Culture fair IQ test: as fair as a picture test can be

In 1940 Raymond Cattell published what he called a culture-free intelligence test. The idea was simple: if a test uses only shapes and patterns, a person's language, education and background should matter less. He later renamed it the Culture Fair Intelligence Test, because "culture-free" claimed too much.

That change of name is the most useful thing to know. A culture fair test reduces the influence of culture. It does not remove it. Three kinds of evidence show why:

• Scores on picture tests rose fastest of all over the 20th century. James Flynn found the largest generational gains on Raven-style matrix tests, larger than on vocabulary tests. Genes do not change that fast, so something in the environment, such as schooling, nutrition and familiarity with abstract puzzles, was pushing scores up.
• Schooling still matters. A 2009 review of Raven's scores from many countries and decades, by Sanne Brouwers, Fons van de Vijver and Dianne van Hemert, found that differences in scores were linked strongly to education.
• Test familiarity matters. People who have grown up with puzzles, worksheets and timed tests start ahead of people who have not, even when every item is a diagram.

So the honest claim for a picture test, ours included, is this: it is fairer to people who do not share the test's language or schooling than a verbal test would be, and it is still not a neutral measuring stick for everyone on earth.

These items need no language at all, only the short instruction above each one. They use the same kinds of task as culture fair tests: series, matrices and simple arithmetic shown with shapes. They are written for this page and are not test questions.

A practical warning about Cattell scores. Because the CFIT often uses a standard deviation of 24, a Cattell score of 148 is the same rarity as 130 on most other tests. If you have a Cattell number, convert it before comparing it with anything else. The Mensa eligibility checker does this for you.

The problem is as old as IQ testing. Early tests in the United States, such as the Army Alpha used in the First World War, asked questions that depended on American schooling and everyday knowledge. Recruits who could not read English were given a separate picture test, the Army Beta, because the written test plainly measured language as much as reasoning.

In 1938 John Raven published his Progressive Matrices, built only from patterns. In 1940 Raymond Cattell published his culture-free test, later renamed culture fair, using series, classification, matrices and conditions. Both aimed to measure what Cattell later called fluid intelligence: reasoning that does not depend on what you have been taught.

Over the following decades two findings tempered the hope. Scores on these picture tests rose sharply across generations, which showed that environment shapes them. And comparisons across countries tracked schooling closely. The modern view is that picture tests are among the fairest tools available, used with care and read as estimates, rather than neutral measures that make background irrelevant.

Whatever test you take, you can remove some of the non-reasoning factors yourself.

1. Read the instruction twice. On a picture test it is the only language there is.
2. Do one or two practice items first. Familiarity with the format is a real advantage, and it is cheap to get. The three above are enough.
3. Use a screen you can see well. Small shapes on a small phone add a visual hurdle that is not reasoning.
4. Choose a calm time. Stress and tiredness lower scores for everyone, and they are not part of what is being measured.
5. Read the result as a range. Background still nudges scores. A few points either way are within the margin.