What Does an IQ Test Consist Of? Inside a Full Test

On this page

1. What is an IQ test?
2. What does an IQ test measure?
3. What does an IQ test consist of, part by part?
4. How does an IQ test work on the day?
5. How is an IQ test scored?
6. How much testing time does each battery need?
7. What does the final score summarize?
8. How is an online reasoning test different?
9. What can an IQ test not tell you?

An IQ test consists of a set of short subtests, each sampling one kind of thinking. Some test reasoning with new problems, others knowledge of words and facts, visual and spatial thinking, working memory or speed. The results are combined into a few index scores and one full-scale IQ. Every score is read against people of the same age.

What is an IQ test?

An IQ test is a standardized test of how well a person solves problems, forms concepts and reasons. The American Psychological Association describes it as an individually given test of mental, verbal and performance tasks of graded difficulty. Those tasks are standardized on a representative sample of the population¹. The Stanford-Binet and the Wechsler Adult Intelligence Scale are its two standard examples.

Two words in that definition carry most of the weight. “Standardized” means everyone gets the same tasks, instructions and scoring rules. “Representative sample” means the test was first given to a large group chosen to mirror the population, and those results became the yardstick. Without that group, a raw count of right answers would have nothing to be compared with.

What does an IQ test measure?

It measures performance on a range of thinking tasks, chosen to sample abilities that psychologists treat as parts of general intelligence. One influential account, the Cattell-Horn theory, separates two main kinds²:

• Fluid ability: the reasoning you use on unfamiliar problems, where nothing you learned gives the answer away.
• Crystallized ability: what you have learned, measured with tasks such as vocabulary and general information.

Later versions of the theory added more abilities, among them visual ability, the kind used to turn a shape in your mind². Modern batteries are built around similar families. The WAIS-5 reports 5 primary indexes: verbal comprehension, visual spatial, fluid reasoning, working memory and processing speed³. The SB5 is organized around 5 factors: fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory⁴.

Seen as measurement, then, an IQ test consists of layers. Each subtest measures one narrow task, each index summarizes a family of tasks, and the full-scale IQ summarizes the lot.

What does an IQ test consist of, part by part?

A full test is a battery: a set of subtests whose scores are combined in layers. A subtest is a separate part of a test with its own identifiable content¹. The two best-known batteries are put together like this:

Sources: Pearson’s WAIS-5 page³ and PRO-ED’s SB5 page⁴.

Notice the design choice in the SB5: every ability is measured twice, once with words and once without. Half of the test needs little or no speech from the person being tested, which the publisher points to for people with limited English or hearing loss⁴. The WAIS-5 takes a different route and reports separate visual spatial and fluid reasoning indexes, which the WAIS-IV combined into one³.

We do not reproduce or describe the actual items of published tests. If you want to see what reasoning puzzles look like, our IQ test questions are original examples of the five kinds we use, with worked answers.

How does an IQ test work on the day?

In a clinical setting, an examiner sits with you and gives the subtests one after another, reading set instructions and recording answers. Items within a subtest get harder as you go, and the test adapts to where you are. On the SB5, for example, a matrices subtest and a vocabulary subtest come first, and their results set the starting level for the nonverbal and verbal sections⁴. That keeps the session from wasting time on items far too easy or far too hard.

The WAIS-5 can be given with paper materials or on Pearson’s digital system, but always one to one with an examiner³. Pearson sells it only to qualified professionals, at its level C, and PRO-ED lists the SB5 as a level C test too⁴. That is the main reason a real WAIS or SB5 is never available as a free download. Our WAIS guide describes the adult Wechsler test in more detail. The Stanford-Binet has its own page.

How is an IQ test scored?

Scoring runs from raw scores up to the full-scale IQ, and at every step the comparison is with people your own age.

1. Raw score. Each subtest gives a count of points.
2. Scaled score. The norms turn that count into a scaled score. On the SB5, subtest scaled scores have a mean of 10 and a standard deviation of 3⁴.
3. Composites. Scaled scores are added up into index scores and the full-scale IQ, which use a mean of 100 and a standard deviation of 15⁴⁶.

The APA defines a norm as the typical performance of a group, such as an age group, used for comparison. The same word covers the conversion of a raw score into something easier to read, such as a percentile or an IQ¹. That is why the same number of right answers can mean different things at 20 and at 70.

The last step is a range. Every score has a standard error of measurement, the amount it would wobble across repeated testing⁶. In Pearson’s WAIS-5 sample report, a full-scale IQ of 111 carries a 95% range from 106 to 116⁵. The IQ scale lists the score bands and their labels.

How much testing time does each battery need?

The size of the battery sets the testing time, so a short reasoning test and a full clinical scale sit far apart:

A clinical assessment can add more than testing: an interview beforehand, and a feedback meeting and written report afterward. The whole appointment, its cost and how to book one are covered in how much an IQ test costs and how long it takes.

What does the final score summarize?

The full-scale IQ is a summary of everything in the battery, placed on a scale where 100 is the middle of your age group⁶. It is useful precisely because it averages over many tasks, so one off subtest moves it only a little.

The same averaging hides detail. Two people with a full-scale IQ of 105 can have quite different profiles, one strong in verbal comprehension and slow in processing speed, the other the reverse. That is why clinical reports list the index scores next to the total, and why a psychologist reads the pattern, not just the headline number.

How is an online reasoning test different?

An online test like ours is not a battery. It has 30 questions in up to 25 minutes, all reasoning puzzles: 8 matrices, 6 series, 6 rotation, 5 logic and 5 math questions. It has no vocabulary or general-knowledge tasks, no memory or speed subtests, and no examiner.

That makes it closer to a single, broad reasoning measure than to a full-scale IQ. While its norms are provisional, its scale comes from a statistical model. A result of 100 usually comes with a 90% range of about 88 to 112, wider than on a clinical test. Unsupervised tests in general also read slightly high, by about 3 points in a large review⁷. Our methodology explains how our questions are generated and scored.

What can an IQ test not tell you?

An IQ test describes performance on its tasks, on one day, compared with people your age. It does not measure character, creativity, effort or knowledge outside its tasks, and a single score is never a diagnosis on its own. For decisions that matter, such as school support or a clinical question, the test is one part of a professional assessment, not the whole of it. How much weight any one score can bear is the subject of are IQ tests accurate.

If you are curious where you stand on reasoning alone, you can take the free IQ test in one sitting and read your result with its range.