AbilityBenchHemispheric asymmetry
Left or right brain test that measures two asymmetries pointing opposite ways
There is no dominant hemisphere to find, so this page finds the asymmetries that are really there: 48 five-letter strings flashed for 150 ms to one side of a cross, which are named faster on the right, and 28 lines with a mark near the middle, which most people judge with a leftward pull. Free, no account, nothing held back, and about five minutes for both parts. The two results run in opposite directions in the same visitor, which is exactly why the quiz version of this idea has nothing to measure.
- 100% free
- No signup
- 48 flashed strings
- 28 bisected lines
- No dominant side
Two tasks, about five minutes together, and they are built to disagree with each other. The first flashes five-letter strings to one side of a cross for 150 ms and asks whether each one is an English word. The second draws a line with a mark near its middle and asks which half looks longer. Neither one produces a side, and the reason is the point of the page.
Part one — 48 flashed strings
Keep your eyes on the cross. A string appears to its left or its right, is gone in 150 ms, and you answer word or not-a-word. Needs a keyboard or a steady tap, and needs both halves of your field of view working.
Part two — 28 bisected lines
No clock at all. A vertical mark sits a little to one side of the middle of a line; you say which half is longer. Seven mark positions, four times each, in a mixed order.
The practice flashes use eight strings that appear nowhere in the scored run, so nothing you have already seen comes back. Prefer to run only the untimed half? . That is the honest route if a scotoma, an eye patch or anything else means one side of your field of view is not delivering — a field difference measured through a field defect is a measurement of the defect.
How to run both halves of this test
One task with a clock on it, one without, and a comparison that only works if you do both.
Sit still and hold your eyes on the cross
The whole word task depends on your gaze not moving. A string appears to the left or the right of the cross and is gone in 150 ms, which is shorter than the time it takes to plan and launch an eye movement — so the string is processed wherever it landed rather than wherever you looked. If you chase it, both hemispheres get a clean look at it and the difference the page reports collapses toward zero.
Answer word or not-a-word with two keys
J says it was an English word, F says it was not, and both are on screen as buttons for a phone or a machine without a keyboard. Half the 48 strings are real five-letter words and half are pronounceable near-misses like GLASP and CHEFT, which is what makes it a decision rather than a glance. Eight optional practice flashes come first, with feedback and no time limit, on strings that appear nowhere in the scored run.
Then judge 28 lines with no clock at all
A horizontal line appears with a vertical mark near its middle; you say which half is longer with the arrow keys or the two buttons. Seven mark positions, symmetric about the true middle, four times each in mixed order. Answer on the first look — measuring with a finger or the window edge removes the very bias the task is built to catch, and there is no time pressure to make you rush instead.
Technical specifications
| Trials | 48 flashed strings — 24 words and 24 non-words, 12 of each per visual field — plus 28 line judgments at 7 mark positions. Eight optional practice flashes are excluded from every figure |
|---|---|
| Stimulus duration | 150 ms for the flashed strings, chosen because a saccade takes longer than that to initiate: the string is off the screen before an eye movement toward it could land. The lines stay up until you answer |
| Where the strings appear | The page measures its own panel and puts the near edge of the string 56-130 CSS px from the cross, which is roughly 1.5-3.2° at a 60 cm viewing distance. The exact offset used is printed with your result, and the viewing distance is an assumption stated out loud rather than something a browser can read |
| Line geometry | The mark sits at 0, ±0.8%, ±1.65% or ±2.5% of the line's own length from the true middle — proportions rather than pixels, so a phone and a desktop run the same task. The measured line length and the resulting pixel offsets are printed with the result |
| Response | Read by physical key position, so F and J are the same two keys on QWERTY, AZERTY and Dvorak; the lines take the left and right arrows. On-screen buttons work throughout and are counted separately, because aiming a pointer costs time that a resting finger does not |
| What voids a trial | A response inside 150 ms of the string painting, a response that arrives before it paints, and any trial during which this tab went to the background — a 150 ms flash cannot survive a tab switch in any useful sense. A voided trial is asked again straight away |
| Reference figures | Strong right-handers 4-4%; Mixed-handed 15-15%; Strong left-handers 27-27% for right-hemisphere language — Knecht et al. (2000), Handedness and hemispheric language dominance in healthy humans, Brain. Group frequencies for context; this page scores you against neither of them and produces no rank |
| What leaves the page | Nothing at all unless you press the copy button, which writes plain text to your own clipboard. That text carries the two headline figures and the run seed, not the 76 individual answers, and there is no network request during a run |
Frequently asked questions
Am I left-brained or right-brained?
