The more accurate instrument found less
A Finnish team followed 260 children for eight years, measuring their movement two ways: a chest monitor that recorded every heartbeat, and a questionnaire that asked them what they’d been doing. The monitor was the objective one. It was the asking that turned something up.
If you wanted to know how much someone moved over eight years, you’d want a device. Something that doesn’t forget, doesn’t flatter, doesn’t guess. A questionnaire is a poor instrument by comparison — it depends on memory, on a parent’s estimate, on whether anyone was paying attention that week.
A study published this month used both. And the results didn’t agree.
What they did
Petri Jalanko and colleagues at the University of Jyväskylä, writing in Pediatric Exercise Science, drew on the long-running PANIC study in Finland. They followed 260 young people who had joined the study aged six to nine, assessing them at the start, again two years in, and again eight years in — by which point they were around fifteen to seventeen.
Most previous work here had taken a single snapshot. The point was to watch the same people as their habits shifted, because habits change a great deal between six and sixteen.
At each assessment, they measured movement twice over. Participants and their parents filled in questionnaires covering organised sports, unsupervised physical activity, screen time, and non-screen sedentary time — reading, crafts, that sort of thing. And each young person wore a chest monitor for at least four consecutive days, recording heart rate and body movement to capture light, moderate and vigorous activity, and total time sitting still.
At the final visit, everyone sat a computerised cognitive battery: reaction times and accuracy across tasks measuring psychomotor speed, attention, working memory and associative learning.
What they found
The chest monitors found essentially nothing. Across the group as a whole, eight years of accumulated device-measured activity and sitting showed no broad association with how the adolescents performed. The physiological intensity of movement — the part you’d think would matter most — didn’t track.
The questionnaires found several things, all of them small, and some of them backwards from what you’d expect.
More reported screen time across the eight years went along with faster reaction times on working memory tasks, and slightly better overall cognitive scores. More non-screen sitting still — books, crafts — went with faster reaction times too, though with poorer accuracy. And more unsupervised physical activity went along with poorer working memory accuracy, a link the authors describe as statistically weak and in need of replication.
Looked at by sex, a few patterns separated. Among boys, more organised sport went with better working memory, while more unsupervised activity went with worse. Among girls, more reported screen time went with better overall cognition, and device-measured light activity went with better working memory — neither of which showed up in the boys.
Where the honesty sits
The researchers are careful, and it’s worth repeating their care rather than the headline.
Cognition was only tested at the end, so nothing here establishes cause. The arrow may well point the other way: young people with stronger cognitive skills might simply gravitate towards screen-based tasks and away from unstructured play. The effects were small — real enough to describe a population, not large enough to say anything about a given child. And the researchers are explicit that none of this recommends unlimited or purely passive screen use.
Both instruments have holes in them, too. Questionnaires lean on memory, and parents routinely misjudge how their children spend an afternoon. Chest monitors don’t know the difference between sitting down with a novel and sitting down with a film.
Which is, we think, the interesting part.
What we keep thinking about
The device was the better instrument. It was also the one that came back empty.
Not because it was wrong. Because of what it was built to record. A chest monitor captures intensity and duration — how hard, how long, how still. It cannot capture what you were doing. The questionnaire, clumsy and forgetful as it is, asked a different question: not how much you moved, but what the hours were made of. Organised sport or roaming about. A screen with homework on it or a screen with nothing on it.
And the thing that showed up lived in that distinction, not in the numbers.
“Screen time” is the clearest case. It’s a category built around a piece of glass, and it groups a maths lesson with a strategy game with a video call to a friend with an hour of scrolling nobody enjoyed. The authors say plainly that some of what sits in that box is likely fine and some likely isn’t, and that lumping them together may be why the research keeps coming back mixed. A measure that broad can be perfectly accurate and still not tell you much.
We think about this a lot, because measuring is the part of wellbeing that’s easiest to get slightly wrong. It’s tempting to build something that counts — hours, streaks, scores, a number that goes up. Counting is satisfying. It’s also, quite often, counting the wrong thing precisely.
