AbilityBenchFree, 16 ratings + 32 scored clock pairs
Aphantasia test with a second half you cannot rate
Rate how much of a picture arrives for 16 ordinary details, get a total on the 16-to-80 span the published cutoffs are defined on, and then take 32 clock comparisons whose answers are arithmetic — which of 4:40 and 9:15 puts its hands closer together. The two results are printed beside each other, never merged, because one is your description of your own experience and the other is a score. Free, about eight minutes for both, no account, and no percentile anywhere, because no published mean and standard deviation exists for either half.
- 100% free
- No signup
- 16 items, 16-80 total
- 32 scored pairs
- No percentile invented
Two halves: 16 ratings, then 32 pairs with a right answer
The first half asks you to picture four ordinary things and rate how much arrives, on a five-point ladder. It takes about three minutes and it is entirely your own report of your own experience. The second half asks something you cannot rate: two times appear as digits, and you pick the one whose hands sit closer together. The angles are arithmetic, so this page already knows every answer.
Half one — 16 vividness ratings
Four scenes, four details each. Totals run 16 to 80 and a high total means vivid imagery — the published questionnaire originally ran the other way round, so the direction has to be said out loud.
Half two — 32 clock pairs
Which of 4:40 and 9:15 puts its hands closer together? Half the pairs are within 26° of each other and half are more than 70° apart, so the run also shows whether the close calls cost you time.
The sixteen items are written for this page. The questionnaire this query usually means is a published instrument; its items and its anchor sentences belong to it and none of them is reproduced here. What carries across is the structure — four scenes, four details, a five-point ladder, a total on the same 16-to-80 span — and the recruitment cutoffs that research groups use on that span, cited in full beside your result.
Either half stands on its own and either can be taken first. Both are reported separately at the end, because a total from one is not evidence about the other.
How to take both halves
The questionnaire measures what you report; the clocks measure what you can do.
Rate four scenes, four details each
Each scene names something ordinary — a person you see daily, a lemon being halved, your own front door, a walk you know — and asks for four details from it: one still, one about color, one about where things sit, one that has to move. Answer for how much of a picture actually turns up, not for how well you know the thing. Knowing your door is oak while no oak appears is a 1, and it is a common answer. All four details open the next scene.
Compare 32 pairs of times
Two times appear as digits and you press F or J for the one whose hour and minute hands sit closer together. Sixteen pairs are within 26 degrees of each other and sixteen are more than 70 degrees apart, shuffled so no more than three of the same kind run together. Four untimed practice pairs come first and name the right answer afterwards; they are excluded from every figure. The hour hand moves as the minutes pass, which decides a good share of the pairs.
Read the two results apart, then together
The left panel gives your total, the band it falls in and the citation for the cutoffs. The right gives how many of the 32 you got right, a 95% interval on that proportion, the probability of doing that well by guessing, and how much longer the close pairs took than the wide ones. A paragraph underneath says what each combination of the two does and does not license — including that a good clock score cannot overrule a low total.
