Did Short Videos Wreck Our Attention? What the Research Says

7 September 2026

Short answer

• Two of the three statistics you have read about attention have no source behind them.
• People switch between screens far more often than they did in 2004. That part is measured.
• No controlled study shows the brain losing its capacity to concentrate.
• Dopamine tracks surprise, so what changed is what your reward system expects, and expectations can be retrained.

Three numbers everyone repeats

Eight seconds, shorter than a goldfish. This traces to a 2015 report from Microsoft Canada, which took the figure from a firm called Statistic Brain, which never produced a source when asked. No peer-reviewed study sits behind it, and nobody measured the goldfish either.

Twenty-three minutes and fifteen seconds to recover from an interruption. The number comes from a 2006 Gallup interview with researcher Gloria Mark, not from a paper. Her published work from the year before, No Task Left Behind?, found something else: interrupted people finished the original task faster, and reported more stress doing it.

Forty-seven seconds on any screen before switching. This one holds up. Mark's team measured the same behaviour five times over two decades: 2.5 minutes in 2004, 75 seconds in 2012, 47 seconds recently, with replications at 44 and 40 seconds. Read it carefully though. It counts how often people switch, not how long they can hold on when they try.

What dopamine actually does

Dopamine gets described as the pleasure chemical, and that description is wrong in a way that matters here. Dopamine neurons signal a reward prediction error: they fire when something turns out better than expected, go quiet when it matches expectation, and dip below baseline when it falls short. Dopamine tracks surprise, and drives wanting rather than liking.

A feed of short videos is built directly on that. Each swipe pays out on an unpredictable schedule, and unpredictable payout is the arrangement that keeps the prediction-error system firing hardest. Slot machines run on the same principle.

The same fact explains why a "dopamine detox" describes nothing real. There is no reservoir to drain and no baseline to reset, as researchers keep pointing out. What shifts with practice is what your reward system has learned to expect, and how anything slower compares against that.

What the studies on short video find

• Heavier short-video use correlates with weaker sustained attention and worse academic results, across a 2025 systematic review of the field. Correlation cannot tell you which way the arrow points: distractible people may simply scroll more.
• Brief exposure changes how people reason. In two experiments published in Cyberpsychology in 2024, Ma and Jiang found that a short stretch of TikTok pushed participants toward fast intuitive answers and away from checking those answers.
• A 2025 study on task switching found people prepared worse for a signalled switch after watching short-form video.

These are small studies, run over short horizons, mostly on students. They establish that something moves. They do not establish permanent change.

What the studies do not find

No controlled evidence shows the brain's capacity to concentrate degrading. Scientific American went through the literature in 2025 and found the opposite of the popular story. Psychologist Michael Posner, who has studied attention since the 1970s, says there is no real evidence that attention has changed since it was first reported on in the late 1800s. Monica Rosenberg at the University of Chicago describes a large gap between what this feels like and what turns up in measurement.

So: people switch more, which is documented. The ability to hold attention when you sit down and try to hold it looks intact.

The honest version

Your reward system learned what a good payout feels like, and it learned from something that answers every second. A document open on a screen pays out in twenty minutes, if it pays out at all. Against that comparison the document loses, and it loses before you have made any decision about it.

That is a comparison problem rather than damage, and comparisons can be changed. The fix is not to make the work more interesting than a feed, which it will not be, but to shorten the distance between effort and something visible.

What shortens that distance

• Put a reward on the work itself. The largest meta-analysis of gamification in learning (Sailer and Homner, 2020) found real if modest effects: g = 0.49 on cognitive outcomes, 0.36 on motivation, 0.25 on behaviour. The cognitive effect survived a re-analysis restricted to the most rigorous studies.
• Make progress visible during the work, not after it. A number that only appears at the end pays out too late to compete with anything.
• Decide the stopping point before starting. This is the one piece of the Pomodoro technique with consistent support behind it. The research on timed blocks goes into where that support ends.
• Cut the switch, not the phone. The measured problem is switching frequency. Anything that keeps your eyes on one surface addresses it.

Where Focus Kitten fits

Focus Kitten reuses the loop that short videos run on and points it at work. Finishing a session earns coins, coins buy furniture and outfits for your kitten's room, and the room fills up over weeks. The payout is small, frequent, and attached to the thing you actually meant to do.

The timer sits in the Mac menu bar, so minutes worked and minutes left stay visible without switching tabs or apps. That covers the part the research is clearest about: the number of switches, and how long it takes to notice you have made one.

Sources

• Mark G., Gonzalez V., Harris J. (2005). No Task Left Behind? Examining the Nature of Fragmented Work. CHI 2005.
• Mark G., Gudith D., Klocke U. (2008). The Cost of Interrupted Work: More Speed and Stress. CHI 2008.
• Ma S., Jiang M. (2024). Swiping more, thinking less: using TikTok hinders analytic thinking. Cyberpsychology 18(3).
• The Impact of Short-Form Video Use on Cognitive and Mental Health Outcomes: A Systematic Review (2025), medRxiv preprint.
• Sailer M., Homner L. (2020). The Gamification of Learning: A Meta-Analysis. Educational Psychology Review 32, 77–112.
• Are Attention Spans Really Shrinking? What the Science Says. Scientific American, 2025.