That’s the honest reason Nagi’s check-in is four taps rather than a test. Mood, sleep, energy — noted as the days pass, not scored, not graded, nothing to be good at. It isn’t trying to be the more accurate instrument. It’s trying to be the one that sits close enough to your actual day to still mean something a month later, when you look back and remember what that week was. The S8 wellbeing measure Nagi uses is in a formal validation study at the moment, because we’d rather find out how it performs than assert it. And none of it goes anywhere: no name, no email, an anonymous ID rather than a name. A calm space that’s just yours.
Noticing doesn’t fix anything, and we won’t pretend otherwise.
The bit we’d underline
Two instruments. Same eight years, same 260 people. The precise one found nothing; the fuzzy one found something small.
There’s a version of this in most people’s lives. You can know exactly how many hours you slept and have no idea why the week felt like that. You can have a full record and no account of it.
Whatever you use to keep track of yourself — an app, a notebook, a vague sense of how things have been going — it’s worth knowing what it’s built to see, and what it isn’t. Not so you can find a better one. Just so you don’t mistake the reading for the whole picture.
If you want a private place to start noticing your own patterns, Nagi is on Google Play and the App Store — or in your browser at app.nagiandme.app. The first 15 messages are free. No name, no email. Just you, checking in.
— The Nagi team
Source: Long-term study links childhood screen time to slightly better cognitive skills in teenagers, PsyPost, 17 September 2026, by Eric W. Dolan. Underlying study: Jalanko, P., Leppänen, M. H., Bond, B., Laukkanen, J. A., Lakka, T. A., & Haapala, E. A., “Associations of Physical Activity and Sedentary Time From Childhood to Adolescence With Cognition in Adolescence: The PANIC Study”, Pediatric Exercise Science. Design: observational longitudinal cohort drawn from the Physical Activity and Nutrition in Children (PANIC) study in Finland, led from the University of Jyväskylä. Sample: 260 adolescents, recruited aged 6–9 and assessed at baseline, at two-year follow-up and at eight-year follow-up, by which point participants were approximately 15–17. Movement was measured two ways at each assessment: self- and parent-reported questionnaires covering organised sports, unsupervised physical activity, screen time and non-screen sedentary time; and a chest-worn heart rate and movement monitor worn for a minimum of four consecutive days, capturing light, moderate and vigorous physical activity and total sedentary time. Cognition was assessed once, at the eight-year follow-up, using a computerised battery measuring reaction time and accuracy across psychomotor function, attention, working memory and associative learning. Findings: cumulative device-assessed physical activity and sedentary time showed no broad association with cognitive performance in the group overall. In the self-reported data, higher cumulative screen time was associated with faster reaction times on working memory tasks and better overall cognitive scores; higher non-screen sedentary time was associated with faster reaction times but poorer working memory accuracy; and higher unsupervised physical activity was associated with poorer working memory accuracy, an association the authors describe as statistically weak and requiring replication. By sex: among boys, more reported organised sport was associated with better working memory and more unsupervised physical activity with poorer working memory; among girls, more reported screen time was associated with better overall cognition, and device-measured light physical activity with better working memory — neither pattern present in boys. Stated limitations: cognition was tested only at the final assessment, so causal direction cannot be established and reverse causation is plausible (adolescents with stronger cognitive skills may choose more screen-based and fewer unsupervised-play activities); observed associations were small and meaningful at population rather than individual level; “screen time” is a heterogeneous category spanning schoolwork, cognitively engaging use and passive scrolling, which may relate to outcomes differently; questionnaires rely on recall and on parental estimates that may misjudge unsupervised play or screen use; chest-worn monitors cannot distinguish between sedentary activities such as reading and watching a film. The authors state the findings are not a recommendation for unlimited or purely entertainment-focused screen use, and note that a randomised cross-over study (VERNA) is underway to test causal questions more directly.
Nagi is a mental health companion, not a medical device, and nothing here is medical advice. If you’re struggling, please reach out to a professional or someone you trust.