Research and science
Every claim on this site has to earn its place. This page keeps the receipts. Only published findings, always with year and source. Correlation is never dressed up as causation. When new research lands, this library gets updated. When a claim does not survive, it gets retired.
Pillar 01 · Time
The baseline numbers for Germany. Measured where possible, and clearly marked when people report about themselves.
a day at the median, measured in a German sample of 1,797 adults. The median also included about 58 device interactions; the Android logs recorded about 144 app episodes. The day rarely disappears in one sitting. It disappears in fragments.
Tóth et al. (2025)1
participants interacted with their smartphones on the median day. The time does not only grow in long sessions, but through many small returns.
Tóth et al. (2025)1
of 18 to 24 year olds in Germany say their own use is too high, and 93 percent of that group report negative side effects. Across all adults, 46 percent check the phone right after waking.
Deloitte Deutschland (2024)2
That was the median number of individual app-use episodes in the Android logs each day. A typical session lasted 6.4 minutes and included several app switches.
Tóth et al. (2025)1
Pillar 02 · Focus
What field studies say about interruptions, and the one randomized trial that shows attention recovering when the pull pauses.
of participants improved in at least one of three outcomes when mobile internet was blocked on their phones for two weeks. Across the intervention group, subjective wellbeing, mental health and objectively measured sustained attention improved. It is the only randomized trial of its kind so far, and the strongest single result in this library.
Castelo et al., PNAS Nexus (2025)3
A meta-analysis of the d2 test of attention pooled 287 samples from 32 countries, about 21,000 people. It found no drop in attention performance in children and moderate gains in adults. Your focus is not broken. Your environment got noisier.
Andrzejewski et al. (2024)4
Interrupted work often still gets finished on time. People compensate by speeding up, and pay with significantly more stress, frustration, time pressure and effort. The cost was measurable after just 20 minutes of interrupted work.
Mark, Gudith and Klocke (2008)5
When a task stays mentally open, part of your attention stays behind with it and performance on the next task drops. A short note on where to resume eases the effect. Closed loops are not a luxury. They are how focus works.
Leroy (2009) · Leroy and Glomb (2018)6
Pillar 03 · Sleep
Nowhere is the picture more consistent than around sleep. It is still largely correlational, and we treat it that way.
Meta-analyses keep finding the same association. More electronic media use is linked with poorer sleep quality and more sleep problems. Sleep is the best documented area in this whole field, which is why clear evening boundaries are especially useful.
Han et al. (2024) · Ahmed et al. (2024)7
of teenagers in Germany say they are often tired in the morning because they were on their phone late at night. Not a lab finding, a yearly national survey, and one of the clearest everyday signals in the German data.
JIM-Studie, mpfs (2025)8
People who sleep badly may also reach for the phone more, so most sleep findings run in both directions. We phrase every sleep claim on this site as an association, and we would rather lose a headline than your trust.
Our reading rule
Pillar 04 · Presence and what works
The evidence behind the principles of ctrl: friction, commitment and plans made in advance.
Patterns of use that feel out of your control show stronger links with depression, anxiety, sleep problems and lower wellbeing than sheer duration does. Not every minute is equal. That is why ctrl aims at control, not at zero.
Huang (2022) · Shannon et al. (2022)9
fewer completed openings of selected target apps after six weeks with one sec, in a non-randomized field study of 280 continuously active users. This suggests that a deliberate pause can help. The study tested one sec, not ctrl.
Grüning, Riedel and Lorenz-Spreen, PNAS (2023)10
a day for three weeks: In a randomized trial, stress, depressive symptoms, sleep quality and wellbeing improved. After the intervention, use quickly moved back toward baseline. A short trial can work; an everyday system should help it last.
Pieh et al., BMC Medicine (2025)11
was the median in one early study, not a universal deadline. Research shows a wide personal range from weeks to months. We design for sustainable repetition.
Lally et al., UCL (2009) · Singh et al. (2024)12
is the meta-analytic edge of if-then plans over vague intentions. And decisions made before the moment of temptation win far more often than decisions made inside it, which is exactly what commitment devices are for. A ctrl session is set up in advance and ends on your terms, not in the scroll.
Bélanger-Gravel et al. (2013)13 · Ariely and Wertenbroch (2002)14
Honesty first
Some popular claims sell well and do not survive contact with the research. You will not find them anywhere on this site.
A marketing meme from 2015. No study ever measured the famous 8 seconds, in humans or in goldfish. We only quote research that exists.
The vending machine picture of the brain is a story, not a finding. We leave neurochemistry to the people who measure it and talk about behavior instead.
Research shows a wide personal range from weeks to months. Sixty-six days was the median in one study, not a universal deadline.12 We would rather set an honest expectation than a catchy one.
No such diagnosis exists for phones or social media. Formal diagnoses of that kind cover only gambling and gaming, which is why that word never appears in our writing about phones.
The evidence about mornings is thin. The strong evidence is about screens late at night. We build for what holds, not for what frightens.
A living page
Research moves, and so does this library. New studies are added regularly, numbers get updated, and weak claims get retired. When something important lands, we write about it.
Modes, streaks, friends and family, and what stays free. The research above is why the app works the way it does.
Explore the app → The KeyA numbered piece of aluminum and ceramic that turns good intentions into one deliberate tap. No battery, nothing to charge.
Pre-order the Key →The evidence is not only about mood. In a randomized controlled trial published in BMC Medicine, three weeks of limiting smartphone screen time to two hours a day improved stress, depressive symptoms, sleep quality and wellbeing in students, with small to medium effects.11 In another experimental study, five days away from Facebook were linked with lower cortisol but also lower life satisfaction.15 The findings are mixed and come from limited samples. We treat them as signals, not as outcome promises for ctrl.
Where research is correlational we write linked with, not caused by. We do not round numbers in our favor, and when the honest answer is that an effect is moderate, we say moderate.
Secure payment · 30 day money back
