Is Writing Notes by Hand Better Than Typing?

The evidence is weaker than either famous version. A 2014 study found students taking notes on laptops performed worse on conceptual questions than those writing longhand, and it was adopted as justification for classroom laptop bans almost immediately. A 2019 direct replication found performance did not consistently differ between any groups, including a group that took no notes at all.

Since then a brain-imaging study has taken over the headlines, finding widespread brain connectivity during handwriting and not during typing. It drew a published commentary in the same journal disputing what can be inferred from it, and one detail of its design is hard to get past.

What the original study found

Mueller and Oppenheimer’s 2014 paper in Psychological Science argued that even when laptops are used solely to take notes, they may still impair learning because their use results in shallower processing. Across three studies, students taking notes on laptops performed worse on conceptual questions than students taking notes longhand.

The proposed mechanism is the part worth keeping: taking more notes can be beneficial, but laptop note takers’ tendency to transcribe lectures verbatim rather than processing information and reframing it in their own words is detrimental to learning.

That is a claim about transcription versus reformulation, not about pens.

What the replication found

Morehead, Dunlosky and Rawson published a direct replication and extension in Educational Psychology Review in 2019, describing the original as a high-profile investigation which concluded that taking notes by longhand is superior.

They replicated it directly and added conditions the original did not have: a group using electronic writing tablets, and a group that took no notes at all.

Their results: some trends suggested longhand superiority, but performance did not consistently differ between any groups across two experiments, including the group that took no notes in the second experiment. Group differences were further decreased after students studied their notes. A meta-analysis combining direct replications revealed small, non-significant effects favouring longhand.

The no-notes group is the finding. If people who wrote nothing performed comparably to both the longhand and laptop groups, the experiment did not detect much benefit from note-taking at all, let alone from its medium.

The brain study that replaced it

A 2024 study in Frontiers in Psychology by Van der Weel and Van der Meer recorded brain electrical activity in 36 university students as they handwrote visually presented words using a digital pen and typewrote the same words on a keyboard, with connectivity analyses performed on data from a 256-channel sensor array.

Handwriting produced widespread connectivity. Typing did not.

The authors’ inference, quoted in the commentary discussed below: as increased connectivity in the brain was observed only when writing by hand and not when simply pressing keys on the keyboard, their findings can be taken as evidence that handwriting promotes learning.

That sentence is doing a lot of work, and it is where the dispute sits.

The published commentary

Pinet and Longcamp published a commentary in the same journal in January 2025, characterising the claim as being that unlike typing, handwriting generates brain connectivity patterns that promote learning.

As a 2026 paper describing their critique reports, the objections are specific: the 2024 study did not directly involve learning or memory retrieval; the typing condition used an atypical input method, a single right index finger, rather than ecologically valid expert bimanual typing; visual feedback differed systematically between conditions; and typing skill was not controlled. We are citing those four points as a later paper summarises them rather than from the commentary directly, so treat them as less firm than the fact of the commentary itself.

The first and second objections are the ones a reader can weigh without expertise. A study of learning that does not test learning is measuring a proxy. And a comparison between handwriting and one-finger typing is not a comparison between handwriting and typing as anyone does it at work.

What survives

Three things, and none is a rule about pens.

The transcription mechanism remains the most plausible account of any behavioural effect. The original’s argument was that verbatim transcription without reformulation is what hurts, and nothing in the replication contradicts that reasoning; it failed to find the downstream difference. If the mechanism is real, the actionable variable is whether you are rewriting ideas in your own words, which you can do or fail to do on either device. That is the same variable at work in cognitive offloading.

The behavioural meta-analytic direction is weakly towards longhand. Small and non-significant is not zero.

And the brain-imaging finding is real as a measurement of brain activity. What is disputed is the step from connectivity to learning, which the commentary challenges and which the study did not test.

What does not survive is the confident version: that typing your notes measurably damages your understanding and switching to paper will fix it.

Common questions

Should I switch to paper notes?

Nothing here requires it. The 2014 study found laptop note takers performed worse on conceptual questions across three studies. A 2019 direct replication with added conditions found performance did not consistently differ between any groups across two experiments, and a meta-analysis combining direct replications found small, non-significant effects favouring longhand. A 2024 EEG study of 36 students found widespread brain connectivity during handwriting and not typing, but a published commentary in the same journal disputes the inference from connectivity to learning, and the study did not test learning or memory retrieval. If you prefer paper, keep using it. If you prefer typing, this literature does not give you a reason to change.

Does the brain study prove handwriting is better for learning?

Its authors argue so and a published commentary disagrees. The 2024 study recorded brain activity in 36 university students handwriting words with a digital pen and typing the same words, using a 256-channel sensor array, and found widespread connectivity only during handwriting. The authors concluded this can be taken as evidence that handwriting promotes learning. Pinet and Longcamp published a commentary in the same journal in January 2025 challenging that claim. As a later paper reports their objections, the study did not directly involve learning or memory retrieval, the typing condition used a single right index finger rather than normal two-handed typing, visual feedback differed between conditions, and typing skill was not controlled.

What was the mechanism supposed to be?

Transcription rather than the medium. The original researchers proposed that while taking more notes can be beneficial, laptop note takers tend to transcribe lectures verbatim rather than processing information and reframing it in their own words, and that this is detrimental to learning. Their framing was that laptops may impair learning through shallower processing even when used solely for notes. The replication did not test or refute that mechanism; it failed to find the performance difference the mechanism was invoked to explain. If the mechanism is right, the useful variable is whether you are rewriting ideas in your own words, which is possible or avoidable on either device.

Does taking notes help at all?

The replication raises a real question, though it was not designed to answer it. Its second experiment included a group who took no notes, and performance did not consistently differ between any groups, including that one. It also found that group differences decreased further after students studied their notes. Both results point away from the act of note-taking during a lecture being where the benefit lies, and towards what happens afterwards. These were studies of students, lectures and test performance, so mapping them onto meeting notes or technical documentation is an extension rather than a finding.

The variable was never the pen

What makes this a useful story is how cleanly the original mechanism and the popular version came apart, and how the same thing then happened again.

The 2014 researchers proposed that transcribing verbatim without reformulating is what costs you. That is a claim about a cognitive operation. What spread was a claim about hardware, because hardware is easier to legislate: you can ban laptops, and you cannot ban shallow processing. The direct replication then found performance did not consistently differ between groups, including people who wrote nothing at all.

The brain-imaging study repeated the pattern in a new register. It measured connectivity, not learning, and the headline became a claim about learning. Its typing condition, according to the published commentary, used one finger.

Which leaves the same takeaway twice over. If you want the version you can act on, it is not a stationery purchase. It is whether, at some point, you put the idea into your own words, which is also what determines whether reading on a screen or media multitasking costs you anything.