← Blog

How do IQ tests work?

Updated 2026-11-18

An IQ test gives you a set of reasoning tasks, counts how many you solve, and then compares that result with the results of a large group of people your own age. The IQ is not a count of correct answers. It says where you stand relative to that group: 100 is exactly average, and every 15 points is one standard deviation away from it.

What an IQ test asks you to do

Tests differ, but most draw on the same few families of task. Each one targets a different part of thinking, and a full test combines several so that no single skill decides the result.

Pattern and series tasks mostly tap what psychologists call fluid ability, the capacity to reason through something new. Verbal tasks lean on crystallized ability, which is accumulated knowledge. The difference is explained in fluid vs crystallized intelligence.

From raw score to IQ

The number of points you collect is called the raw score. On its own it means little: 32 out of 40 could be excellent or ordinary, depending on how hard the items are. To give it meaning, the publisher first gives the test to a large sample of people chosen to resemble the general population. This sample is the norm group.

Your raw score is then compared with the people in the norm group who are the same age as you. Age matters because raw performance changes over a lifetime. A ten-year-old and a thirty-year-old with the same IQ did not solve the same number of items; each did equally well compared with their own age group. More on this in IQ and age.

Finally the comparison is put on a standard scale. Most modern tests set the mean at 100 and the standard deviation at 15. Scores follow a bell curve, so most people land near the middle and few at the ends.

Score rangeShare of people
85 to 115about 68%
70 to 130about 95%
above 130a little over 2%

An IQ of 115 therefore means you did better than roughly 84% of the people your age in the norm group. IQ percentiles are often the clearer way to read a result. Some older scales use a different standard deviation, so the same position can carry a different number.

The original formula, mental age divided by actual age times 100, is no longer used. Modern tests compare you with other people of the same age instead.

What makes a test reliable

Reliability means consistency. If the same person takes a reliable test twice, a few weeks apart, the two scores are close. The different parts of the test also agree with each other instead of pointing in different directions.

No test is perfectly reliable. Sleep, mood, distraction and lucky guesses all move a score a little. This is why professional reports give a range of a few points around the score, not a single exact figure. A difference of two or three points between two people, or between two attempts, tells you nothing.

Length helps. Each item is a small sample of what you can do, and more samples average out luck. A ten-question quiz cannot be as reliable as a test with several dozen items across different task types.

What makes a test valid

Validity means the test measures what it claims to measure. A bathroom scale that always shows two kilos too many is reliable but not valid: consistent, and wrong. For an IQ test, evidence of validity comes from several directions:

Validity also has limits. An IQ test measures certain kinds of reasoning under time pressure. It does not measure creativity, motivation, practical judgement, kindness or how well you deal with people.

Clinical tests and self-administered tests

The tests with the strongest evidence behind them, such as the Wechsler scales and the Stanford-Binet, are given one to one by a trained psychologist. A self-administered test, online or on paper, is something you take on your own.

Supervised clinical testSelf-administered test
Who runs itA trained psychologist, one to oneYou, on your own
ConditionsQuiet room, fixed instructions, exact timingWhatever your surroundings happen to be
TasksMany subtests, including spoken and hands-on onesMostly what works on a screen: patterns, numbers, words
LengthOften an hour or moreUsually much shorter
NormsLarge, documented samplesVaries, and often not documented
ResultA report that interprets scores in contextA number or range, to be read as an estimate

A self-administered test gives an estimate. That applies to the test on this site as well. It can tell you roughly where your reasoning stands, but it is not a clinical assessment and cannot diagnose anything. How far such an estimate can be trusted is the subject of are online IQ tests accurate. If you need a score for a formal purpose, talk to a qualified psychologist.

What the number tells you in the end

An IQ score is a comparison, taken on one day, on one set of tasks, against one norm group. Read that way it is useful: a rough indication of how you handle abstract problems compared with other people your age. Read as a fixed measure of your worth or your potential, it claims more than any test can deliver.

Keep reading

Are online IQ tests accurate?

A well-built online IQ test can estimate your reasoning ability, but it is not a clinical assessment. What limits accuracy and how to spot a fair test.

Can you increase your IQ?

IQ scores can move by a few points, but general intelligence is hard to change. What has good evidence, what has weak evidence, and what simply helps.

How to prepare for an IQ test

What helps on the day of an IQ test: sleep, a quiet place, reading the instructions, knowing the question types. And what does not: cramming and tricks.