Technical specifications
| Questionnaire | 16 items written for this page, in 4 scenes of 4, rated 1 to 5. 5 is "as sharp as looking" and 1 is "nothing appears", so a HIGH total means vivid imagery — the 1973 instrument scored it the other way round and the two conventions give opposite readings of the same number |
|---|---|
| Total span and bands | 16 to 80. Bands printed as Aphantasia range 16-32 VVIQ points; Typical range 33-74 VVIQ points; Hyperphantasia range 75-80 VVIQ points, from Marks (1973) with the 32 and 75 cutoffs from Zeman et al. (2020). These are the thresholds research groups recruit on, not a boundary anyone crosses |
| Objective task | 32 scored comparisons of two clock times shown as digits, plus 4 untimed practice pairs. Chance is 50%, so the results panel prints the exact one-sided binomial probability of the score against a coin flip rather than leaving 20 of 32 looking like a result |
| Angle arithmetic | Hour hand at 30 degrees per hour plus 0.5 per minute, minute hand at 6 per minute, folded to 0-180. Times fall on 5-minute marks; pairs whose angles differ by under 8 degrees are excluded as coin flips, and the two bands are 8-26 degrees apart and 70-170 degrees apart |
| Reaction time | Measured from the frame that painted the two times to the timestamp the browser put on your key, so on a 60 Hz screen about 16 ms of every figure is the display. Answers under 150 ms are discarded as anticipations, a key pressed before the times appear is discarded as premature, and a pair the tab was hidden during is requeued |
| Response window | 12 seconds per pair, which is long enough that a slow comparison is not truncated into an omission. The median correct comparison usually lands between 1.5 and 4 seconds, and the near-minus-far difference is reported with a 95% interval so a 40 ms gap is not read as an effect |
| Percentile | None, on either half. A percentile needs a published mean and standard deviation from a sample tested the way you were tested, and neither half has one — the questionnaire's figures are recruitment cutoffs and the clock task has no browser norm at all. The raw score and the reference range are printed instead |
| What leaves the page | Nothing. The 16 ratings and the 32 trial records are held as JavaScript objects in this tab, summarized once, and never written to storage or a URL — which is why a reload starts you over and why the copy button exists |
Frequently asked questions
Is this the VVIQ?
No — the sixteen items here were written for this page, and none of them is a rewording of a published one. What carries across is the format that makes the published cutoffs readable: four scenes, four details each, a five-point vividness ladder, and a total on the same 16-to-80 span. The anchor sentences are ours too, which matters more than it sounds: how the five points are worded moves totals by several points on a scale whose lowest band is seventeen points wide, so the band this page reports is an orientation and not a score from that questionnaire.
Why does a high score mean vivid imagery here when other sites say the opposite?
Because the original scoring ran downward and most online versions have since flipped it, and both are still in circulation. In the 1973 direction, 1 meant perfectly vivid and 5 meant no image, so a low total was the vivid one; the cutoffs everybody now quotes — 32 and below for the aphantasia range, 75 and above for the hyperphantasic one — are stated in the reversed direction, where high means vivid. A page that does not say which way its own scale runs is handing you a number that means the opposite thing on half the sites you compare it against. This one runs high-is-vivid, and says so on the item screen, in the results panel and in the copied summary.
I scored in the lowest band. Does that mean I have aphantasia?
It means sixteen of your ratings summed to a number inside the range research groups use when they are recruiting people to study, which is a description of an answer pattern rather than a finding about you. Two things keep it from being more than that. The cutoff is a recruitment convenience — it was chosen to get a clean sample, not to divide a population, and the trait it sits on is continuous with no gap at 32. And the whole measurement is self-report: it depends on what you took the anchor words to mean, and there is no external check on that inside the questionnaire. The clock half exists precisely because there is nothing else to check it against.
What does the clock task have to do with imagery?
It asks for a spatial fact about something that is not in front of you, and it has a right answer. Paivio's 1978 finding was not that people can compare mental clocks — it was that the time they take rises as the two angles get closer together, which is the pattern you get from inspecting a representation and not the pattern you get from looking a fact up. That makes it usable as a check on a questionnaire, because it needs no self-description at all. It is deliberately not a picture task: nothing is drawn during the run, so what is on screen is two four-character strings.
My ratings were low but I got most of the clock pairs right. Which one is true?
Both, and the combination is the commonest one among people who arrive at this word. Object imagery — the picture, its color, its texture — and spatial imagery — where things are relative to each other — come apart, and Kozhevnikov, Kosslyn & Shephard (2005), Spatial versus object visualizers: a new characterization of visual cognitive style, Memory & Cognition characterized them as two separable styles rather than two amounts of one thing. People who report no voluntary visual imagery routinely perform in the ordinary range on spatial tasks, so a solid clock score does not weaken a low total. It locates it: it says the difficulty is with generating a picture, not with knowing where things sit.
