What this Free IQ Test Online measures

The questions cover five kinds of thinking, and you get a separate score for each. Words and deduction asks you to follow a statement to what it does and does not allow. Logic and patterns asks you to find the rule behind a set of figures. Numerical reasoning works with series, matrices and relationships between quantities. Spatial visualisation covers folding, stacking and seeing an object from another side. Mental rotation asks whether a shape is the same one turned, or a mirror image of it.

Those last two are usually bundled together as a single spatial score. They are separable abilities, and someone can be strong at one and ordinary at the other, so this IQ Test keeps them apart.

Most of the questions are pictures. That is deliberate: a figure means the same thing in every language, so a score here is comparable whether you took the test in English, Japanese or Polish. Only a few items use sentences, and those were written separately in each language rather than translated.

Why this Free IQ Test is more accurate than most

Most free tests online run to between twenty and thirty-six questions. This one runs to forty, and the difference is not cosmetic. Reliability rises with length in a way that is predictable, and at forty items the internal consistency of this bank is about 0.91, which puts the margin of error near four and a half points. At twenty items the same bank would carry a margin close to seven. That is the difference between a result worth reading and a result that is mostly noise.

The second difference is breadth. A great many free tests measure one thing, usually figural patterns, and then report it as an IQ. Pattern questions are a reasonable proxy for general ability, but only a proxy, and a test built from one question type rewards whoever happens to be good at that type. This Free IQ Test draws on five domains, so an unusual strength or weakness in any one of them moves your total far less.

The third is that you are shown the breakdown. A single number hides the thing most people actually want to know, which is where they are strong and where they are not. Four sections solid and one weak is a different result from five mediocre sections, and the two can produce the same total.

The fourth is the part nobody can see from the outside. Every item was checked by enumerating the rules that fit the stimulus, not by eyeballing it. That process found and removed several questions where a second rule fitted the evidence just as well as the intended one and pointed at a different option, which is the most common defect in tests of this kind and the hardest to notice. A question with two defensible answers does not measure reasoning. It measures whether you guessed which answer the author had in mind.

None of that makes this a clinical instrument, and the next section is about what it is not.

What your result is worth

Your result is shown as a single number and as a range of about ten points around it. A Free IQ Test Online of forty questions can place you reasonably well in the broad middle of the distribution and cannot do much better than that. Anybody quoting you a single exact number from a short free test is overstating what they know.

One or two questions move the figure by several points. If you misread a question, or tapped the wrong tile once, you have already moved your own result further than the difference most people read into two different scores.

The range matters more than the midpoint. If your range is 104 to 114 and a friend scores 108 to 118, you have not been separated by this test. You have both landed in the same part of the distribution.

What an IQ Test actually measures

An IQ Test measures how well you handle a particular family of reasoning problems, compared with other people of your age. It does not measure a substance in the head, and the number is not a property of you in the way your height is. It is a position in a distribution.

The reason the exercise works at all is a finding that has held up since Charles Spearman noticed it in 1904: performance on almost any mental task correlates positively with performance on almost any other. Someone good at verbal analogies tends to be better than average at spatial puzzles too. That shared variance is called g, or general ability, and it is what a well-built test is reaching for through whatever question types it happens to use.

One hundred is the middle because it was defined to be. Scores are scaled so the average of the reference group is 100 and the standard deviation is 15. About 68 of every 100 people fall between 85 and 115, about 95 between 70 and 130, and about 2 in 100 above 130. A score of 130 does not mean thirty per cent more of anything. It means two standard deviations above the middle of the group you were compared against.

Which group that is matters enormously, and it is why tests are re-normed every decade or two. Raw performance on these tasks rose steadily through the twentieth century, roughly three points per decade across many countries, a pattern named the Flynn effect after James Flynn, who documented it. Someone scoring 100 against a 1950 reference group would score well below 100 against a modern one. The rise appears to have slowed or reversed in several countries since the 1990s. This is the single best reason to treat an IQ figure as a comparison rather than a measurement.

What an IQ Test predicts is real but partial. Scores correlate with school and job performance well enough to be useful in aggregate and badly enough to be a poor guide to any individual. They say nothing about conscientiousness, judgement, creativity, or whether somebody is good company, and those things decide most outcomes that matter.

How IQ scores change with age

A chart of IQ against age would be a flat line at 100, because scores are re-normed within each age group. A sixty-year-old with an IQ of 100 is average for sixty-year-olds, not average for twenty-year-olds. The flat line is a convention, not a finding.

The question people are really asking is what happens underneath, and there the answer is well established and not flat at all. Abilities split into two broad kinds that move in opposite directions across adult life.

