You probably know someone who describes themselves as good at multitasking. Maybe you describe yourself that way. They study with music playing, the television on, messages pinging, a browser tab open to something interesting, and they claim, often persuasively, that they take in more this way than they would with fewer inputs. They can hold multiple things in their head. They don’t get bored. They’re, as the claim usually goes, just wired differently.
In 2009, a Stanford communications researcher named Clifford Nass set out to study these people — the heavy media multitaskers, the ones who genuinely believed and reported that they could handle many streams of information at once. He expected to find that they had developed real cognitive advantages. He didn’t.
What he found, to his own surprise and disappointment, was that heavy multitaskers performed worse than light multitaskers on nearly every cognitive test he could design. Worse at filtering irrelevant information. Worse at switching between tasks. Worse at holding information in working memory. The people who most confidently described themselves as multitaskers were, in fact, the worst at the mental operations multitasking requires. Nass, shortly before his death in 2013, described the finding as one of the most uncomfortable results of his career. “They’re suckers for irrelevancy,” he told an interviewer. “Everything distracts them.”
The study has since been replicated and extended. And it points to something worth understanding carefully, because most of us — students, workers, ordinary humans living on phones — are heavy media multitaskers by Nass’s measure, whether we recognise it or not.
What multitasking actually is
To understand why this research matters, it helps to be clear about what multitasking really is. In the narrow technical sense studied by cognitive psychologists, very little of what we call multitasking is actually happening in parallel. The brain is not, in most cases, doing two cognitively demanding things at once. It’s switching rapidly between them — fast enough that it feels simultaneous, but in sequence, one at a time, with a small cost at every switch.
The research on task-switching costs goes back decades, with foundational work by the University of Michigan psychologists David Meyer and David Kieras in the 1990s. What they and others have documented is that every time you switch between two tasks, there’s a time penalty for reorientation. Your working memory has to clear the first task, load the second, and re-engage. These switches are fast — a fraction of a second, usually — but they accumulate. Ten switches a minute means several seconds per minute lost to switching. Over an hour, that’s a meaningful proportion of your productive time evaporating into the overhead of the switches themselves.
Worse, the switches don’t just cost time — they cost accuracy. People switching between tasks make more errors on both tasks than people doing each task alone, even when you correct for the time spent. The mental residue of the previous task lingers, and it interferes with the current one. This is especially pronounced for cognitively demanding tasks — the kind where you’re trying to think, learn, write, or problem-solve. For simple, well-practised tasks, the cost of switching is small. For hard ones, it’s substantial.
The Stanford study in detail
Nass and his collaborators, Eyal Ophir and Anthony Wagner, designed a series of tests to compare heavy and light media multitaskers. They recruited about 260 Stanford students, measured their typical media use through a questionnaire, and split them into the extreme ends of the distribution — the people who reported almost always being engaged with multiple media streams simultaneously versus those who typically engaged with one at a time.
The tests were designed to probe the cognitive skills that multitasking should, in theory, train. One measured the ability to filter out distractions — participants were shown arrangements of red and blue rectangles and asked to note whether the red ones had changed orientation, ignoring the blue ones. Heavy multitaskers were worse at ignoring the blue. Another measured working memory — holding information in mind across short delays. Heavy multitaskers were worse here too. A third specifically measured task-switching ability — the thing multitaskers should, of all things, be good at. They were worst of all.
Across every measure, the heavy multitaskers performed worse. The effect held after controlling for intelligence, for technology familiarity, for a range of demographic variables. The only cognitive skill where heavy multitaskers equalled light ones was the ability to describe themselves as skilled multitaskers. They were not skilled. They were simply more confident.
Why we think we’re good at it
One of the most interesting questions raised by this research is why the self-perception is so durable in the face of contradicting evidence. Several factors seem to contribute.
The first is that multitasking produces a feeling of efficiency. Switching between tasks gives the brain a continuous sense of novelty, small dopaminergic hits every time attention lands somewhere new. This feels like engagement, like productivity, like being on top of things. The subjective experience is genuinely pleasant, which is why heavy multitaskers report enjoying their mode of work. The fact that they’re actually getting less done, and doing it worse, isn’t directly available to introspection. You can’t feel the cognitive cost of a task switch. You can only feel the stimulation of the switch itself.
The second is that multitaskers rarely get feedback on their actual output. A student who studies with distractions may not know that their test score would have been five per cent higher with focused study, because they never ran the comparison. A worker who writes an email while on a video call may not realise that the email is weaker than it would have been if written alone, because they never see the counterfactual. The felt experience of “I got it done” is the only available data.
