Two weeks without internet on the phone: attention and adherence

Keep the phone, take away always-available internet

A phone can be where you answer a call from family and where you drift from a message into the next video. Counting both as screen time obscures what the device is doing for you. A randomized study published by Castelo and colleagues in 2025 tested what happens when internet access on the phone is blocked. For digital wellbeing products, the study also raises a question about how easily someone can leave the screen when an experience ends. [1]

A questionnaire and an attention task answer different questions

Over the first two weeks, the early group improved on subjective wellbeing, a composite of self-reported psychological symptoms, and performance on a sustained-attention task. No statistically significant pre–post change was detected on those three outcomes in the waitlist group during that period. All three measures improved in the waitlist group when it received the intervention. The wellbeing questionnaire combined life satisfaction with positive and negative affect. Attention performance came from a five-minute online task. [1]

The task gives the study a measure beyond questionnaires. Discriminating when to press a key over five minutes and finding it easier to concentrate at work or at home still call for separate tests. The composite tracks changes in self-reported symptoms over the study period. Clinical recovery would require separate assessment. A waitlist gives people no alternative activity they expect to help to a similar degree. What participants inferred about the purpose of the study could have influenced their responses.

467 started, and 119 met the adherence criterion

Of 467 participants, 266 set up the block and 119 met the reported prespecified adherence criterion, or 25.5%. A total of 313 completed all three self-report surveys. Primary analyses used complete cases for each measure, retaining people with complete data in their assigned groups regardless of adherence. Those who completed all three surveys had better baseline symptom scores and attention-task performance. [1]

The results section describes adherence as blocking access on at least 10 of 14 days. The methods section reports that 119 of a subset of 249 people who completed all three surveys disabled the block fewer than five times. Days and interruptions are different measures. The 25.5% figure uses the 119/467 count reported in both sections. [1]

Those denominators answer different questions. Getting a setting to work tells us little about whether it fits daily life, and completing follow-up is different from following the instructions throughout. Most people in the trial were already highly motivated to cut back on phone use, yet many did not meet the adherence criterion. The paper did not establish each person's reasons. A design team could ask about them separately. One person might need navigation or a work login. Another might rely on internet messaging to check on family. Someone else might simply find setup cumbersome. Those situations give a team more to work with than a single label of nonadherence.

Where did the freed-up time go?

In analyses pooling the two groups, participants reported more offline activity and less media consumption. The change in time spent on digital communication did not reach statistical significance. Including factors such as offline activity, social connection and sleep in statistical models accounted for some change in wellbeing or symptom scores. These pathways relied on self-report and statistical analysis. The study did not randomize those activities separately. The analysis found no statistically significant indirect effects for sustained attention. [1]

A digital wellbeing product could compare hiding content entry points at times the user chooses with simply offering usage reminders. Both designs could keep necessary communication and a temporary exit available. A test could record who completes setup, starts, keeps the setting and stops, along with their reasons. It could also ask whether reduced use moved to another screen or left room for rest, conversation or something the person already wanted to do. These designs remain to be tested. The trial examined a particular phone restriction, and four weeks of observation cannot establish long-term benefits or costs.

This article includes AI-generated content. AI assisted with search, verification of original sources, and writing and editing the Chinese and English texts. Luoboli conducted no hands-on product test or original interview. The operating team develops Six Realms and provides Heart to Heart and Sleep, creating a commercial interest. This article does not evaluate those products or endorse the tool used in the study.