How to remember what you read: what the research says

Here's how to remember what you read, according to the research: test yourself on it. Close the article, try to recall the main points or answer a few questions about it, and then do the same thing again a few days later and about a week after that. Get a full night's sleep after you read something you want to keep. Rereading and highlighting feel productive, but on their own they do little for memory.

The rest of this post goes through each step with the studies behind it, and ends with a routine that takes about ten minutes for an article.

Why you forget most of what you read

Forgetting is fast at first and slow after that. Hermann Ebbinghaus measured it on himself in the 1880s with lists of nonsense syllables, and the shape he found is now called the forgetting curve: a steep drop in the first day, then a long, slow slide. In 2015 two researchers in Amsterdam, Jaap Murre and Joeri Dros, ran the experiment again and got almost the same curve.

Real reading behaves the same way. In a 2006 study at Washington University, students read a short prose passage and then simply studied it again. Five minutes later they could recall 81% of it. A week later they recalled 40%.

So if an article you read last month feels like a blur, nothing is wrong with your memory. You read it once and never brought it back.

Rereading and highlighting do little on their own

In 2013 a team led by John Dunlosky reviewed ten common study techniques for the journal Psychological Science in the Public Interest. They rated rereading, highlighting and summarizing as low utility, and gave the top rating to two techniques: practice testing and spreading practice out over time.

Rereading is easy to fall for. The second time through, the words look familiar, and familiar feels a lot like knowing. The real test is whether you can bring the idea back with the page closed.

Highlighting has the same problem. It marks what seemed important while you were reading, and it doesn't make you retrieve anything later. A highlight you never look at again works like a bookmark. If you read on paper, annotate the book instead: a few words in the margin, in your own words, make you think about the idea in a way a line under it doesn't.

Test yourself as soon as you finish

Pulling information out of your memory strengthens it more than putting it in again. Psychologists call this the testing effect or retrieval practice, and study guides call it active recall.

The Washington University study above, by Henry Roediger and Jeffrey Karpicke, compared two groups. One group reread the passage several times. The other read it once and then wrote down everything they could remember, three times, without looking. A week later the group that tested themselves recalled 61% of the passage and the rereaders recalled 40%, even though the rereaders had read it about 14 times and the testers about 3 or 4.

To use this on an article:

1. When you reach the end, close it.
2. Write or say the two or three main points from memory.
3. Open it again and check what you missed or got wrong.

Questions make this easier, because a good question asks for a specific answer. "What did the article say about sleep?" is vague. "Which night of sleep matters most after you study?" has one right answer, and you either know it or you don't.

This is the idea Bundle is built on. When you finish an article in Bundle, a Quick check asks you a few questions about it, and you keep the good ones as cards.

QUICK CHECK · SCIENCE Which night of sleep matters most after you study? The night after. It is at least as important as the night before, because that sleep cements what you learned. Sleep On It: How Snoozing Strengthens MemoriesShow in article

SkipKeep card

Review it again a few days later

One test right after reading helps. A second and a third, spread over days, help more. A 2006 review of 317 experiments by Nicholas Cepeda and colleagues found that spaced reviews beat reviews crammed together, and that the best gap between reviews grows the longer you want to remember something.

A simple schedule for something you read today:

• the next day or two
• about a week later
• about a month later

You may have seen this called the 2-7-30 rule: review after 2 days, 7 days and 30 days. It isn't from one specific study, but it follows the same idea, and it's easy to remember.

Spaced repetition apps go one step further and fit the schedule to you. If you answer a card easily, it can wait longer before it comes back. If you forget it, it comes back soon. Anki added a scheduler like this, called FSRS, in 2023 as an option you can switch on. Bundle uses FSRS too: every time you remember a card, it waits longer before the next review, and a card you forget comes back the next day.

Write it in your own words, with the article closed

Summarizing scored low in the Dunlosky review, but how you summarize matters. Copying the main sentences while the article is open is mostly copying. Writing the main ideas with the article closed is a recall test, which is the part that works.

Two formats that help:

• The blank sheet, which Shane Parrish describes on Farnam Street: after a chapter or an article, write what you remember on an empty page, then check the source and fill the gaps.
• The 3-2-1 summary that teachers use: three things you learned, two you found interesting, and one question you still have. The open question gives you a reason to come back.
• A commonplace book, one notebook where you copy the passages worth keeping from everything you read and write what you think next to them.

Connect it to something you already know

New ideas are easier to keep when you tie them to things you already know. While you read, ask yourself why the claim is true, where you have seen something like it, and what you would do differently if it is true. Reading with one question in mind before you start does something similar, because you notice the parts that answer it.

You don't have to do this for every paragraph. Doing it for the two or three ideas you want to keep is enough.

Sleep on it

Sleep helps the brain keep what it learned that day. In NIH's article on sleep and memory, Harvard sleep researcher Robert Stickgold says that the sleep you get the night after you study is at least as important as the sleep you get the night before.

In practice this means you shouldn't trade sleep for one more article. If you read something important in the evening, go to bed, and do your first review the next day.

Decide what is worth remembering

You don't need to remember every article you read. Pick the one to three ideas worth keeping from each article, turn them into questions, and let the rest go.

This also keeps reviews short. Ten good cards a day take a few minutes. Two hundred cards made out of every highlight turn into a chore.

It's also fine to stop reading an article that isn't giving you anything. For books, a reading log with one line on what each book gave you shows over a year which ones you kept.

A 10-minute routine for any article

1. Before you start, skim the headings and write down one question you want answered. (1 minute)
2. Read without notifications. Highlight a line only if it says something you want to keep.
3. When you finish, close the article and write the main points from memory. Then check it and fix what you missed. (2 minutes)
4. Turn the two or three ideas worth keeping into questions with short answers. (2 minutes)
5. Answer those questions again tomorrow, in about a week and in about a month. (1 or 2 minutes each time)
6. Sleep.

How Bundle does this for you

I made Bundle so this routine happens without the busywork. You save an article from any app with the share button, and Bundle files it under a topic. When you finish reading, the Quick check asks you a few questions about it, and you keep the good ones as cards. One reminder a day brings back the cards that are due, which usually takes about five minutes, and each card waits longer every time you remember it.

Bundle is on the App Store for iPhone, and you can try it free for 14 days.