HomeResearch
What the evidence actually says
One well-known study found that a phone within reach cost people attention even when they never touched it — face-down and switched off. A later replication did not find the same effect, so we treat it as a good question rather than a settled answer. Spark doesn’t move your phone; it moves the unlock.
None of it is ours. These are independent studies by other people, published in journals and conference proceedings you can read for yourself. We have not run a study, commissioned a survey or measured a single Spark user — so there is no figure on this page that we are the source of, and every claim below is followed to its author.
Every link below last followed to its source and re-checked on .
Four studies, all of them somebody else’s
Each card states the finding in the authors’ own terms and links the paper. If a card and the paper disagree, the paper is right and the card is a bug.
before interrupted work was picked up again
Researchers observed and timed 24 information workers — managers, analysts and developers — for more than 700 hours. When interrupted work was resumed the same day, it took on average about 25 minutes to get back to it, with more than two other activities in between.
Mark, Gonzalez & Harris — “No Task Left Behind? Examining the Nature of Fragmented Work” (CHI 2005) ↗beat a phone on the desk in one well-known study — a later replication didn't find it
Participants who left their phone in a different room did better on working-memory and fluid-intelligence tasks than those whose phone sat on the desk, even when it was switched off. A later pre-registered replication of the experiment found no difference, so treat this as one result, not a settled fact.
Ward, Duke, Gneezy & Bos — “Brain Drain” (Journal of the Association for Consumer Research, 2017) ↗when a phone is simply visible on the table
Pairs of strangers who talked with a phone in view reported lower relationship quality, less closeness and less trust than pairs who talked with no phone present. Nobody had to touch it.
Przybylski & Weinstein — Journal of Social and Personal Relationships (2013) ↗after light-emitting screens in the evening
Reading on a light-emitting device before bed suppressed melatonin, pushed the circadian clock later, delayed sleep onset and reduced next-morning alertness compared with print.
Chang, Aeschbach, Duffy & Czeisler — PNAS (2015) ↗“In another room”, and what it does not prove
The Brain Drain study shaped how we think about Spark. Its participants did not have to use their phone for it to cost them: in that experiment, having it on the desk, face-down and powered off, went with lower scores on working-memory tasks than leaving it in another room. It is a well-known result and not a settled one — a pre-registered direct replication did not find the effect, so we hold it as a reason to try distance, not as proof that distance works.
It also does not describe what Spark does. Spark doesn’t move your phone; it moves the unlock. Your phone stays with you and calls always get through; the key is what goes in another room, so getting your apps back means getting up, walking to it and holding your phone against it. Locking is one tap in the app and never needs the key. Whether that walk helps is exactly what nobody has measured yet. How that works in practice.
How much India is on its phone
Three pages of other people’s numbers: who is on a phone and for how long, what the word “addiction” does and does not mean here, and what the courts and the law have actually said about phones in schools.
Phone use in India, state by state
Phone and internet use in every state and union territory, by age, from the government's own 2025 survey.
See the mapThe dataPhone addiction in India: what the data shows
The hours, what teenagers do with a phone, and why there is no addiction rate to quote — for phones, unlike gaming, there is nothing clinical to count.
Read the figuresLaw and policyPhones in Indian schools
The Delhi High Court called a complete ban “undesirable and unworkable” and set out nine principles instead. Quoted in full.
Read the principlesWhat this evidence does not say
Cited research is easy to over-read. These are the specific things the papers above do not establish.
- None of them tested Spark. Nobody has. They describe the problem Spark is aimed at; they are not evidence that our answer to it works.
- The best-known one has not held up everywhere. A pre-registered direct replication of the Brain Drain experiment, using the same tasks and conditions, did not find the effect: Ruiz Pardo & Minda — “Reexamining the ‘brain drain’ effect: A replication of Ward et al. (2017)” (Acta Psychologica, 2022). One study for and one against is not a verdict either way.
- An average is not a prediction.Each figure describes what a group of people did under one study’s conditions. It is not a law of attention, and it is not a forecast of your afternoon.
- These are modest samples under specific conditions. An effect that shows up across a group is not a promise about any one person on any one day, and none of these studies followed anyone for months.
- Nothing here measures a phone-free habit over time. They measure what a nearby phone costs in the moment. Whether that cost compounds, and by how much, is not something we can source.
The row we cannot fill in
Every figure above is somebody else’s, published, and each one is cited so you can go and read it. We have none of our own: nothing has shipped, so nothing has been measured, and a number invented to fill this row would be worth less than the empty space it replaced.
When there is a study of actual Spark users it will appear here with its method, its sample size and its limitations stated in the same breath as its result. Until then this row stays blank on purpose.
Getting the most out of it
Usage advice, not outcome claims — none of these says how much time anyone saved, because we do not know. The app is free and the key is a one-time purchase, so the only thing you are spending here is attention on where you put it. See what comes in the box.
Put the key where the friction is useful
Not the next desk — the next room. Far enough that fetching it is a decision, close enough that a genuine emergency is never a crisis.
Lock before you need to, not after
The moment to lock is when you sit down, not when you notice you have already lost twenty minutes.
Start with three apps, not thirty
A mode you actually use beats an exhaustive one you abandon. Add the rest once the habit holds.
Pair it with something you already do
Lock when the kettle goes on. Lock when you sit down to eat. Attach it to a trigger that already exists.
Where the argument goes next
Why a physical key
The case for putting the unlock outside the phone, and the honest reasons you might not want to.
Read the caseMechanicsHow Spark works
Modes, the one-tap lock, the hold-to-unlock key, and what each operating system actually permits.
See the loopBackgroundWhy we built it
The kitchen drawer that worked better than every app we tried, and what we changed to keep it.
Read the storyDistance you don’t have to maintain.
Spark doesn’t move your phone; it moves the unlock. Lock in one tap and leave the key in another room, and the decision gets made once, in the morning, instead of being a choice you have to keep winning all day.