What a brain scan can and cannot tell you
80 min
Two hosts talk the lesson through. The voices are synthetic; the script was written from this lesson and checked against it, and asserts nothing the lesson does not.
- State what the EEG study measured and what it did not measure
- Give the three objections the published commentary raises, in its own words
- Say what a measurement of brain activity licenses about a claim on learning, and what it does not
Lesson 5 left the medium question with five answers and no verdict. This lesson is the same question asked with a completely different instrument, and it's the cleanest example in this course of a result that says less than it sounds like.
The study
In 2024, two researchers published an EEG study comparing handwriting with typing.1 EEG is electroencephalography: electrodes on the scalp reading the brain's electrical activity, 256 of them here. It's open access, it was widely reported, and you've probably met a version of it.
Their abstract, quoted in full, and it is the whole of what this course read of this paper apart from one sentence about the typing condition:1
As traditional handwriting is progressively being replaced by digital devices, it is essential to investigate the implications for the human brain. Brain electrical activity was recorded in 36 university students as they were handwriting visually presented words using a digital pen and typewriting the words on a keyboard. Connectivity analyses were performed on EEG data recorded with a 256-channel sensor array. When writing by hand, brain connectivity patterns were far more elaborate than when typewriting on a keyboard, as shown by widespread theta/alpha connectivity coherence patterns between network hubs and nodes in parietal and central brain regions. Existing literature indicates that connectivity patterns in these brain areas and at such frequencies are crucial for memory formation and for encoding new information and, therefore, are beneficial for learning. Our findings suggest that the spatiotemporal pattern from visual and proprioceptive information obtained through the precisely controlled hand movements when using a pen, contribute extensively to the brain's connectivity patterns that promote learning. We urge that children, from an early age, must be exposed to handwriting activities in school to establish the neuronal connectivity patterns that provide the brain with optimal conditions for learning.
Read it again and count what was measured.
Thirty-six university students. Single words, presented visually. A digital pen against a keyboard. Connectivity patterns from a 256-channel EEG array.
Now count what was not measured. Nobody learned anything, and nothing anybody remembered was tested. There is no recall measure in that abstract, no delay, no outcome at all.
Where the learning claim comes from
The abstract does make a claim about learning, and it's worth reading slowly to see where it enters.
It enters here: "Existing literature indicates that connectivity patterns in these brain areas and at such frequencies are crucial for memory formation and for encoding new information and, therefore, are beneficial for learning."1
That's a step from other people's work, not from this study's data. The study measured connectivity. The claim that this particular connectivity is good for learning comes from a literature the study cites and this course has not read.
So the structure of the argument is, and this is this course's reading rather than the paper's own description of itself, that we measured A, other work says A goes with B, therefore B.3 Drawn out, with the measured step separated from the asserted ones, it looks like this.
Only the top box was measured here. That's a legitimate thing for a discussion section to do, and it's a completely different kind of statement from a measured result. A reader who meets this study as "handwriting is better for learning" has been handed the conclusion of a syllogism as though it were a finding.3
Before you read on. If you wanted to know whether handwriting beats typing for learning, and you had this study's 36 participants for an afternoon, what would you add to the design?
Show the answer
The answer is short, and how short it is matters more than what it is.
A test. Have them write some of the words by hand and some by keyboard, wait, and then ask them which words they can recall. That's a study about learning, and the design change is small.
Notice what you wouldn't need. You wouldn't need the EEG at all. A recall measure answers the question the headline asks, and the connectivity measure answers a different question about what the brain is doing while the writing happens.
Which isn't a criticism of the study for being what it is. It was designed to look at connectivity and it looked at connectivity, on a 256-channel array, which is careful work. The gap is between what it measured and what it is quoted for, and the abstract's own last two sentences are where a reader would pick it up.3
The published objection
In 2025 the same journal published a commentary on it.2 This course read the four fragments quoted below, which make three objections, and nothing else of it.
You have the design now: thirty-six students, single words, a digital pen in one hand, a keyboard. Before you read the objections, write down the one you would raise.
Show the answer
There's no right answer to match, and the exercise is worth more if you write yours before you look.
What's worth comparing afterwards isn't whether you matched the commentary but where you aimed. The sample is the easiest thing to aim at, because thirty-six sounds small. None of the commentary's three objections is about the sample. All three are about the design of the comparison and about what the measure means, which is a different instinct and, for this kind of claim, the more useful one.
