The phone could still ring.
That detail matters in a 2025 experiment by Noah Castelo and colleagues. Participants were asked to block mobile internet on their phones for two weeks. Calls and texts remained available, and they could use the internet on other devices. This was a change in access, not exile from the human world.[1]
The randomized trial reported improved well-being and a small improvement in performance on a five-minute sustained-attention task. It also showed how difficult the intervention was to maintain: of 467 participants, 266 set up the block and 119 met the study’s compliance rule. Only 313 completed all three surveys. The primary analyses retained noncompliers but required completed measures.
These motivated American and Canadian iPhone users cannot stand for everyone, and a short test is not a whole life. Still, the study offers a reason to take a change in the environment seriously. People need not be permanently stuck with their current pattern of connection.
We should carry both parts of the result into practice: there is room to improve, and making the change work is part of the problem.
A ban removes more than a temptation
Imagine that you decide to keep a particular application closed for a week. You use it too often without intending to, and you want to find out what happens when it is less available.
On the first evening, you discover that a friend uses the application to arrange a visit. You also remember that a group you care about posts its meeting details there. The service had been providing more than a stream to scroll. It was part noticeboard, part address book, part place where people happened to be.
That does not mean the experiment must end. It means you need to preserve the functions that matter. Ask the friend to contact you another way. Note where the meeting details can be found. Decide whether a brief planned visit to the group would serve the purpose better than a complete block.
Without these arrangements, the experiment can confuse two changes: less unwanted consumption and less access to people. If you feel worse, you will not know which change did the damage.
The same issue arises with a phone-wide restriction. A device may hold travel information, work access, payment functions, or tools needed for daily tasks. A broad block can be useful for someone whose arrangements allow it. For another person, it can create practical trouble that has little to do with a weakness for distraction.
A rule should be designed around the activity you are trying to change. If the problem is repeatedly opening short videos between paragraphs, there may be no reason to remove the map that gets you home. If the problem is a work channel carrying routine questions into the evening, the repair may belong in the team’s expectations rather than in every application on the device.
We can learn from strong interventions without copying their exact form into every life. A trial tells us what happened under its conditions. Your task is to find a change that can survive yours.
This requires a little honesty about what you would miss. Name the benefit before removing the service. It is easier to protect a valued relationship when you acknowledge that it exists than when you declare the whole platform worthless and later creep back to find your friends.
You are allowed to like something that you need to use differently.
Less use is not the whole result
A major Facebook experiment offers a useful complication. Hunt Allcott and colleagues studied paid deactivation for four weeks around the 2018 U.S. midterm election. Deactivation improved an overall measure of subjective well-being, but reduced factual news knowledge. More time went to offline activities, including both socializing with friends or family and watching television alone.[2]
Freed time did not automatically become the same activity for every person. Nor did one improvement cancel every loss.
Imagine two people who each recover an hour from a platform. One uses it to prepare for a class. The other spends it watching a show they genuinely enjoy. Both may be pleased. A third person switches to another feed and feels little change. A fourth misses a community that had helped them through a difficult period.
The screen-time reduction is identical. The lives are not.
This is why a sensible experiment needs an aim beyond making a number smaller. Perhaps you want to finish dinner without checking for updates. Perhaps you want fewer unwanted interruptions during reading. Perhaps you want to stop arriving late because an intended five-minute visit has expanded. The aim gives the result a meaning.
A person who uses a screen to speak to a distant parent may rightly spend more time on it after arranging their attention better. A person who reads a long book on a phone may record a large total while doing exactly what they intended. No universal screen-time target can decide these cases without looking at the activity.
There are costs to measure as well. Did the change make work harder? Did you miss a message that mattered? Did another household member become responsible for watching a channel on your behalf? Did the new rule create an argument that a clearer agreement could have prevented?
A successful change should be able to describe what improved and what had to be given up. If the only defense is that the number fell, the evaluation is too small.
The aim is a better relation to the tool. Sometimes that means far less use. Sometimes it means moving one use to a different time or form. Sometimes it means leaving a particular service. The answer should follow the experience rather than precede it as a declaration of purity.