Neither, and the question has no measurable answer. Resting-state connectivity was examined in over a thousand people specifically to look for individuals whose whole brain leaned one way, and the lateralized networks turned up while the lateralized people did not. Particular jobs are lateralized — language usually left, some spatial attention usually right — and they are lateralized in the same direction in the same person at the same time, which is precisely what this page shows you by running one task of each kind.
Why does the string disappear before I have read it?
Because a string you have time to look at is no longer telling you anything about hemispheres. Presenting to one visual field only works while the eyes are still, and an eye movement takes about 200 ms to plan and launch, so 150 ms is the standard compromise: long enough to identify a short familiar word, short enough that your gaze cannot get there. Versions of this test that leave the word up for a second are measuring reading, with the word landing in both hemispheres, and the difference they report between sides is mostly noise.
My right-field advantage came out negative. Does that mean my language is on the right?
Almost certainly not, and a browser cannot tell you either way. The half-field method does index cerebral language dominance when it is run properly — with many trials, controlled eccentricity and a fixation check — and the case for it was made against imaging in exactly those conditions. This run has 24 word trials, no way to verify you held fixation, and no control of how far you sat from the screen, so a reversed or flat difference on any single run is well inside what noise produces. Run it twice on different days before you find it interesting.
Why is the second task about lines instead of something creative?
Because the creative half of the folk story has no task behind it, and line bisection does. Give people a horizontal line and ask them to mark or judge its middle and the average adult errs leftward — a bias documented across a large meta-analysis of these studies and attributed to the right hemisphere carrying more of spatial attention. It is unglamorous, it is measurable in a browser, and it points the opposite way to the word task, which is the only property the page needs from it.
Does being left-handed change what I should expect here?
It shifts the group frequencies and settles nothing about you. Measured across 326 adults chosen to span the handedness range, language activation was stronger on the right in 4% of strong right-handers, 15% of mixed-handers and 27% of strong left-handers — so handedness moves the odds substantially and still leaves nearly three-quarters of strong left-handers with left-lateralized language. This page never asks your handedness, because knowing it would not let it say anything more about your own two numbers than it already can.
Can I do this with one eye, or with a visual field defect?
The line task, yes; the word task, no, and it says so before you start rather than after. Flashing to one side of fixation measures which hemisphere received the string first, and it only does that when both halves of your field of view are delivering normally. With a scotoma, a patch or an untreated squint, the difference between the sides is a measurement of the field defect wearing a hemisphere label. The start screen therefore offers the line task on its own as a first-class route rather than as a consolation.
Why are there non-words at all, and why do they look almost real?
Because without them you could answer every trial correctly without reading anything. If everything on screen were a word, pressing the word key 48 times would score 100%, and the reaction times would be a measurement of how fast you can press one key. Non-words force the decision, and near-misses like BREAP and STOME force it to go through the lexicon: a random consonant string is rejected on sight, but something one letter from BREAD has to be looked up before it can be turned down. The page also reports a d-prime per field, which separates actually telling them apart from a habit of pressing one key.
What is really lateralized, and what the quiz version borrowed from it
The folk version has a real ancestor. Split-brain surgery in the 1960s severed the corpus callosum in a handful of patients with intractable epilepsy, and the resulting experiments showed something genuinely startling: with the hemispheres disconnected, a word flashed to the left visual field could be acted on by the left hand and not spoken about, because speech lived on the other side of a cut. That work is sound, and what it established was that specific operations have addresses. What it did not establish — what nothing has established — is a person-level trait built out of those addresses. The quiz you were probably looking for asks whether you prefer plans or surprises and returns a hemisphere; there is no path from that question to that answer.
