We run a typing site whose lessons are built to be finished in minutes, so we have an obvious interest in the answer to this question — read accordingly, and check the sources, which are all linked. That said, the answer is unusually solid. How to schedule practice is one of the oldest questions in learning research, the results have pointed the same way for a century, and the classic experiment happens to be about learning to type.
The postmen who learned to type
In 1978, the British postal service needed to teach thousands of postal workers to type. Researchers split the trainees into four schedules: one hour a day, two hours a day, one two-hour session a day, or two two-hour sessions a day. Everyone got the same instruction; only the timetable differed. The result was clean: the one-hour-a-day group learned the most per hour of practice, and the four-hours-a-day group the least1. The more the practice was piled up, the less each hour of it bought.
The study is remembered for a second finding, though. The one-hour group — objectively the most efficient learners in the building — were the least satisfied with their own progress, and the group on the heaviest, least efficient schedule was the most satisfied. Their calendars agreed with their feelings: the efficient group needed more weeks to reach the same standard, because their efficient hours came one per day. Per hour spent, they were far ahead. It did not feel that way from inside.
The schedule that feels slowest is the one that wastes the fewest hours.
A century of the same result
The postmen were not a fluke. The general version of this finding is called the spacing effect, and it is about as close to a law as learning research gets. A quantitative synthesis pooling 317 experiments found that practice spread across separate sessions reliably beats the same total amount of practice massed into one, across ages and materials2.
There is even an answer to how far apart the sessions should be. A study that tested 1,354 people over gaps as long as a year found that the best interval depends on how long you need to remember: the longer the horizon, the longer the ideal gap, and for anything meant to last months, gaps of days to weeks beat both cramming and very long pauses3. A child learning to type is learning for life — the far end of that horizon. Practising daily, or close to it, sits comfortably in the zone the data recommends.
Why practice that feels worse works better
The satisfaction result from the typing school generalises too, and it matters for parents more than any other finding here. Across motor and verbal learning, the conditions that make practice feel harder in the moment — spacing it out, varying it, holding back feedback — are the ones that reliably produce better retention and transfer later. Performance during practice turns out to be a poor measure of what is being learned4.
For typing, this means a session that felt clumsy was not a wasted session, and a fast Tuesday does not mean more was learned than on a slow Wednesday. It also means the daily words-per-minute number is the wrong thing to watch. What was learned shows up next week, not on today's meter.
The part everyone still gets wrong
None of this is new, which makes its absence from real timetables the strangest part of the story. Sixty years after the postmen — and a full century after the spacing effect was first measured — a review in the American Psychological Association's flagship journal asked why one of the most robust findings in learning research was still barely used in classrooms5. Its answer is the satisfaction paradox again: spaced practice loses the feeling contest. A long session ends with visible progress and a sense of accomplishment. Ten minutes ends just as the child warms up, and the improvement it caused is invisible until tomorrow. Schools, apps and parents all keep buying the feeling.
What this means at your kitchen table
The honest limits first: the postmen were adults, most spacing experiments use word lists rather than keyboards, and no study hands down an exact number of minutes for an eight-year-old. What the evidence supports is a direction, not a prescription — but the direction is unambiguous:
- Ten minutes a day beats an hour on Sunday. Same total time, measurably more learning — this is the single best-evidenced scheduling fact in the field.
- Stop while it still goes well. The improvement from a session is banked in the first minutes; the frustration is loaded at the end.
- Do not judge a session by its words-per-minute. Learning and today's performance are different things, and the meter measures the wrong one.
- Expect calendar time. Short daily practice needs more weeks to reach any given standard — that is the price of wasting fewer hours, not a sign it is failing.
- When your child sighs that they are getting nowhere, remember the postmen: the group that felt that way most strongly was learning fastest.
This is why every lesson on this site is one real sentence — something a child finishes in a few minutes, with a clear end. A session that ends by itself while things still go well is not a product limitation. As far as the evidence goes, it is the optimal dose.