The third is that modern environments are specifically designed to reward multitasking-like behaviour. Phones, messaging apps, notifications, social media — these are all built to capture attention briefly and hand it back, building a habit of short, switchable engagement. A person marinating in this environment for hours every day is being trained, implicitly, to treat attention as a constantly-renewable short-term resource rather than as a scarce, protected commodity. The sense of being good at this kind of rapid switching comes partly from being practised in it — even if the practice is actively degrading the deeper cognitive capacities the person would use if they were trying to do something demanding.
The honest counter-thread
It would be misleading to leave this without acknowledging the places where the simple anti-multitasking story breaks down.
Some task combinations really are simultaneous in a meaningful sense. You can walk and talk. You can drive on a familiar road while carrying on a conversation. You can listen to music while doing mechanical work. The research on task-switching costs applies specifically to tasks that demand the same cognitive resources — working memory, language processing, sustained attention. Tasks that recruit different systems don’t compete in the same way. The bias in the multitasking literature is toward combinations of cognitively demanding tasks, because those are where the costs are most visible. For tasks that don’t compete, the costs are small.
There’s also a question of whether “multitasking” as measured in these studies maps onto the variety of things people mean when they use the word. A person who studies with instrumental music in the background is technically multitasking in some sense, but the cognitive cost is probably small, because instrumental music makes few demands on the language or working-memory systems. A person writing an essay while in a lyrical-music conversation would pay a much higher cost. The research finding “multitasking is bad” is less a universal claim than a specific one: demanding cognitive tasks done in parallel with other demanding tasks degrade both.
Finally, the measurement of “heavy multitasker” in the original study was by self-report, and self-reports of media use are notoriously inaccurate. The people in the heavy-multitasker group might differ from the light-multitasker group on other variables that the study couldn’t fully control for. The effect has been replicated multiple times and looks real, but it’s worth being honest that any single variable like “multitasking behaviour” is hard to isolate cleanly from everything else it correlates with.
What this should change about your habits
For most students, the practical implication of this research is straightforward. The belief that you can study effectively with phone notifications, multiple browser tabs, a television in the background, and a messaging app checked every few minutes is almost certainly wrong. Not wrong in the sense that you’re lying to yourself; wrong in the sense that your subjective experience of working in that environment is disconnected from how much is actually going into your memory.
The cleanest intervention, supported by almost every replication of this research, is to protect genuine single-tasking time. Not eight hours a day. Even thirty or forty-five minutes at a stretch, done with the phone in another room and tabs closed, produces meaningfully more output and meaningfully better memory than the same time spent with divided attention. The evidence for this is overwhelming.
For work that’s particularly important — writing an essay, learning difficult material, preparing for a serious conversation, doing creative work — the protection should be more deliberate. Most knowledge workers in professional settings now have so thoroughly lost the habit of sustained attention that they need to rebuild it in small increments. Cal Newport, whose work on deep focus we’ll cover separately, has documented the cognitive capacities that can be restored with a few weeks of deliberate attention training. It isn’t quick, but it isn’t impossible either.
What doesn’t work, the research is clear, is trying to train yourself into multitasking ability. There is no practice regime that makes humans good at genuinely simultaneous cognitive work, because humans don’t have that architecture. What practice produces is the feeling of comfort with switching, coupled with a degradation of the ability to do the thing you were supposedly training. This is a specific and painful trade.
The question that remains
The most useful thing you can take from the multitasking research is probably a reframe of what’s actually happening when you’re working with distractions.
You’re not doing multiple things at once. You’re doing one thing at a time, very quickly, at a constant cost. You’re not being efficient. You’re generating the feeling of efficiency while producing less output of lower quality. You’re not training a cognitive superpower. You’re training a habit of shallow engagement that will, over the years, make the demanding cognitive work of your life — whatever it turns out to be — harder than it needed to be.
The question to sit with, especially if you’re someone who has identified strongly as a multitasker:
Of the things you’re proud of in how you work, how many of them have you actually tested — against a version of yourself that tried the same work with a single thing in front of you and nothing else? The answer, for most people, is almost none. And for most people, the test, if you ran it, would surprise you.
Key research referenced: Eyal Ophir, Clifford Nass and Anthony Wagner’s 2009 PNAS paper on media multitasking; David Meyer and David Kieras’s research on task-switching costs; subsequent replications and extensions including Uncapher and Wagner’s 2018 meta-analytic review.