What a collection of studies can tell us
The word detox invites a clean story. Something bad has entered; abstinence will remove it; the person will emerge restored. The story is easy to sell because it supplies a villain, a ritual, and an ending.
It is not an adequate account of the research.
A systematic review by Theda Radtke and colleagues examined 21 studies involving 3,625 participants. The interventions varied, and the findings included benefits, no clear effects, and some worse outcomes. Many studies used small or student samples and short periods of disconnection.[3] The review predates the 2025 mobile-internet trial, so it cannot settle that newer evidence. Together, the work supports asking which change helps which person under which conditions.
That is less tidy than a cleanse. It is also more useful.
If we call every difficult urge a chemical problem and every period offline a reset, we can stop looking at the ordinary facts. What did the application provide? When did checking occur? Which tasks became easier? Was the person lonely, bored, anxious about a genuine obligation, or simply following a well-practiced routine?
Those questions have different answers. A person who needs company may benefit from arranging company. A person who keeps opening a service whenever a task becomes unclear may benefit from a clearer next step and a little friction around the service. The same block can interrupt both patterns while solving only one underlying problem.
This does not mean you must understand everything about your behavior before changing a setting. Small changes can teach you things. It means the explanation should remain open to what the change reveals.
A week without a feed may show that the feed had been occupying empty moments. It may also show why the moments felt empty. The second discovery deserves attention too.
Timing the delivery
There are options between continuous delivery and complete silence.
In a randomized field experiment with 237 Android users in India, Nicholas Fitz and colleagues tested different notification schedules. Delivering notifications in three daily batches improved several reported experiences, including attentiveness and sense of control, compared with usual delivery. Turning notifications off entirely produced greater anxiety and fear of missing out, with fewer benefits.[4]
These were self-reported outcomes in that study. They do not establish that three deliveries is the correct schedule for every job or that silencing any notification is harmful. They do suggest that predictability may be worth trying alongside reduction.
If you know when ordinary updates will be available, you may not need to keep guessing whether you should check. But the schedule must fit the work. Someone coordinating same-day transport may need more frequent access than someone reviewing a weekly newsletter. A useful plan respects the reason a channel exists.
Consider a small community group planning an event. Messages arrive throughout the day, but most decisions can wait until evening. One member keeps checking because they do not want to delay the others. The group could agree that routine choices will be gathered and settled at a particular time. A separate call remains available for a problem that cannot wait.
The arrangement does more than reduce pings. It tells people what a delay means. Silence during the afternoon is no longer an ambiguous failure to participate. It is the agreed form of participation.
This is an example we can design for ourselves. No experiment has proved that this exact community rule will improve every outcome. Its logic is transparent enough to test: group the decisions that can wait, preserve the path for those that cannot, and see whether the event is still organized well with less repeated checking.
If it fails, ask where. Did a decision really need an earlier answer? Was the meeting time inconvenient? Did someone continue to send urgent-looking messages for routine matters? A failed first arrangement can provide the information needed for a better one.
A habit does not explain itself
A different kind of research compares people who habitually combine media with people who do so less often. This asks another question from an experiment that interrupts someone during a task.
Douglas Parry and Daniel le Roux’s 2021 meta-analysis found small, varied associations between media-multitasking habits and cognitive control. Links were stronger for people’s reports of their control difficulties than for performance on tests.[6] Such associations do not establish that the habit caused a lasting loss of ability. The direction of influence and the surrounding circumstances remain important.
A person might turn repeatedly to media because the current task is hard to engage with. Repeated switching might also make that task harder. A stressful period could affect both. Several paths can produce a relationship between two measurements. A survey taken at one point cannot tell us which path brought each person there.
This is why immediate effects, habitual behavior, and lasting capacity should remain separate in our thinking. A message can interrupt a sequence now. A person can report frequent switching across a week. A test can measure one aspect of attention under specified conditions. None of those facts, by itself, tells the whole story of the others.
