that you have made some quite useless questions. One sign of that is when they make you ask “what does this question really ask?” After a while, you’ll figure out the answer, but then realize that the question really sucks.
In that case, you should reformulate them into better questions.
For example, while you’re doing your reviews and you see the question, “What are interface circuits?,” you’ll eventually say, “What the heck are you talking about?”
That’s when you should reformulate the question into a better one:
Decent question:
Q: What are interface circuits?
A: Circuits that are used to connect systems with different signal characteristics
Better question:
Q: Why are interface circuits called “interface” circuits?
A: Because they are circuits that are used to connect systems with different signal characteristics
This is how you make your questions useful in the long run while developing your question-making skills along the way. If you’re aiming for long-term retention, then questions that are only comprehensible in the short-term are useless.
“I’m studying a lot of subjects, so doing this probably won’t make me remember anything”
Look, if you think that way, it’s because you’re mistaking your short-term memory for your long-term memory. You probably think that if you don’t hold the information actively in your head, then you will eventually “lose grip” and forget it.
The fact is this:
You don’t need to keep things at the top of your head to remember them.
As soon as you’ve formed coherence, you’ve already made that knowledge easier to retrieve. And when you combine retrieval practice, you’re making that memory even more accessible. To finish it off, spaced repetition makes that memory trace increasingly strong.
So if you’re able to make the material coherent AND answer the questions you’ve formulated, then that’s already the sign that YOU HAVE ALREADY ABSORBED IT. All that’s left is spaced repetition to help long-term retrieval.
If you ever forget anything, it’s because that information didn’t have enough elaborations by nature, your questions may be poorly formulated, or both.
When you process raw information and use the conservative spacing I’ve recommended, however, I highly doubt you’ll forget them. (Assuming you formulate questions from insights, that is.)
“Is the spaced repetition algorithm is more important for retention?”
Not necessarily. The specific algorithm is less important so long as you have processed the material at a deeper level and have spaced intervals for knowledge maintenance.
I can’t stress this enough: In my experience, when information is more elaborated and is more meaningful, you can be less conservative of retrieval schedule and use a longer spacing.
Thus, instead of wasting your time in finding the perfect algorithm, focus on the bottleneck — the hardest part — of your system: Processing.
It’s better to focus on improving your processing skills and just use a good enough or conservative spaced repetition schedule.
To put it bluntly, this is much better than having a perfect algorithm but merely regurgitating information that you don’t even understand.
Summary: To maintain knowledge, use rehearsal strategies
The rate at which you learn is directly determined by your prior knowledge, because having a lot of prior knowledge makes “decoding” new information easier. It’s similar to how you need to learn the alphabet before trying to learn sentences. Thus, when you’re able to constantly increase the amount of prior knowledge, you’ll theoretically be able to decode more complex information. It’s a virtuous learning cycle.
But the path to this virtuous cycle is blocked by knowledge gaps. Knowledge gaps mostly stem from a far interval between acquiring knowledge and elaborating on it. It seems that knowledge gaps are just normal, because forgetting is a natural process.
The good news is that rehearsal strategies help counter natural forgetting. Yet, they’re easy to mess up. This is why I recommended 3 steps:
- Formulate prompts from your insights
- Put the questions and answers directly into your spaced repetition software
- Review on a spaced schedule, then reformulate to retain usefulness
Make your acquired knowledge permanent
Aim for useful retention instead of mere remembering. If we’re not able to apply our knowledge to learn more complex materials, or worse, when we don’t even understand what they mean, retention becomes useless. Put another way, it doesn’t matter if we can remember a lot if we’re not able to use what we’ve retained.
I strongly believe that maintaining knowledge should have higher priority than learning new stuff. After all, if you’re a student, it doesn’t matter how much material you’ve finished if you can’t remember them for your exams.
And even if you’re not taking exams and are just learning for your own sake, would you really let your hard-earned knowledge fade into the abyss? I don’t think so.
Action Steps
- Take the notes you have from the last lesson.
- Use those notes to formulate your questions. (Warning: This is hard for most beginners.)
- Recall these questions using the instructions laid out in this lesson, and reformulate as needed.