On the typing condition. The study's participants, in its own words, "only used their right index finger for typing to prevent undesired crossover effects between the two hemispheres".1 The commentary's objection, in its words: "participants were instructed to type using only their right index finger, making the typing condition quite different from typical typing".2 It adds that "removal of visual feedback decreases typing speed and impairs error monitoring processes".2
On what the measure shows. In the commentary's words: "the interpretation of increased theta/alpha connectivity as an unequivocal indicator of a brain state favorable to learning and remembering is problematic".2
Theta and alpha are bands of the brain's electrical rhythm, and the objection is about the first link in the abstract's own chain: from a pattern in those bands, to a brain state, to learning. The commentary's point is that the first link is not one-to-one, because the same pattern accompanies states that are not favourable to learning. That is the same step this lesson has already counted, arriving from somebody who studies the measure.
On what was reported. In its words: "only the difference between handwriting and typing is reported in the results, not connectivity patterns for each condition separately".2
Take the first objection seriously, because it is the one with a real answer on the other side. The study had a reason for the one-finger instruction and states it: to avoid crossover effects between the hemispheres. Typing with both hands moves both of them at once, and the comparison here is with one-handed handwriting, so a two-handed typing condition would have confounded the difference the study was trying to measure with a difference in how many hands were moving. That is a real cost of the alternative design rather than a convenience. And the commentary's point stands anyway: a conclusion about handwriting against typing needs a typing condition that resembles typing, and whatever the reason, this one doesn't.
Both things are true at once, and that's what a good methodological disagreement looks like. The design choice was made for a reason and it limits what the result can be used for.
The third objection is the subtlest
Only the difference was reported, not each condition separately.
Why that matters takes a sentence, and this is the course's reasoning from what the commentary says rather than the commentary's own elaboration.3 A difference between two conditions can be produced by either one moving. One condition could have been higher than the other because handwriting rose, or because typing fell, or because both moved in opposite directions by half as much.
Those three possibilities support three different stories, and the reported result can't tell them apart.
Two papers, one saying handwriting engages the brain more and one objecting to how that was shown. Which should you believe?
Show the answer
Neither, in the sense the question means, and the reason is the most useful thing in this lesson.
They aren't two answers to the question you care about. The study measured brain connectivity while people wrote single words. The commentary objects to how it measured it. Neither measured anybody learning anything, so neither is evidence about whether handwriting helps you learn.
What the study does establish, at the level its design supports, is that these two motor activities are accompanied by different patterns of brain activity. That is genuinely interesting and it is not what the headlines said.
What the commentary establishes is narrower still: that three specific features of the design limit what the result can be used for. It doesn't show the finding is wrong.
And the reason this lesson exists is that a brain measurement feels like harder evidence than a test score, and it's the other way round for this question. A test score is a direct measurement of the thing the claim is about. A connectivity pattern is a measurement of something else that other work associates with the thing the claim is about. The number of steps between the measurement and the claim is what to count, which is this course's own generalisation and not something it has measured, and imaging results seem to this course to have one more step than they appear to.3
The recommendation about children
The abstract's last sentence is a recommendation, and it is about children: "We urge that children, from an early age, must be exposed to handwriting activities in school to establish the neuronal connectivity patterns that provide the brain with optimal conditions for learning."1
That's theirs and this lesson gives it in their words, because it's part of what the paper says, and a course that quietly dropped it would be editing the source.
And this course has nothing to say about it. Every study it read was on adults. The EEG study's own participants were university students. Nothing read for this course bears on children at all, in either direction, and a lesson that used an adult finding to take a side in an argument about schools would be doing the thing lessons 1 to 5 were about.
The same count, outside the brain
The step count isn't really about brains. It works anywhere a number stands in for the thing somebody is claiming, so here's one with no scanner in it. This case is invented for the lesson rather than drawn from a source, because what's being shown is the arithmetic and not the case.
A school district spends a year on a new reading programme and reports a result: library check-outs are up forty per cent. The press release says literacy is rising.
Count. The district measured check-outs. Other work, which the release gestures at without naming, associates how much children read with how well they read. The claim is about literacy. That's two steps from the measurement to the claim, and the release has evidence for the first one, which is the shape of the EEG study's argument in a building with no electrodes in it.
The alternatives are easier to see here than they were in the brain, which is why this example is worth the detour. A check-out is a book leaving a shelf. It can rise because children are reading more, or because a teacher started requiring a book a week, or because the library extended its hours, or because the programme moved the popular shelves to where children walk past.
One measurement would close the gap between the check-outs and the claim. What is it?
Show the answer
A reading test, given to the same children at the start of the year and at the end of it.