The distinction protects hope as well as accuracy. If you interpret every difficult day as evidence that technology has permanently damaged you, a practical change may seem too small to matter. If you interpret every difficulty as a habit you should be able to stop immediately, you may overlook a need for more support.
Start with what can actually be observed. There is an activity you want to do. There is a recurring pattern that gets in its way. A change in the pattern may help, and you can examine whether it does. You do not have to settle the history of your entire attention system before making the next interval less difficult.
Equally, a successful week is not proof that all attention problems are voluntary. People differ, circumstances differ, and some difficulties persist despite careful effort. A useful practice should make those differences easier to notice, rather than give us another way to judge one another.
The phone on the table
Should the phone leave the room?
It might help. Distance can make an unplanned check less convenient. It can also make a boundary visible to the people around you. A device on a shelf may be easier to leave alone than a device already in your hand.
But we should be careful about the stronger claim that the mere presence of a silent phone reliably drains everyone’s mental ability. Research syntheses have not produced one simple answer. A 2024 meta-analysis by Andree Hartanto and colleagues, covering 53 samples, found no significant overall effect of phone presence on the cognitive measures considered. Another meta-analysis by Douglas Parry found a small negative working-memory effect, with no statistically significant pooled effect for several other outcomes.[5]
That uncertainty does not forbid moving the phone. It changes the reason we give for doing so.
You can say: when the phone is beside me, I often pick it up; I want to make that action less automatic. That is a claim you can observe in your own routine. You need not add that the device is emitting an invisible tax on your intelligence while it lies silent.
The reason matters because it helps you adapt. If you need the phone for an authentication code or a call you are expecting, the rule can allow that use. If a desk drawer works as well as another room, you can choose the drawer. You are adjusting the path to a behavior, not protecting a sacred perimeter from a contaminated object.
Keep the practical intervention smaller than the mythology around it.
An experiment you can actually learn from
Choose one recurring situation. For example, you want to read for a while after dinner without repeatedly opening a news application.
Before changing anything, describe what currently happens. You sit down with the reading, encounter a difficult paragraph, reach for the phone, and return several minutes later. You may discover that the difficult paragraph is the trigger. Or perhaps the phone vibrates first. Or perhaps you are so tired that neither activity is getting much from you.
Do not turn this observation into surveillance of every minute. A short note about a few evenings may be enough to see the pattern. The purpose is to answer a question, not to create an administrative career in monitoring yourself.
Now choose one change that fits the pattern. Silence the nonessential alert. Move the application off the first screen. Keep the phone out of easy reach while preserving a necessary contact route. Choose a shorter reading interval. Decide what to do when a paragraph becomes difficult: reread it once, make a note, or stop at a clear point instead of fleeing into another stream.
For the first attempt, one change can make the result easier to understand. If you change six things and the evening improves, you may not know which one mattered or which burdens you can safely drop. Sometimes several changes are necessary together; a phone rule without a contact arrangement may be incomplete. Use judgment rather than treating simplicity as another rigid command.
Decide in advance what would count as help. Perhaps you want to follow the chapter and recall its main point. Perhaps you want a calmer evening, or to go to bed when you intended. Include any cost you do not want to impose on others.
Then give the arrangement a fair try across more than one unusual day. A quiet Sunday is not proof that the same rule fits a busy workweek. Nor does one interrupted evening prove that the rule is useless. Look for a pattern strong enough to guide a practical decision, while remembering that your informal test cannot isolate every cause.
If you are changing the rule every time you feel an urge to check, you may be avoiding the test. If the rule repeatedly obstructs something necessary, keeping it unchanged is not discipline. It is refusing to use the information you have gathered.
At the end, keep what helps. Adjust or discard what does not. You are not obliged to defend the first plan because you announced it with confidence.
What happens to the recovered time?
An empty interval does not remain empty just because you fought to make it.
Other tasks can expand into it. You may notice a chore you had postponed. Someone may ask for help. An employer may treat improved efficiency as evidence that more work will fit. You may do that to yourself.
