New lesson
Why Typing Your Own Words Feels Slower Than Copying
A typing test hands you the words. Real writing makes you invent them. That single difference explains why a respectable test score can collapse the moment the text is yours — and why drilling more tests is usually the wrong fix.
Three tasks that share a keyboard
Copying, taking dictation, and composing look identical from across the room: hands on keys, words on screen. Research on typing tasks treats them as different jobs, because they are. Copying asks the eyes and hands to convert visible text into keystrokes. Dictation adds listening, holding sounds in memory, and converting sounds into letters. Composing — generating your own text — adds the heaviest load of all: deciding what to say, in what order, in what words, while also typing it. A study of university students performing copying and dictation tasks, discussed alongside earlier text-generation research, found that the intervals between key presses were shortest when students typed a well-known text from memory, longer when they copied, and much longer again when they generated the text themselves. The hands had not changed. The thinking stacked on top of the hands had.
Working memory is the bottleneck the study points to: composing requires more of it than copying does. Every unit of attention spent choosing the next sentence is attention not spent on the next keystroke, so the keystrokes slow down, pause, and bunch up around the hard decisions. That is not poor technique. It is a queue. Two jobs are sharing one channel, and the slower job — deciding what the words should be — sets the pace for both.
| Task | What the brain must do besides type | Typical effect on speed |
|---|---|---|
| Copying on-screen text | Read ahead, hold a few words, convert to keystrokes | Closest to your test score |
| Copying from paper | All of the above, plus gaze switching and place-keeping | Noticeably slower; place-losing adds pauses |
| Dictation | Listen, hold speech in memory, convert sounds to spelling | Varies; spelling decisions add load |
| Composing your own text | Decide content, order, and wording while typing | Slowest of the four; pauses cluster at decisions |
Why the test score misleads you
A typing test is a copying task with the reading made as easy as possible: short, familiar words, presented a line at a time, scored the instant you finish. It measures the fastest of your four jobs under the friendliest conditions. Comparing that number with your composing speed is comparing two different scores — and typing scores are not interchangeable even within one task. Gross words per minute, net words per minute after errors, and accuracy-adjusted speed (gross speed multiplied by accuracy) give three different numbers for the same performance. In a 2026 study of 10,613 U.S. adults, the median accuracy-adjusted speed on a computer was about 37.8 words per minute — a reminder that everyday, accuracy-counted typing is far slower than headline test scores suggest, before composing load is added at all.
There is a second trap in the paper direction. Community typing advice collected from learners makes the point from the other side: a typist scoring 50 words per minute in a familiar online test slowed down sharply when copying from a printed page, because the eyes must switch between page and screen, hold a chunk of words in memory, and find their place again after every switch. If even copying changes speed when the source changes, composing — where the source is a half-formed thought — was never going to match the test.
Measure both speeds properly: a one-sitting protocol
- Take a standard copy test. Use your usual test, at its usual length. Record gross speed and accuracy, then compute the accuracy-adjusted score (gross multiplied by accuracy). Write down which scoring the test reports — never mix gross from one test with adjusted from another.
- Rest two minutes. Stand up, look away from the screen. The compose test must not inherit copy-test fatigue.
- Compose for the same duration. Set a timer for the same length as the copy test and write about a familiar topic — a description of your morning routine works — without stopping to edit. Count the words produced and divide by the minutes. That is your raw composing rate, thinking pauses included.
- Mark the pauses. Read the composed text and note where you remember hesitating: at the start of paragraphs, mid-sentence, or on particular words. The location of the pauses is the diagnosis, as the next section shows.
- Repeat on another day. One sitting is an anecdote. Two sittings with the same gap are a measurement.
What success looks like: you finish with two honest numbers — a copy score and a compose rate, measured the same way on two different days — and a pause map. Most learners find the compose rate is a third to a half below the copy score. That gap is normal physiology, not failure, and now it has a size you can work on.
Read the pauses: a decision framework
Pauses at sentence and paragraph starts mean the bottleneck is planning, not typing. The fix is structural, not mechanical: outline before drafting, even three rough bullet points, so the composing brain makes its big decisions before the hands start. Planning load moved out of the typing window stops taxing keystrokes.
Pauses mid-sentence, on ordinary words point to spelling or word-finding friction. The remedy is volume of real writing, not word-list drills: composing is trained by composing, in short daily stretches on easy topics where the wording comes cheaply, so the word-finding pathway gets its repetitions.
No long pauses, but a slow, even pace suggests the hands genuinely are the limit even under load — the one case where classic speed and accuracy drills are the right prescription. Fast bursts separated by long stalls usually mean editing-while-drafting: the writer is composing, judging, and deleting in rotation. Separating the passes — draft first with the delete key rationed, edit afterwards — removes a third job from the channel.
Training that respects the diagnosis
Whichever cause the pause map names, two habits help every composer-typist. First, practise the task you want to improve: community guidance for the paper-copying slowdown says exactly this — measure and train the target task itself rather than expecting an unrelated test to carry over — and the principle applies unchanged to composing. Ten minutes of daily free-writing on an easy topic trains the composing pathway in a way no copy test does. Second, protect accuracy while you compose. Errors made at speed must be found and fixed later, and the correction pass costs more time than the errors saved. A slightly slower draft that needs light editing beats a fast draft that needs surgery, every working day of the year.
Questions learners ask
Is a big copy-versus-compose gap a sign of bad technique? No. The research pattern runs the other way: key-press intervals lengthen under generation load for typists generally, because the task is heavier. Technique problems show up in the copy score itself.
Should a job application quote the test score or the compose rate? The test score, honestly labelled — that is the number employers' own tests will measure. Just do not budget your own writing deadlines around it.
Does composing speed ever catch up to copy speed? It narrows with writing practice, especially once outlining becomes habitual, but the planning load never fully disappears. Expect a durable gap, and plan deadlines with the compose rate, not the test score.