Learning a Hard Thing in Twenty-Minute Pieces

One Saturday a few years ago, I cleared the whole day for statistics. I had a thermos of tea, a second-hand textbook with someone else's pencil marks in the margins, and a firm belief that I was about to get it done. By evening I had read three chapters and felt wonderfully virtuous. By Thursday, a friend asked me what the difference was between a mean and a median, and I gave an answer so vague that we both looked at our biscuits.

That is the whole story of my first attempt, and I tell it because it taught me something I now rely on. The best way I've found to learn a daunting subject is to take it in short, regular sessions, about twenty minutes at a time, rather than in a few heroic stretches. Short sessions suit the way memory seems to work, and they make starting much less painful. That second benefit sounds minor, but I've come to think it matters more than the first.

Why a heroic weekend feels productive and mostly isn't

A long session gives you a strong feeling of progress. The material is fresh, the pages turn, and everything makes sense because you've just read it. That feeling is real, but it tells you about the next hour, not about next week.

Memory researchers have been aware of this for a long time. In the 1880s the German psychologist Hermann Ebbinghaus memorised lists of nonsense syllables and tested himself at different intervals. He found that forgetting is steep at first and then flattens out, and that relearning something is easier than learning it the first time. Later work has repeatedly found a related effect, usually called the spacing effect: the same amount of study tends to stick better when it is spread out than when it is packed together.

I should be honest about the limits of what I know here. I'm a reader of this research, not a researcher, and the details of the best spacing for any one person are still debated. But the broad pattern is well established, and it matches my Saturday experience. Cramming can carry you through tomorrow. It is much less reliable at carrying you through next month.

A common misconception is that more hours in one sitting must mean more learning. What seems to matter more is how many separate times you have to pull the idea back out of your head.

What short sessions do for memory

Think of a path across a paddock. Walk it once and you've flattened some grass. Walk it for ten minutes in a single afternoon and you've flattened the same grass a bit more. Walk it once a day for a fortnight and you have a path.

Short sessions work the same way, with one extra ingredient: the gap. If you open your notes the next day and have to reconstruct what a median is, that small struggle does real work. Researchers call it retrieval practice. The act of pulling something out of memory strengthens it more than reading it again does. Rereading feels easy because the answer is on the page. Recalling feels harder because it isn't, and the difficulty seems to be part of the benefit.

Benjamin Franklin described a version of this in his autobiography, though he didn't have a name for it. To improve his writing, he took short notes on the sentences of essays he admired, put them aside for a few days, and then tried to rebuild the essays in his own words. Afterwards he compared his versions with the originals and fixed the differences. The gap and the rebuilding are exactly the two ingredients above.

Seneca, in his letters, gave advice that I read as a cousin of this idea. He warned against flitting from book to book, and described the good reader as something like a bee that gathers from many flowers but then digests what it has gathered. Whatever you read, give it time to settle into something of your own.

Why twenty minutes lowers the dread

Here is the part I only understood after my Saturday failed. A hard subject carries a feeling with it, a sort of weather. You look at the textbook and your shoulders go up. Starting a six-hour session means committing to that weather for six hours, so you put it off, and the putting off becomes its own weight.

Twenty minutes is small enough that the dread has less to hold on to. You are not promising to learn statistics. You are promising to sit with it until the kettle has boiled and the tea has cooled a bit. Most days, I find, the hardest part is the first two minutes, and after that the subject is just a subject again.

There's also a kindness in the arrangement. A missed day costs you very little. You don't have to confess to a streak or make up for lost time. You open the notes the next day and carry on. I think a method you can fail at gently is a method you are more likely to keep.

Twenty is not a magic number. It's what fits between my tram stop and the front door, or before the first meeting of the day. If your attention settles better at fifteen or thirty, use that. Pick a length that makes you slightly relieved rather than slightly afraid.

A twenty-minute session, step by step

This is the shape I use now. I've adjusted it many times, and you should feel free to as well.

1. Recall first (about 4 minutes). Before opening anything, write down what you remember from last time. Don't peek. Gaps are fine and in fact useful, because they show you where the next session should go.
2. Check (1 minute). Compare with your notes and mark what you missed.
3. One new idea (10 minutes). Learn a single small piece, not a chapter. Work one example by hand.
4. Explain it (4 minutes). Write two or three sentences as if telling a friend who has never heard of it. If you can't, that is the useful discovery.
5. Leave a note (1 minute). Write down where you'll start tomorrow. It makes starting easier, because you are following a note rather than making a decision.

A worked example: mean versus median

Suppose I note how many minutes I wait for the tram on five mornings: 4, 7, 5, 9 and 5.

The mean is the total divided by the count. The total is 30 and there are five mornings, so the mean is 6 minutes. The median is the middle value once they're in order: 4, 5, 5, 7, 9. The middle one is 5.

They're close, so why do both exist? Now imagine a sixth morning when the tram is very late, and I wait 40 minutes. The six values are 4, 5, 5, 7, 9 and 40. The total is now 70, so the mean is 70 ÷ 6, about 11.7 minutes. The median is the average of the two middle values, 5 and 7, which is 6.

One strange morning nearly doubled the mean, while the median barely moved. That is the point. The mean uses every number's size, so extreme values pull on it. The median only cares about order, so it resists them. Neither is "right". They answer slightly different questions: "what is the total waiting spread evenly?" versus "what does a middle morning look like?"

That one example, worked by hand, is a perfectly good twenty minutes. If I had met it on my Saturday, between chapters three and four, I think it would have vanished.

A week of statistics in twenty-minute pieces

To show how the pieces fit together, here is how I might lay out a first week. It is only an illustration, and the order is a matter of taste.

• Day 1: Mean and median, with the tram example above.
• Day 2: Recall day 1, then the mode, the most frequent value (5 in my first set).
• Day 3: Recall, then deviations: how far each value sits from the mean.
• Day 4: Recall, then notice something small. For my first five mornings the deviations are −2, +1, −1, +3 and −1, and they add up to zero. That is true for any set of numbers, and it hints at why people later square the deviations before averaging them. If you don't, the positives and negatives simply cancel.
• Day 5: Recall everything so far, with no new material. A review day is allowed and often does the most good.

Notice that each session begins by dragging the earlier ideas back out. That is the spacing doing its work, and it's also why a tiny new idea per day adds up faster than it seems it should.

Where this might not suit you

I'd be overselling if I said this works for everyone. Here is what I've noticed, and what I'm still working out.

Some subjects need long, unbroken stretches. Learning to write a complicated proof or to follow a long argument can fall apart if you are interrupted every twenty minutes. For those, I use short sessions for the groundwork and a longer one for the joining-up.

Some people simply find it hard to stop after twenty minutes once they've started, and that's fine. If you're enjoying it, carry on. The twenty minutes is a promise to start, not a limit.

Results also vary from person to person, and from season to season. In a busy week I manage four sessions out of seven. I used to count the missed ones as failures, but I now count the four as the thing that happened.

Finally, I can't tell you the ideal gap between sessions, and I'm cautious of anyone who claims to know it for you. Daily works for me. You may find every second day suits your life better.

Something to wonder about

My library returns desk taught me something about patience that no book has: a pile of returned books gets shelved one at a time, and by closing time the trolley is empty. Nobody has ever cleared that trolley in a single heroic push, as far as I've seen.

I wonder how much of what we call talent is really the quiet accumulation of small sessions that nobody saw. And I wonder what the right size of "small" is for you, whether twenty minutes, ten or forty. If you try the method, I'd be glad to hear what you find. Thank you for reading, and I hope your first session goes gently.