Attention and IQ Scores: Why Focus Moves Results
Two people with the same reasoning ability can leave the same IQ test with different scores. One sat down settled, followed every instruction the first time, and held each problem in mind until it was solved. The other was tired, kept losing the thread halfway through items, and rushed the timed sections to get them over with. The test recorded both performances faithfully. What it could not do is separate the reasoning from the attention that carried it — because on a real test, the two arrive together.
Attention is not one thing
Psychologists split attention into jobs that sound similar but fail in different ways. Sustained attention is staying on one task over minutes without drifting — the skill a long, repetitive test section demands. Selective attention is picking the relevant detail out of clutter and ignoring the rest. Divided attention is splitting focus across two streams at once, which in practice usually means switching quickly between them and paying a cost each time. And inhibitory control is stopping an automatic response — not blurting the first answer, not tapping the phone that just lit up.
None of these is intelligence itself. But every IQ subtest runs on top of them. Lose the instruction, lose the middle digits of a sequence, or check out for ten seconds of a timed coding task, and the score drops for reasons that have nothing to do with how well you reason when you are fully present.
Where attention hides inside the WAIS-5
The WAIS-5 does not have an “attention index,” but attention is built into its design — as our WAIS-5 explainer describes, the expanded Working Memory coverage explicitly includes focus of attention alongside auditory, visual and spatial working memory. The new Running Digits subtest requires a progressive shift of attention as spoken numbers stream in at a natural rate. Symbol Search, part of the Processing Speed Index covered in our processing speed guide, draws on sustained attention and concentration as well as visual scanning. Coding rewards staying locked onto a repetitive symbol-pairing task for two full minutes.
This is why clinicians read a profile, not just a total. A person whose verbal and reasoning scores sit well above their working memory and processing speed scores may not have a reasoning problem at all — they may have an attention bottleneck that taxes exactly the subtests where focus has to be held longest. Our working memory guide covers the workspace itself; attention is what decides what gets into that workspace and stays there.
What the 2025 evidence actually shows
The largest recent synthesis is a 2025 meta-analysis by Nguyen and colleagues in Psychological Bulletin, pooling 71 studies with 98,299 participants on short-form video use — TikTok, Reels, Shorts and similar feeds. Heavier use was associated with poorer cognition overall (a correlation of about −.34, which the authors describe as moderate), and the strongest links were precisely with attention (about −.38) and inhibitory control (about −.41). Read that at its true size: it is a group-level association from mostly correlational studies. It cannot show whether heavy scrolling weakens attention, whether people who struggle to sustain attention scroll more, or whether a third factor drives both. Anyone who turns it into “short video destroys your brain” has gone past the evidence.
The stronger design arrived the same year. Castelo and colleagues ran a preregistered randomized trial published in PNAS Nexus in February 2025: 467 iPhone users had all mobile internet blocked on their phones for two weeks — calls and texts still worked, and laptops were untouched — with compliance tracked by an app. Sustained attention was measured objectively with the gradCPT, a continuous performance task in which city and mountain images morph into each other and you respond to one while withholding for the other. The blocking group improved on that task; the authors note the gain was roughly the size of ten years of age-related decline on the same measure, run in reverse. In total, 91 percent of participants improved on at least one of the three tracked outcomes: sustained attention, mental health, or subjective well-being. Two honest caveats: strict compliance was hard — only a minority kept the block fully in place for the required days — and part of the gain ran through how time got re-spent, with more in-person socialising, exercise and time in nature. The study shows attention is movable by environment. It does not show that any individual’s IQ rose, because IQ was not what it measured.
Smaller lab work points the same direction, more narrowly. In a 2025 experiment with 72 participants, 30 minutes of TikTok-style video removed the normal benefit of extra preparation time on a task-switching test, while a documentary and a no-video control kept it — a temporary effect on one component of cognitive flexibility, not a permanent change. And a study published in Scientific Reports in October 2026, drawing on 14 years of Fitbit records from 3,093 adults in the NIH All of Us programme, found that higher daily step counts went with slightly higher sustained attention scores — a small, observational link the authors do not present as proof that walking causes focus.