That's the whole answer, and what's worth noticing is how ordinary it is. A report reaches for a number like check-outs because the number is already being collected, and that is also why a study built around a 256-channel array reports connectivity rather than a test score. It's what the equipment measures.
Neither report is dishonest. Each gives you a real number about something next to the question you asked, and the counting is left to you.
Three things people get wrong about brain-imaging claims
"It lights up, so it works." Activity is a measurement of activity. The claim is usually about an outcome, and the outcome is a different measurement.
"Imaging is more objective than a test." Both are measurements and both have designs. The test has the advantage of measuring the thing the claim is about.
"A published commentary means the paper was wrong." It means somebody raised objections in print, which is how the literature is supposed to work.
Practice
Take 20 minutes.
Find a claim that cites brain activity. An app, a headline, a training programme, a book jacket.
Write it as two columns.
- What was measured. The actual thing the study measured and reported, as precisely as the source states it.
- What is claimed. The sentence being sold.
Then count the steps between them, and name each one. A step is anything of the form "and X is associated with Y".
Then one line: how many of those steps does the source give you evidence for?
Usually the answer is the first one only, and the rest arrive as "existing literature indicates", which is the phrase to look for. That expectation is this course's own rather than something it has measured.3
Take 25 minutes.
Both papers in this lesson are open access. Read the study's abstract and the commentary's three objections, both linked below.
Then write two things.
- Which of the three objections you find strongest, and why, in three sentences.
- What the study's authors could say in reply to that one, at their strongest. Not a strawman: their actual best answer.
The second is the harder half. The study states a reason for one of the three and this lesson quotes it. For the other two you will have to construct the reply yourself, from what the paper is trying to do.
Connections
Back. Lesson 5 is the same question measured with tests rather than with electrodes, and the contrast between the two lessons is the point of having both. Lesson 2's denominator question is the ancestor of this lesson's step-counting: there, a count with nothing to divide by; here, a measurement with nothing to attach it to. Memory lesson 7 is the sample question, and thirty-six people is where it applies.
Forward. Lesson 7 is what to do with the notes you have, which is the constructive half of this course. Lesson 8 is where counting the steps becomes a habit.
Go deeper
- Handwriting but not typewriting leads to widespread brain connectivity: a high-density EEG study with implications for the classroom (Frontiers in Psychology, 2024). Open access. Read at abstract level by this course, plus one sentence about the typing condition. Read the abstract and notice where the word "learning" first appears.
- Commentary: Handwriting but not typewriting leads to widespread brain connectivity (Frontiers in Psychology, 2025). Open access. Read in part by this course: the three objections quoted above. Reading it beside the study is the exercise above, and the pair is the best free example this course found of a methodological disagreement conducted in public.
Sources
- F. R. Van der Weel and Audrey L. H. Van der Meer, "Handwriting but not typewriting leads to widespread brain connectivity: a high-density EEG study with implications for the classroom", Frontiers in Psychology 14, 2024, article 1219945. Open access. Read at abstract level, plus the one sentence about the typing condition, and the body says so. Supports: the abstract as quoted in full, including the thirty-six university students, the single words, the 256-channel array, the step through "existing literature", and the recommendation about children; and the quoted sentence about the right index finger.
- Adrien Pinet and Marieke Longcamp, "Commentary: Handwriting but not typewriting leads to widespread brain connectivity: a high-density EEG study with implications for the classroom", Frontiers in Psychology 15, 2025, article 1517235. Open access. Read in part: the four quoted fragments, which make three objections, and nothing else, which the body says. Supports: those four fragments and nothing beyond them. The commentary has no abstract, so there is none to quote.
- The reading of the abstract's learning claim as a step from other work rather than from this study's data is this course's own, said as such where it appears, and the syllogism it is set out as is the course's framing rather than anybody's description of the paper. So is the account of why reporting only a difference matters, labelled at the head of that section as reasoning from what the commentary says rather than as the commentary's own elaboration. So is the closing observation that the number of steps between a measurement and a claim is the thing to count, and that imaging results seem to this course to have one more than they appear to, both labelled where they appear. So is the judgement about where the gap between the measurement and the quotation opens, which is a judgement about the abstract rather than a report of it. The expectation that a source usually supports only the first step is an expectation rather than a measurement, and the exercise says so. Nothing in this lesson is a claim about children, and the section carrying the authors' recommendation says why.
Check your understanding
This lesson has a 6-question quiz. Pass it and the questions come back on a schedule in Review, so what you learned stays learned. Your progress is saved in your browser; no account needed.