Can I pass the clock half without imagining anything?
Yes, and that is the honest limit of it. The angles can be computed — thirty degrees an hour, six a minute, and remember that the hour hand has moved — and someone who does the arithmetic will score well without generating any image at all. This is why the page reports the near-minus-far time difference beside the accuracy rather than accuracy alone: a computed answer takes roughly as long whether the two angles are close or far apart, while a compared one takes longer as they converge. Neither figure is proof of a strategy, and the results panel says so rather than reading a flat difference as a finding about you.
Does having no visual imagery affect memory or dreams?
The research says the picture is patchier than the headlines, and this page cannot add to it. People reporting no voluntary imagery describe reduced autobiographical recall — remembering that a holiday happened without being able to re-see any of it — while semantic memory, the facts, is unaffected. Involuntary imagery is the striking part: a sizeable share of people who cannot summon an image on demand report visual dreams anyway, which suggests the machinery is present and the voluntary route to it is not. None of that is measured here, and no number on this page speaks to it.
What a vividness rating can and cannot establish
The word is fifteen years old and the observation is a century older. Francis Galton polled his colleagues in 1880 about their breakfast table and found some of them reporting nothing at all, then filed it and moved on; the trait had no name until Zeman and colleagues gave it one after a patient lost his imagery following a cardiac procedure. The prevalence figure that now travels everywhere — 1-4% of adults — comes from that line of work, and it is a range rather than a decimal for a reason worth stating plainly: it is a count of people below a questionnaire cutoff, so it moves when the cutoff moves. Full citation, since these things get repeated: Zeman, Dewar & Della Sala (2015), Lives without imagery — congenital aphantasia, Cortex. Anyone quoting a single precise percentage for this is quoting one study’s threshold as though it were a census.
The questionnaire everybody means has been in use since 1973, and its structure is what makes the bands legible: sixteen ratings, four scenes of four, a five-point ladder from no image to as-vivid-as-seeing. Two things about it trip people up. The first is direction — the original ran downward, most online versions run upward, and the cutoffs in circulation assume upward, so the same raw total means opposite things depending on the page. The second is that the instrument was designed to be given twice, once with the eyes open and once closed, and the two administrations do not always agree; a single pass, like the one here, is the common modern shortening and it loses that comparison. Neither the items nor the anchor sentences are reproduced on this page, because they belong to the instrument, and because anchor wording measurably shifts totals — which is the honest reason the band here is an orientation rather than a score from that questionnaire.
What almost every version of this test on the web has in common is that it stops there, and a self-report has a ceiling no amount of careful phrasing raises: it measures how you describe an experience nobody else can see, using words whose meaning you had to guess at. So the second half is Paivio (1978), Comparisons of mental clocks, Journal of Experimental Psychology: Human Perception and Performance’s clock comparison, which asks for a spatial fact with a right answer and never asks you to characterize anything. The interesting part is not the accuracy but the ordering — close angles taking longer than distant ones is the signature of something being inspected — and the results panel prints both, with an interval, so a small difference is not read as a large one. The same instinct runs through the neighbors: the synesthesia test replaces “does the letter A look red to you” with whether you pick the same red three times, the face blindness test has no self-report in it at all, and the mental rotation test reports the slope of time against angle rather than an average. If you would rather have statements than tasks, the alexithymia test and the sensory processing test are next door, and the left or right brain test is the page that explains why a popular framing of this kind can be wrong outright.
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.
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 aphantasia. Only a qualified professional, working with more than a browser, can make that judgment.
Where your 16 ratings and 32 trials are added up
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.
Nothing on either half is written to local storage or to the address bar, so there is no resume and no share link to leak: close the tab and the 16 ratings and the 32 trial records are gone. The copy button is the only way anything leaves, and it puts the text on your clipboard, which is your machine.