What this means for an IQ score
Put the pieces together and a practical picture emerges. Attention behaves partly like a trait — some people sustain it more easily than others — and partly like a state, moved by sleep, stress, environment and recent habits. An IQ test samples one person, on one day, in one state. That is one reason scores carry a margin of error, and why a distracted sitting tends to drag down working memory and processing speed first, while vocabulary and general knowledge — which lean far less on moment-to-moment focus — hold steadier. It is also why our guide to sleep and IQ test scores matters here: poor sleep is one of the most reliable ways to arrive at a test with attention already spent.
What attention research does not license is the claim that focus exercises raise IQ. Improving on an attention task shows improvement on that task and its close relatives. The 2025–2026 studies measured attention, well-being and task-switching — not Full Scale IQ — and their authors stay inside those limits. So should any app or course quoting them.
How to test on a focused day
- Remove the phone from the room, not just the desk. The trial evidence is about constant connection, not willpower. For a practice test or study session, another room beats a silenced phone in your pocket.
- Single-task the hour before. Arrive having done one thing at a time. A morning of rapid feed-switching is the opposite warm-up for a test that rewards staying with one problem.
- Protect sleep the night before. Sustained attention is among the first capacities to slip on short sleep, and it is the one timed subtests lean on hardest.
- Read the profile, not just the total. If working memory and processing speed sit far below your other indexes, consider attention and test-day state as explanations before concluding anything about reasoning.
- Retest in a better state before believing a low score. A score earned while exhausted, ill or distracted is a measurement of that sitting. Standard tests expect retest variation even under good conditions.
None of this inflates a score dishonestly. It does the opposite: it stops an avoidable state from being misread as a fixed ability.
Go deeper with AI
This article is the starting point. Copy any of these prompts into your favorite AI assistant to learn more about attention and IQ test scores:
🔎 Read your own score profile like a clinician would
Here are my WAIS-style index scores: Verbal Comprehension [score], Visual Spatial [score], Fluid Reasoning [score], Working Memory [score], Processing Speed [score]. Explain what it could mean if Working Memory and Processing Speed sit well below the other three, which non-ability factors (attention, sleep, test-day state) could produce that pattern, and what questions I should ask before treating the Full Scale score as my fixed ability. State clearly that you are interpreting a pattern, not diagnosing anything.
🧭 Design a fair two-week attention experiment on yourself
Design a two-week self-experiment modelled on Castelo et al. (2025, PNAS Nexus): one week with mobile internet blocked on my phone outside set times, one normal week, order randomised by a coin flip. Give me a simple daily sustained-attention self-check I can do in five minutes, a log sheet for sleep, screen time and exercise, and the rules for reading the result honestly — including why my self-check is not an IQ test and why one good week proves nothing on its own.
💡 Audit a “focus training raises IQ” claim
Here is a claim about attention or focus training: [paste the advert, post or course page]. Audit it against the 2025-2026 evidence: does it confuse the Nguyen et al. (2025) correlation between short-form video use and attention with proof of cause, does it quote the Castelo et al. phone-blocking trial as if it measured IQ, and does it show transfer to an untrained test or only improvement on its own exercises? Rewrite the claim so it only says what the evidence supports.
Tip: replace the bracketed parts with your own situation — the more specific your prompt, the more useful the answer.
The short summary
Attention is the carrier wave of an IQ test: sustained attention, selective attention and inhibitory control sit underneath WAIS-5 subtests such as Running Digits, Symbol Search and Coding, even though the battery has no separate attention index. A 2025 Psychological Bulletin meta-analysis of 71 studies and 98,299 people linked heavier short-form video use with poorer attention (about r = −.38) and inhibitory control (about r = −.41), without proving cause. A 2025 PNAS Nexus randomized trial with 467 participants then showed that blocking mobile internet for two weeks measurably improved sustained attention — evidence that focus is a movable state, not only a trait.
The single most important point: an IQ score records reasoning plus the attention state you brought to the sitting — so protect focus and sleep before any important test, and read a low working memory or processing speed score as a possible attention problem before reading it as a reasoning limit.
Want better prompts, in your inbox?
One sharper AI prompt + one new brain training and IQ topic every week. Free forever, unsubscribe anytime.
We never share your email. One email a week, that's the deal.