Fluid intelligence test: reasoning with rules you have never seen

In the 1940s the psychologist Raymond Cattell split general intelligence into two broad parts, and his student John Horn developed the idea. Fluid intelligence, often written Gf, is reasoning with new material: spotting patterns, forming rules and solving problems you have no recipe for. Crystallised intelligence, Gc, is what that reasoning leaves behind: vocabulary, facts and learned methods.

The two are related. People high on one tend to be above average on the other, because fluid reasoning is how knowledge gets built. Modern models of intelligence, such as the Cattell-Horn-Carroll model, keep both as broad abilities under a general factor.

This test is deliberately one-sided. Every item is a picture puzzle with a rule to discover, so it leans almost entirely on Gf. That is what makes it a fluid intelligence test rather than a general knowledge quiz.

Each item has a rule you probably have not met in exactly this form. That is the point: fluid reasoning is what you use when there is no method to remember. These are practice items for this page, not test questions.

Fluid reasoning develops fast through childhood and the teens, peaks in early adulthood and then declines slowly. Timothy Salthouse, who has studied this for decades, found that some measures of reasoning and speed begin to fall in healthy, educated adults in their twenties and thirties, long before anyone would notice in daily life.

Crystallised knowledge tells a different story. Vocabulary and general knowledge usually keep rising into middle age and hold up well into later life. That is why older adults often do better than younger ones on a knowledge test and worse on a timed puzzle test, even when they are equally sharp.

One honest note: this test scores everyone on the same scale and is not adjusted for age. If you are over 50, a professionally normed test would compare you with people your own age, and your standing there would usually be higher than a raw comparison suggests. The IQ test for seniors page explains this in detail.

Fluid reasoning is what you lean on when experience runs out. A few everyday examples:

• Learning a new system. A new job, a new app, the rules of an unfamiliar game. The first week leans on fluid reasoning; after that, experience takes over.
• Spotting what is odd. Noticing that a set of figures does not add up, or that a plan has a hidden flaw.
• Solving a problem with no manual. Fixing something you have never fixed, or planning a route with a new constraint.
• Transferring an idea. Seeing that a trick from one area solves a problem in another.

This is also why fluid reasoning relates to how quickly people learn on the job, especially early on. Over time, knowledge and practice carry more of the load, which is one reason experience can make up for a dip in puzzle speed at work.

In 2008 a study by Susanne Jaeggi and colleagues reported that training on a working memory game, the n-back task, raised scores on a fluid reasoning test. It made headlines and helped launch the brain-training industry.

Later research was less encouraging. Meta-analyses, including one by Monica Melby-Lervåg and Charles Hulme in 2013, found that working memory training improves the trained task and similar tasks, but shows no convincing lasting gain on fluid reasoning itself. The current consensus is that you get better at the game, not at reasoning in general.

• What does help a test score: sleep, a quiet room and familiarity with the item types. Practice items like the ones above can add a little on a first retest.
• What helps reasoning over a lifetime: education. A large 2018 meta-analysis by Stuart Ritchie and Elliot Tucker-Drob estimated that each extra year of schooling adds roughly one to five IQ points.
• What protects it with age: the same things that protect the rest of the body and brain, such as physical activity and good health. The evidence here is mostly about slowing decline, not boosting a score.