Why Your Attention Span Feels Broken

You used to read novels. Now you open one, read a page and a half, and check your phone without deciding to. A three-minute video feels long. You have started skipping the intros of things you chose to watch.

The obvious conclusion is that something in you has been damaged - that your attention span used to be a certain size and is now smaller, the way a muscle wastes.

That conclusion is worth examining before you go looking for how to fix your attention span, because the most famous number supporting it has no traceable evidence behind it, and the most cited research on the subject does not measure what people think it measures.

The short answer

There is no reliable evidence that human attention capacity has shrunk. What has changed is the environment your attention operates in, and the habits you have built inside it.

The practical consequence is that most advice on how to improve attention span underperforms. Attention-span "training" - willpower, discipline, focus streaks - treats a systems problem as a personal one. The evidence points more strongly toward changing conditions than toward training a supposedly damaged capacity: reducing the number of available switches, lengthening the default unit of engagement, and noticing that roughly half of the interruptions people experience in these studies are self-generated.

First, the number that isn't true

If you have read that the human attention span has fallen to eight seconds, and that this is shorter than a goldfish, you have encountered one of the most successful pieces of misinformation of the last decade.

The claim is usually traced to a 2015 Microsoft Canada consumer report. According to published accounts of a 2017 BBC News investigation by Simon Maybin, the eight-second figure did not come from Microsoft's own research at all - that research surveyed 2,000 Canadians and measured brain activity in 112 people, and reportedly never mentions eight seconds or goldfish. The number came from a firm called Statistic Brain, which listed the US National Center for Biotechnology Information and the Associated Press as its sources. When the BBC contacted those sources, according to these accounts, neither could produce any record of research supporting the figure.

A note on how solid that is: the BBC's own page could not be retrieved for direct inspection, so this chain is reported here as documented by independent secondary accounts of that investigation rather than verified first-hand against the original. What can be said with confidence is narrower and still sufficient - the eight-second statistic has no traceable credible evidentiary basis, and nobody has produced the study it supposedly comes from. The comparison to goldfish is not supported either; research on goldfish memory finds retention over periods of weeks to months.

There is also a deeper objection. Attention researchers quoted in coverage of the myth have pointed out that attention is task-dependent - how long someone stays with something varies enormously by what it is, why they are doing it, and what else is available. On that view the problem is not that eight seconds is the wrong number. It is that a single number for "the human attention span" is not a coherent measurement in the first place.

That does not mean nothing has changed. It means the thing that changed has to be measured properly.

What the real research actually measured

The most credible work here belongs to Gloria Mark, a professor at UC Irvine who has spent two decades measuring attention in real workplaces rather than laboratories, using computer logging, sensors, brief in-the-moment probes and observation.

Her findings are consistent and striking:

• In 2003-2004, people spent an average of about two and a half minutes on any screen before switching.
• By 2012, that average had fallen to 75 seconds.
• In recent years it has been around 47 seconds, with other research groups measuring 50 and 44 seconds. The median observed was about 40 seconds.

At the level of whole projects rather than individual screens, people worked around 10.5 minutes before switching, and took roughly 25 and a half minutes to return to an interrupted project.

Now the critical distinction, and it is the one almost every article on this topic skips.

What Mark measured is switching behaviour, not attention capacity.

These are recordings of how long people actually stayed on a screen in a working environment full of notifications, tabs, messages and obligations. They are not a test of a person's underlying ability to pay attention when they want to. A figure of 47 seconds describes an environment as much as it describes a mind.

This matters because it changes the diagnosis entirely. "My capacity has shrunk" implies repair. "I am in an environment that rewards switching every 47 seconds, and I have adapted to it" implies redesign.

The hidden system: an environment you have already internalised

The most uncomfortable finding in Mark's work is not the 47 seconds. It is where the interruptions come from.

"It turns out half the time we interrupt ourselves," she has said. "We self-interrupt because we're just so tempted to look up information or we're tempted to check email."

Half. Not notifications. Not algorithms. You.

This is the hidden system, and it is more interesting than the usual story. Interruption is not only something done to you from outside. Repeated switching can become self-initiated even when no notification arrives - the tab opened for no particular reason, the phone picked up before the sentence is finished. Mark's finding matters precisely because a large share of interruptions turn out to be generated by the person rather than externally triggered.

What that observation does not establish is the mechanism. It is tempting to describe this as an environment being internalised, or a rhythm being learned, and that may well be what is happening - but Mark's data are observational records of when people switched, not evidence of conditioning. The honest version is that the behaviour is frequently self-started, and that any account of why is still an interpretation rather than a finding.

Mark is careful about attributing this to any single villain. "Everyone tends to blame tech companies and algorithms," she notes, "and that's certainly a big part of it, but it's not the only part." She also points to the structure of the internet itself - organised by association and links, which is "conducive to distraction" by design - and to ordinary human factors: personality, emotional and social needs, curiosity. "We're social beings. We're curious."

That is the honest version. Not a machine stealing your attention, and not a moral failure on your part, but a well-fitted loop between an environment where switching is unusually easy and a mind that was always going to be curious. It is also consistent with something people who feel easily distracted often describe: a pull that seems to come from nowhere in particular.

The counterweight: the science here is weaker than the headlines

There is a widely repeated claim that heavy media multitasking measurably damages cognitive control. It comes from a 2009 study by Ophir, Nass and Wagner, and it has been treated as settled for years.