This is why it helps to name what the change is for. If the aim was to read, put the book where the recovered interval begins. If the aim was to speak to a friend, arrange the call. If the aim was rest, resist the idea that the time must justify itself by producing something new.
A named purpose is not a command that every saved minute obey you. It is a way to prevent the benefit from being absorbed so quietly that you wonder why the change made no difference.
Suppose the news application is no longer part of the evening. For three nights, you use the interval to catch up on work. By Friday, the screen-time report congratulates you, but you have not read and do not feel more rested. The application rule may have worked exactly as designed. The plan for the freed time was missing.
There is also a social question. If one person’s reduced checking means another person now handles every update, the household total may not have improved. The quiet has been transferred. A fair experiment asks who carries the work before and after the change.
These are not reasons to abandon a useful practice. They are reasons to follow its effects far enough to see where the benefit goes.
Attention advice often ends when the switch is turned off. Your day continues. The next choice is where the recovered space becomes a part of your life.
Turning something back on
At the end of the imagined week, you may decide to keep the application closed. You may decide to use it at particular times. You may restore it because the costs of absence were too high and try a narrower change instead.
Returning is not automatically a failure. You have learned something if you return with a clearer purpose and a better boundary.
Perhaps you now go directly to the group whose messages matter. Perhaps you disable a channel that had no useful role. Perhaps you stop blaming yourself for needing a service on which work or friendship actually depends, and begin asking whether the unwanted parts can be separated from the useful ones.
You may also find that the service is difficult to use selectively. Then leaving or reducing it more sharply can be a reasonable choice. The important thing is that you can explain the choice in terms of your life, rather than a story that treats every minute of connection as either salvation or damage.
The phone can still ring. The person you wanted to reach can still be reached. A necessary task can still be done.
And between those uses, there may now be a little more room for an activity that does not have to compete with every other possibility at once.
Sources and notes
[1] Noah Castelo, Kostadin Kushlev, Adrian F. Ward, Michael Esterman, and Peter B. Reiner, “Blocking Mobile Internet on Smartphones Improves Sustained Attention, Mental Health, and Subjective Well-Being,” PNAS Nexus 4 (2025), pgaf017. Randomized delayed-intervention design; short follow-up, unblinded participants, substantial noncompliance and missing follow-up data. Compliance required blocking on at least ten of fourteen days. Paper
[2] Hunt Allcott, Luca Braghieri, Sarah Eichmeyer, and Matthew Gentzkow, “The Welfare Effects of Social Media,” American Economic Review 110 (2020), 629–676. Paid four-week Facebook deactivation among eligible volunteers; the results concern that service, sample, and election period. Paper
[3] Theda Radtke, Theresa Apel, Konstantin Schenkel, Jan Keller, and Eike von Lindern, “Digital Detox: An Effective Solution in the Smartphone Era? A Systematic Literature Review,” Mobile Media & Communication (2022; online 2021). The 21 studies included 12 randomized trials; differing interventions and outcomes limit a single universal conclusion. Review
[4] Nicholas Fitz, Kostadin Kushlev, Ranjan Jagannathan, Terrel Lewis, Devang Paliwal, and Dan Ariely, “Batching Smartphone Notifications Can Improve Well-Being,” Computers in Human Behavior 101 (2019), 84–94. A randomized field experiment with 237 participants. Self-reported attentiveness is not an objective test of sustained-attention capacity. Paper
[5] Andree Hartanto and colleagues, “The Effect of Mere Presence of Smartphone on Cognitive Functions: A Four-Level Meta-Analysis,” Technology, Mind, and Behavior 5 (2024), 53 samples and 4,368 participants. Review. Douglas A. Parry, “Does the Mere Presence of a Smartphone Impact Cognitive Performance? A Meta-Analysis of the ‘Brain Drain Effect,’” Media Psychology 27 (2024; online 2023), 737–762. Review
[6] Douglas A. Parry and Daniel B. le Roux, “Cognitive Control in Media Multitaskers” Ten Years On: A Meta-Analysis, Cyberpsychology 15 (2021), article 7. The synthesis includes 118 assessments; associations vary by how habits and control are measured. Review