Does Meditation Make You Better at Paying Attention?
Most people who download a meditation app are not looking for enlightenment. They are looking for the ability to read one page without checking their phone.
The pitch writes itself. Mindfulness is usually defined as paying attention to the present moment. Practise paying attention, get better at paying attention. It sounds less like a spiritual claim than like going to the gym.
The research on that promise is now large enough to check. It does not say what the pitch says.
Does meditation improve attention?
In 2022 a 32-author team published the first meta-analysis of objective cognitive outcomes across multiple domains from randomised meditation studies in adults. They searched seven databases, found 56 studies covering 2,931 people, and pooled 45 of them covering 2,238.
Across every cognitive domain at once, mindfulness programmes beat their comparison groups by a small margin, g = 0.15, with a confidence interval of 0.05 to 0.24.
Then they split it by domain, and the pattern inverted the theory.
Executive function showed a small significant effect, g = 0.15. Working memory, treated here as part of executive function, showed the largest effect in the review, g = 0.23 on an interval of 0.11 to 0.36, and was the only subdomain to reach significance.
Attention did not. In the authors' words, the programmes "did not significantly outperform comparators for attention, declarative memory, nor cognitive aging."
The sharpest version sits in their supplementary analysis. Six studies used the same measure, the Attention Network Test, which splits attention into three networks: alerting, orienting and executive control. Meditation did not outperform the comparison groups on any of the three.
So the domain meditation is defined by is where the evidence is weakest. The domain it rarely advertises, holding information in mind, is where the effect is strongest.
Why is the attention finding so weak?
Three reasons, and only one is the expected one.
The first is the comparison group. Given a control group with nothing to do, meditation won: g = 0.20 across 28 studies. Given a control group with a real activity of matched length, the effect fell to g = 0.07, on an interval running from minus 0.04 to 0.19. Not significant, across 22 studies. The authors name what that leaves open: therapeutic alliance, social stimulation and treatment expectancy, all shared by both arms.
The second is age. In studies of adults under 60 the pooled effect was g = 0.11 and did not reach significance. In studies of adults over 60 it was g = 0.21 and did. If you are 34 and tired, the age group the evidence supports is not yours.
The third is a non-reason, worth stating because everyone assumes it. This is not hidden null results. The review compared trials registered in advance against trials that were not, and the registered ones showed a larger effect, g = 0.23 against g = 0.09. The sloppier literature is not inflating this result. It is diluting it.
What the review is blunt about is blinding: 45 of the 56 studies, 80 percent, were rated at high risk of bias for blinding participants, close to unavoidable when the intervention is a course you knowingly attend.
A separate 2021 meta-analysis arrived at the same place. Pooling 29 studies of group-based mindfulness programmes and executive functioning, it found small effects for inhibition, working memory and verbal fluency, nothing significant for attention shifting, and summarised itself in one sentence: the effect "is not robust."
Does a short daily practice do anything?
In 2019 a group at New York University randomised non-experienced meditators aged 18 to 45 to either 13 minutes a day of guided meditation or 13 minutes a day of podcast listening for eight weeks. Note the control: same length, same daily commitment, same headphones. A harder test than most studies in the meta-analysis above.
Against the podcast group, the meditators improved on attention, working memory and recognition memory, and reported less negative mood and less anxiety under a laboratory stress test.
The crucial detail is the timing. The paper's phrasing is that this held at eight weeks but not at four. Four weeks of daily practice, roughly 6 hours of sitting, produced nothing measurable. Eight weeks did.
What about one short sit, right now?
A single session does something smaller and stranger than the training effect.
Two experiments in 2018 gave novices a 10-minute meditation recording or a control recording, then tested them immediately. In the first, the meditation group was more accurate on the hard flanker trials without getting slower. In the second, they finished an Attention Network Test faster without getting less accurate. Both patterns point to better allocation of a fixed resource rather than more of it.
But the effect depended on who you are. Neuroticism moderated both results, and the brain measure appeared only in participants lower in neuroticism. Ten minutes helped some novices and not others, worth knowing before you conclude it did not work for you.
How much practice did the clearest attention gain cost?
The cleanest sustained-attention result in healthy young adults belongs here, because of its price.
A UCSF team built a closed-loop app that lengthened each breath-focus trial as users improved, and ran it for six weeks against off-the-shelf apps as an active placebo. The meditation group gained on sustained attention and working memory, with matching changes in two EEG signatures of attentional control.
Now the two numbers that matter. The daily dose started at 20 minutes in weeks 1 and 2, rose to 25 in weeks 3 and 4, and finished at 30 in weeks 5 and 6. The meditation group had 22 people.
So the strongest demonstration that meditation trains attention in young adults took 20 to 30 minutes a day for six weeks, in a sample small enough to fit in a classroom. Five minutes a day is not that experiment, and no honest reading says otherwise.
Does consistency matter more than length?
In 2020 researchers assigned 121 firefighters to mindfulness training, to an active relaxation programme matched session for session and taught by the same trainer, or to no training. Both programmes asked for 10 to 15 minutes of daily practice at home.
Resilience scores rose more in the mindfulness group than in either comparison. Then the authors read the practice logs, and found that performance on an objective attention task, along with positive affect, rose more over time in the people who practised on more days per week. That relationship held in the mindfulness group and not the relaxation group.
Be careful with it. This is a correlation inside one arm, not a randomised test of frequency, and people who practise more may simply be different people. But it is the variable that moved, and it was days, not minutes per sit.
What does this mean for five minutes?
Do not start meditating to fix your focus. That is the claim with the thinnest support, it fails against active controls, and in under-60s it does not clear significance at all. Anyone selling attention training in five minutes a day is ahead of the evidence.
Expect the effect that did survive. Working memory, holding a few things in mind without dropping them, is the one domain where the number held up across 2,238 people. That looks like remembering what you walked into the room for, not grinding through a spreadsheet.
And give it longer than you want to. The best-controlled brief-practice trial found nothing at four weeks and something at eight. Which is the argument for a short session: five minutes is the length you can still do in week seven, and week seven is where the evidence starts.
One thing to do today
Stop measuring your practice by how focused the sit felt.
That number tells you about today's sleep and today's inbox, not about whether anything is accumulating. A calm sit and a scattered sit count the same in these studies, because what the trials recorded was attendance.
So take the one metric the research rewarded, days practised, and ignore quality for eight weeks. In this app that is the Streak, and the Grace that keeps one missed day from erasing it. Sit for five minutes whether or not you expect it to go well, then judge the practice by the count in December rather than by how any single sit felt.
Sources
This article shares general information about meditation research. It is not medical advice. If you have concerns about your mental health, talk to a qualified professional.
Five minutes. Every day. That is the whole app.