It is not settled.

Wiradhany and Nieuwenstein ran two replication studies and a meta-analysis. Across 14 tests with an average statistical power of 0.81, only five produced significant effects in the predicted direction - and under more conservative Bayesian analysis, only two survived. Their meta-analysis of 39 effect sizes found a weak but significant association between media multitasking and distractibility that became non-significant once corrected for small-study effects.

Their conclusion is direct: "these findings lead us to question the existence of an association between media multitasking and distractibility in laboratory tasks of information processing."

This should change how confidently anyone writes about this subject, including us. The claim that digital media has produced measurable cognitive damage is not well supported. The claim that digital environments change moment-to-moment behaviour is very well supported. Those are different claims, and conflating them is how the goldfish statistic survived for ten years.

So the careful statement is this: the available evidence is much stronger for changes in switching behaviour than for a broad decline in underlying attention capacity. That is not the same as proving nobody's capacity has changed, and it is not a claim about any individual reader. It is a statement about where the evidence is solid and where it thins out.

What follows from this diagnosis

If the stronger evidence is about behaviour and environment rather than capacity, several things follow - less as instructions than as a different account of where the difficulty sits.

Available switching opportunities are part of the problem. Environments with fewer immediate alternatives create fewer chances to self-interrupt. That is the practical significance of Mark's self-interruption finding: if roughly half the interruptions originate with the person, then the number of easy exits in the immediate environment matters more than it would if all interruption were external. It also explains why phone and browser defaults tend to do more work than intention does - defaults operate whether or not anyone is paying attention to them. Removing an option is a structurally different lever from trying to maintain focus against it.

Interruptions carry a cost that is easy to underestimate. If returning to an interrupted project takes something in the region of 25 minutes on average, the real price of a short check is not the length of the check. This reframes what "quick" means, and it is one of the clearer places where a measured figure changes how a familiar behaviour looks. We have written separately about where that time actually goes.

Attention is not uniform across a day. Mark's work also indicates that attentional capacity varies over the course of a day, which weakens the common assumption that every working hour should be capable of producing the same kind of concentration. Treating a low hour as a discipline failure misreads what is being measured.

Recovery appears to be a real variable, not a soft one. Mark points to modest interventions - time outdoors, brief meditation, keeping a current goal in view - and the notable thing about them is their scale. They are small because the underlying claim is small: depleted attention is not obviously restored by applying more attention.

The reference point shifts, and that shift is not the same as damage. If the ordinary unit of engagement is under a minute, longer things will register as slower - not because they changed, but because the comparison did. That is a change in expectation rather than evidence of a lost faculty, and the two are easy to confuse from the inside.

What none of this supports is the framing of attention as a personal challenge to be won. Mark's own emphasis is on structural responses - organisational quiet time, right-to-disconnect provisions - rather than individual effort. That is not a motivational point. It follows from the finding that the environment is doing a large share of the work.

Why "how to fix your attention span" may be the wrong question

The phrasing assumes a broken part. The evidence does not really support a broken part. It supports a well-adapted person in a badly designed environment.

That reframe is not a consolation. It is a redirection of effort. If the problem were capacity, the answer would be training, and you would have to become a different person. If the problem is conditions, the answer is arrangement - and arrangements are much easier to change than people.

Feeling less able to concentrate does not by itself establish that the underlying ability to concentrate has permanently shrunk. What has demonstrably changed is the set of conditions under which staying focused on one thing was the path of least resistance - and conditions are more tractable than faculties.

Common questions

Is it true that the human attention span is eight seconds?

No. The figure came from a 2015 Microsoft Canada report that sourced it from a firm which could not provide credible evidence for it, as a 2017 BBC News investigation reported. No study supporting it has ever been produced, and the goldfish comparison is not supported either.

Has my attention span actually gotten shorter?

Observed switching behaviour has changed substantially in workplace studies. Gloria Mark's measurements show average time on a screen falling from about 2.5 minutes in 2004 to roughly 47 seconds recently. Whether your underlying capacity has changed is a different and much less settled question - the research claiming measurable cognitive damage from media multitasking has largely failed to replicate.

How long does it take to refocus after being interrupted?

Mark's research on project-level work found people took roughly 25 and a half minutes to return to an interrupted project. That figure is an average from workplace observation, not a fixed law, but it is a useful order of magnitude when deciding whether to "just quickly check" something.

Can you actually fix your attention span?

Most searches for how to fix attention span, or for how to increase your attention span, assume the answer is training. Sustained attention can improve with practice and interventions like meditation show modest effects, but the evidence points more strongly at conditions than at capacity - roughly half of interruptions are self-generated, which makes the number of available switching opportunities a larger variable than willpower.

Do phones cause short attention spans?

The honest answer is that they contribute to an environment that rewards switching - social media feeds most of all - and switching behaviour has measurably increased. Direct evidence that phones have damaged attention capacity is weaker than commonly claimed. Mark explicitly resists blaming technology alone, pointing to internet structure, personality, curiosity and social needs as well.

Why is my attention span so short now?

Most likely because the number of easy, available switches around you has increased enormously, and you have adapted to them - including by generating your own. The feeling of decline is real. The explanation is usually environmental rather than neurological.

Related reading

• Screens, Attention, and Focus: The Complete HOS Guide
• Your phone isn't the problem. Its defaults are.
• Why you sit down to work and lose 40 minutes without noticing