Handwriting or typing?

The study everyone cites

In 2014 Mueller and Oppenheimer published The Pen Is Mightier Than the Keyboard. US college students who took lecture notes by hand scored higher on conceptual questions than students typing on laptops, even though the typists wrote more words HND-01. The finding was irresistible: it was tidy, it confirmed something people already suspected, and it produced a headline that has been repeated more or less continuously since.

Two things are worth knowing about it. The first is what it actually measured — university students taking notes from a recorded lecture in a laboratory. Not children. Not composition. Not an examination. The second is that it did not hold.

The replications that failed

In 2019 a pre-registered replication and extension tightened the design, adding an eWriter condition and a no-notes control. Longhand note-takers did not reliably outperform laptop note-takers on conceptual questions. Its authors concluded that deciding which method is superior seems premature HND-02.

In 2021 a direct replication of the original Study 1 — 142 participants, 87 undergraduate co-authors — reproduced the surface result that laptop users transcribe more verbatim content, and then found no significant advantage for longhand on the quiz at all, including on the conceptual questions that were the whole point HND-03.

This is ordinary science working as intended, and it is the part that rarely travels. A practice that teaches students to check what a source actually says should say so plainly: the most-cited evidence for handwriting is the weakest evidence for handwriting.

The newer study everyone cites

Since 2024 the headline has had a successor, and it is the one a parent is most likely to raise. Researchers recorded high-density EEG from 36 Norwegian university students and found that writing words with a digital pen produced “far more elaborate” brain connectivity than typing the same words — 32 significant connectivity differences across parietal and central regions HND-04. It ran in Scientific American, on NPR, and through the education press as evidence that handwriting is better for learning.

A peer-reviewed commentary published the following year makes an objection that settles what the study can be used for. The participants never learned anything: “the protocol did not include any learning. Participants repeatedly wrote well-known words without any requirement for memory encoding.” Nobody was taught, and nobody was tested. The French researchers writing the commentary also note that the Norwegian study’s participants were told to type “using only their right index finger”, that there was an “absence of visual feedback on the screen” while typing but not while handwriting, and that typing skill was never assessed HND-04.

Be careful what that does and does not establish, because the temptation is to over-claim in the other direction. The connectivity differences may well be real; the commentary does not dispute them, and no lab has tried and failed to replicate the paradigm. What it disputes is the inference everyone drew. The honest statement is that handwriting and one-finger typing produce different patterns of brain activity while copying familiar words, which is genuinely interesting and is not a finding about learning to write.

The same shape twice in a decade is the lesson worth keeping. Both times a tidy result travelled further than its design could carry it, and both times the correction arrived quietly years later. The evidence that actually supports teaching handwriting is listed below, and none of it is either of these studies.

Why it matters in Victoria anyway

Whatever the laboratory evidence says about note-taking, the question is settled for a Victorian student by the rules of the assessment. VCAA prohibits laptops and computers as general-use materials in standard VCE written examinations; a device may be brought in only under an approved Special Examination Arrangement HND-11. VCE English is written by hand, for three hours, on the day.

That has a straightforward consequence. A student who drafts everything on a laptop and has never written continuous prose by hand for three hours is not under-prepared in argument — they are under-prepared in the physical act of the exam. Handwriting stamina is not a virtue here; it is a requirement of the format. The VCE English exam overview sets out what the three sections ask.

Earlier in school the mode shows up in the marks. In NSW research, after controlling for task, student and school-level factors, primary students in Years 2 to 5 scored consistently lower on computer-based writing tests than on paper — across every year examined, for both narrative and persuasive genres, by roughly 15 to 20 scaled points HND-10. NAPLAN writing is now online — but not at every level. ACARA sits the writing test “with a keyboard in Years 5, 7 and 9, and in handwriting in Year 3” HND-14, so a child hand-writes their first NAPLAN and types every one after it. The medium changes at precisely the point the NSW research found it mattered most, and nothing in ACARA’s framework addresses whether the scaling between year levels accounts for that. It is not a reason to distrust a result; it is a reason not to read a Year 3 and a Year 5 writing score as the same measurement taken twice.

One finding worth setting against the others, because it points somewhere more useful than “paper is better”. Before the rollout, ACARA’s own device study reported that NAPLAN can be sat on laptops or tablets “without device effects across content domains, item types and year levels”, and that the key factor in how students interact with online items is familiarity with the device they are using LAN-15. It is vendor work commissioned before the rollout it supported, and it compares screens with screens rather than screens with paper, so it does not answer the NSW finding above. But the two agree on the part a parent can act on: what matters is whether the child is fluent in the medium they will actually be writing in. What NAPLAN is and what a result can tell you covers how those results are reported, and the shorter answer for a parent deciding what to do about it sets out the study’s limits and the practical split by year level.

One more finding is worth holding lightly, because it is about markers rather than students. A 2024 review by Cambridge Assessment, the UK exam board, found no consistent overall pattern of typed and handwritten scripts scoring differently — but did find that markers tend to award handwritten versions of the same essay higher scores than the typed version HND-12. That is a fact about human marking, not about writing.

What to actually do

The best-evidenced claim on this page is also the least fashionable: teaching handwriting improves writing. The meta-analysis puts composition quality at an effect size of 0.84 and the amount students write at 1.33 HND-07, and handwriting fluency alone accounts for 7.4 per cent of the variance in primary writing quality, in Norwegian data HND-08. The mechanism is unglamorous — a student who has to think about forming letters has less attention left for the sentence.

So: for a young child, handwriting practice is worth the time, and the evidence for that does not depend on the 2014 study at all. For a VCE student, write by hand under time, because that is the examination. For everyone in between, the honest answer is that typing is not harming your child, and the question is smaller than the internet suggests.

Where that advice does not apply

Everything above is about students for whom writing by hand is merely a skill to build. For some students it is the obstacle itself, and every study cited on this page was conducted on typically developing learners — which is worth saying plainly rather than leaving a parent to assume the advice covers their child.

Australia’s national education research body reviewed the writing literature in 2022 and reported that, for students with learning disabilities, dictation to a scribe or technology used to circumvent handwriting and spelling difficulties improved writing quality at an effect size of 0.55 — that figure and the others in this section come from United States meta-analyses the Australian review is summarising HND-15. The same review reports that students who compose primarily by word processor tend to convey ideas better than those composing primarily by hand (0.50), and recommends students be taught typing alongside handwriting rather than instead of it.

The caveat belongs in the same breath, because the review puts it there. Dictation being effective “does not negate the need for explicit and systematic handwriting and spelling instruction” for those students, in that same Australian review of United States evidence HND-15. Quoting either half alone misrepresents it: this is not permission to abandon handwriting, and it is not a reason to withhold a keyboard from a child who needs one.

Two limits on how far this goes. The review uses “learning difficulties and disabilities” as an umbrella and gives no separate figure for dyslexia, and its underlying studies span primary and secondary rather than secondary alone. If handwriting is the thing standing between your child and the sentence they meant to write, that is a conversation with the school and, where relevant, an assessment — not something this page can settle.