Does handwriting still matter if everything is typed?

Children who learn a new letter by writing it end up knowing it better than children who learn the same letter by typing it. That has now been found with brain imaging [1], with accuracy tests on letters and words [2], and in a seven-week classroom study with 145 children [3].

What is not established is anything grander. Nobody has shown that handwriting makes a child cleverer, and nobody has shown that typing damages a child who can already read. The finding is narrow and it is about the period when a child is first learning what letters are.

In 2012 two researchers took 15 children, aged between 4 years 2 months and 5 years old, none of whom could read yet. Each child learned letters three ways: writing them freehand, tracing them over dotted outlines, and typing them on a keyboard. Then each child was shown those letters while a scanner recorded what their brain did [1].

The pattern that fluent readers show when they look at text, which involves the left fusiform gyrus and two other regions, appeared only for the letters the child had written by hand. Not for the traced ones. Not for the typed ones [1].

Fifty children in their last year of preschool in Spain, average age 5.4 years, were randomly assigned to one of four groups: copying by hand, tracing, typing with several different fonts, and typing with one font [2]. They had three sessions of 45 minutes on three days in a row.

The clever part is what they learned: nine letters from the Armenian and Georgian alphabets, which no Spanish five-year-old has ever seen, each given a sound, plus 16 made-up words built from them [2]. Nobody could arrive already knowing them, which is the flaw that makes studies of familiar letters hard to read.

Two rows deserve a second look. Picking a letter out of a lineup was easy for everybody, between 90 and 93% whatever they had done, so that task tells you nothing [2]. Writing a word from memory produced the widest gap on the table: 69.2% against 7.8%[2]. Doing something with your hand helps most when the test also asks your hand to do something, which is less mysterious than it first sounds.

In 2020 a German team ran the same comparison at proper length: 145 kindergarten children, average age 5 years 10 months, in three groups using a pencil, a tablet stylus or a keyboard, for 28 sessions of 25 minutes across seven weeks [3].

The pencil group gained more letter knowledge than the keyboard group [3]. So far, the same story. But the group differences in writing letters were not there, and at the word level the reliable differences were between the keyboard and the stylus rather than between the keyboard and the pencil[3].

One more caveat that cuts against the tidy version: those groups were matched rather than randomly assigned [3], while the 2025 study randomly assigned but was much smaller[2]. Neither study is the one you would design if you could have anything you wanted.

Read together, the honest summary is this. The advantage for learning letters by hand shows up in every study we could open. The advantage for words is real in one study and not clearly repeated in the larger one. Handwriting looks like a good bet during letter learning, not a magic ingredient.

Here is the answer almost nobody gives: nobody knows. We looked for a peer-reviewed study that establishes an age at which keyboarding should be taught, and did not find one. We also looked for a study showing that teaching a child to type improves their reading, and did not find one of those either.

What the three studies here tested is different: whether the keyboard is a good tool for the specific job of learning what a letter is, in the year or two before a child can read. The answer to that is no, or at least not as good as a pencil[1][2][3]. That says nothing about an eight-year-old who can already read and wants to write a story faster than their hand allows.

We took that seriously enough to remove something. A touch-typing stage was planned for the end of this app's path and was cut rather than shipped, because we could not point to an age threshold that would justify putting it in front of a particular child. It is easier to ship a feature than to admit you have no evidence for it.

It means an app should be honest about the one thing it cannot do. Plimmery can say a sound out loud, wait, and say it again without ever getting tired of a child who needs the twentieth repeat. It cannot watch a hand hold a pencil and it cannot tell whether a letter was formed from the right starting point.

So after a letter lesson it points at the paper: a prompt to go and draw the letter the child has just been working on, away from the screen. That is not a feature we are proud of building. It is the evidence on this page, applied honestly to a product that would rather keep you on the screen.

Nothing here replaces a teacher or a parent, and a tablet that nudges a child toward a pencil is still a tablet.