Fluid ability is reasoning on problems you have not seen before, which is most of what this test asks for. It peaks in the twenties and declines slowly and then less slowly from there. Crystallised ability is what you know and can put into words, and it keeps climbing into the sixties before drifting down.

So a fifty-year-old and a twenty-five-year-old with the same IQ are not doing the same thing. The younger one is doing more of it with raw processing, the older one with accumulated knowledge and better strategy. Because this test leans on figures and patterns, it asks more of fluid ability than a vocabulary-heavy test would, which is worth knowing if you are over fifty and comparing yourself with a younger relative.

Within a single person, scores are fairly stable from the late teens onward. Across a few weeks they move by a handful of points from mood, sleep and practice alone.

Taking a Free IQ Test Online versus a professional assessment

A supervised assessment such as the WAIS or the Stanford-Binet takes an hour or two with a psychologist, uses materials that are not public, and is normed on a carefully constructed sample. It produces several index scores, it is administered under controlled conditions, and it can be used for diagnosis. It also costs money and usually requires a referral.

A Free IQ Test Online is a different thing doing a smaller job. Nobody is watching to see whether you looked something up, took it three times, or did it at two in the morning. The reference group is whoever happened to take it rather than a stratified sample. The honest claim is that it puts you in a band, quickly and at no cost, and that is genuinely worth something as long as nothing is hanging on it.

The question families here are the public ones that research instruments are also built from: figural matrices in the tradition of Raven’s Progressive Matrices, series completion, mental rotation of the kind Shepard and Metzler used, paper folding, and verbal deduction. The specific items are original. No published test is reproduced here, and any site offering you "the real WAIS questions" is either lying or infringing.

If a number is going to be used for anything, a school placement, an assessment, a diagnosis, then it needs to come from a psychologist. If you are curious about roughly where you sit, this Free IQ Test Online will tell you that in eighteen minutes.

Autism, ADHD and uneven profiles

A Free IQ Test Online is not a screening tool for anything. This one cannot detect autism, ADHD, dyslexia or a learning disability, and a low score is not evidence of any of them.

What is worth saying is that uneven profiles are common and normal. A large gap between your best and worst section is not a malfunction and not a diagnosis. Plenty of people are a standard deviation apart on verbal and spatial reasoning and go through life without the difference ever mattering.

Autistic people in particular are often poorly served by timed tests, which measure speed of adjustment alongside reasoning. This test has no time limit, which removes one source of distortion. People with ADHD tend to be affected most by sustained-attention demands rather than by the reasoning itself, and forty questions in one sitting is a real demand. Stopping halfway and coming back is fine.

If an uneven profile matches something you have noticed elsewhere in your life and you want it looked at properly, that is a conversation with a clinician, not with a web page.

What can distort your score, and why there is no timer

Short tests are sensitive to things that have nothing to do with ability. Sleep is the big one: a night of five hours costs more on a reasoning test than most people expect. Hunger, illness, alcohol the night before, a noisy room and a phone lighting up all cost points. So does taking it on a small screen, where the figures are harder to compare.

Misreading is the quietest source of error. On a forty-question test, two misread items can move the result by five points, and you will not remember having misread them.

Practice raises scores. If you have done several of these recently you will do better than someone equally able who has not, and the gain is in test-taking rather than in reasoning. This is also why a repeat score a day later is not a cleaner reading of the same quantity.

Time pressure is the one distortion this test removes outright. There is no timer. A clock on screen tells you how long you have taken, and nothing happens when it grows. Timed tests measure speed as well as reasoning, and the two are separable abilities. Two people can reason equally well and differ substantially in how fast they do it, and a clock scores that difference as though it were intelligence. Removing the timer costs a little information and removes a large source of unfairness, which is a trade worth making for a Free IQ Test that nobody is supervising anyway.

Take it once, rested, on the largest screen you have, and do not take the first number too seriously.

Before you start

This Free IQ Test Online is untimed and runs to forty questions, which takes most people about eighteen minutes.

Each question opens on its own page, so your answers are kept as you go and the back arrow takes you to the previous one. Answers live in this browser tab only. They are never sent anywhere and they are gone when you close it.

There is no account, no email address and nothing to pay, at the start or at the end.

Sources

• Spearman, C. (1904). General intelligence, objectively determined and measured. American Journal of Psychology.
• Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin.
• Salthouse, T. A. (2009). When does age-related cognitive decline begin? Neurobiology of Aging.
• Horn, J. L., & Cattell, R. B. (1967). Age differences in fluid and crystallized intelligence. Acta Psychologica.
• Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science.
• Condon, D. M., & Revelle, W. (2014). The International Cognitive Ability Resource: Development and initial validation of a public domain measure. Intelligence.
• Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence.
• Trahan, L. H., et al. (2014). The Flynn effect: A meta-analysis. Psychological Bulletin.