The two tasks here are the honest remains of the idea. Flashing a string to one side of fixation is the oldest behavioral handle on cerebral asymmetry there is — the retinal locus effect was published in 1952 — and it works because each half of the visual field projects first to the opposite hemisphere, so a word landing right of your gaze reaches the left hemisphere's language machinery a synapse sooner. The half-field method holds up against imaging when it is run carefully, which is a real qualification and one this page is in no position to meet: it cannot see whether you held fixation, cannot know how far you sat from the screen, and gives you 24 word trials rather than several hundred. The line task points the other way for a different reason — bisection errors run leftward on average across a large body of studies, which is read as the right hemisphere doing more of the work of distributing attention across space. Face recognition is the third classic entry on that list and the one with the clearest everyday consequence when it goes wrong, which is why it gets a test of its own. Neither number here is a diagnosis of where anything sits in you. Together they are a demonstration that “which side” is a question about a job, not about a person.
It is worth naming what the story is usually reaching for, because there is a real individual difference buried under it. When somebody says they are right-brained they generally mean something about imagery, or about how they think when they are not using words — and imagery does vary enormously between people, in a way that has a name and a measurement: aphantasia, reported by 1-4% — Zeman, Dewar & Della Sala (2015), is the absence of voluntary visual imagery and has nothing to do with hemispheres. Synesthesia is another such difference, and it is measured by consistency rather than by preference. The same substitution runs through this corner of the internet: a stable, interesting fact about a person gets re-labeled as a hemisphere, or as a learning style, or as time blindness. The label travels; the measurement usually does not. And if you want to see the spatial-attention system this page's second task leans on measured directly rather than as a bias, that is what the Posner cueing task does.
Why there is no score for a dominant hemisphere
There is nothing to norm. Imaging of over a thousand people found no evidence that individuals have a globally dominant hemisphere; particular functions are lateralized, people are not. Any quiz assigning someone a dominant side is assigning a horoscope.
Nielsen, Zielinski, Ferguson, Lainhart & Anderson (2013), An evaluation of the left-brain vs. right-brain hypothesis with resting state functional connectivity magnetic resonance imaging, PLOS ONE
Right-hemisphere language dominance by handedness: Strong right-handers 4-4%; Mixed-handed 15-15%; Strong left-handers 27-27%
Knecht et al. (2000), Handedness and hemispheric language dominance in healthy humans, Brain
These are group frequencies for one function measured with blood-flow imaging, not a scale anybody sits on. They are worth knowing for the opposite of the usual reason: language is the most reliably lateralized thing the brain does, and even it is on the right in a quarter of strong left-handers. Nothing on this page detects which side yours is on, and a browser task cannot.
Sources for the two tasks: Mishkin & Forgays (1952), Word recognition as a function of retinal locus, Journal of Experimental Psychology; Hunter & Brysbaert (2008), Visual half-field experiments are a good measure of cerebral language dominance if used properly, Neuropsychologia; Jewell & McCourt (2000), Pseudoneglect: a review and meta-analysis of performance factors in line bisection tasks, Neuropsychologia.
A reaction time here is the interval between the frame that painted the stimulus and the timestamp the browser attached to your key, both read from the same monotonic clock. What neither can see is the display pipeline behind it, so on a 60 Hz screen roughly 16 ms of every figure below is the machine rather than you. That is the timing floor: two numbers closer together than that are the same number, and this page reports no precision it cannot support.
The word task is where that floor bites. Its headline is a difference between two of your own conditions, so the display and the keyboard largely subtract out of it — but a difference of a few milliseconds between the sides is smaller than one frame, and this page will not dress such a figure up as an asymmetry. The line task carries no clock and no floor.
This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish a language disorder, hemispatial neglect, or an unusual arrangement of your own hemispheres. Only a qualified professional, working with more than a browser, can make that judgment.
Where your 76 answers are worked out
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Your answers, your reaction times and your score are never uploaded, logged or kept — which is also why the test carries on working after you disconnect from the network, and why nothing here can be held back behind an email address.
The two blocks keep their records in one JavaScript array inside this tab, and the array is read exactly twice: once to subtract one field's mean from the other's, once to count how often you chose the left half. Nothing is written to storage and nothing is sent anywhere, which is why reloading loses a finished run and why the copy button exists.