Guide
Fluid vs crystallized intelligence
In 1963 Raymond Cattell split general ability into two broad kinds: fluid intelligence, reasoning with problems you have never seen, and crystallized intelligence, the knowledge and skill you have built up. The split explains a lot, from why different tests disagree to why some abilities peak earlier in life than others.
Want your own number? The test is 25 visual puzzles, about 12 minutes, no sign-up.
Two kinds of ability
Raymond Cattell distinguished fluid intelligence, reasoning with novel material, from crystallized intelligence, acquired knowledge, in 1963 (Journal of Educational Psychology, volume 54) (Wikipedia: Fluid and crystallized intelligence). The idea built on earlier work on a single general factor. Charles Spearman had proposed in 1904 that one general ability, g, runs through performance on many different mental tests (Wikipedia: g factor). Cattell's point was that this general ability has at least two broad parts that behave differently.
A simple way to hold the difference: fluid intelligence is what you use when you do not know what to do, and crystallized intelligence is what you use when you do. Working out the rule in an unfamiliar puzzle is fluid. Recalling a word's meaning, or applying a method you learned at school, is crystallized.
What fluid intelligence looks like on a test
Fluid items present something new and ask you to find structure in it. The classic example is a matrix puzzle: a 3x3 grid of figures with the last cell missing, where the shapes change by rules that run across rows and down columns. Raven's Progressive Matrices, first published in 1938, are non-verbal matrix puzzles widely used as a measure of fluid, culture-reduced reasoning (Wikipedia: Raven's Progressive Matrices).
Other common fluid formats include:
- Pattern sequences: which tile comes next?
- Number series: which number continues the rule?
- Figure analogies: A is to B as C is to what?
- Odd one out: which figure breaks the shared rule?
What these have in common is that your knowledge does not hand you the answer. You must notice, try and check rules on the spot, while holding several of them in mind.
What crystallized intelligence looks like on a test
Crystallized items draw on what you have learned. Vocabulary questions, general knowledge, and word problems that rely on familiar methods all lean this way. They depend partly on education, language and reading habits, which is one reason nonverbal tests exist: they try to reduce that dependence. Our guide to culture-fair IQ tests covers what that approach can and cannot promise.
Crystallized ability is not a lesser kind of intelligence. Expertise in any field is mostly crystallized: a doctor recognizing a pattern of symptoms or a mechanic hearing what is wrong with an engine is using knowledge built over years. It is simply harder to measure fairly across people with different backgrounds.
Try one · Abstract Reasoning
Which tile completes the grid?
Tap the option you think is right, or show the answer.
Show the answer
The missing tile has one solid circle (option 3).
How to solve it
- Start with the first row and name what changes from tile to tile.
- Every row and every column contains each shape (circle, triangle and diamond) exactly once.
- Every row and every column contains each number of shapes (two, one and three) exactly once.
- The fill (solid) stays the same in every tile.
- Check the rules on the second row, then apply them to the third.
- So the missing tile has one solid circle.
That was 1 puzzle. The full test has 25.
How the two change with age
The two kinds of ability follow different paths across adult life. Large studies find that processing speed and some reasoning measures peak in the late teens and twenties, while vocabulary and accumulated knowledge keep rising into the sixties (Hartshorne & Germine, 2015) (Salthouse, 2009). In plain terms, many people get somewhat slower at brand-new puzzles over the decades, while knowing more and more.
That does not mean an IQ score falls with age. Modern IQ tests compare you only with people your own age, with the mean set to 100 in every age group (Wikipedia: Intelligence quotient). A typical age-related change in raw speed does not lower your score by itself, because your peers have changed in similar ways. Our guide to IQ and age goes further, and the glossary entry on age norms explains the mechanics.
Why the distinction matters for your score
Because different tests weight the two kinds of ability differently, the same person can score differently on different tests. A full clinical battery samples both. A nonverbal test leans heavily on the fluid side. If your strengths are mainly in language and knowledge, a purely visual test may undersell you; if you are a strong puzzle solver who reads little, a vocabulary-heavy test may do the same.
The two are related, though. People who do well on fluid tasks tend to do somewhat better on crystallized ones too, which is the overlap Spearman's g describes. Read more in what is the g factor?
Which one IQ Octopus tests
IQ Octopus is a fluid-style test. Its 25 puzzles are visual and number puzzles in five sections: Pattern Recognition (5), Abstract Reasoning (9), Spatial Reasoning (5), Quantitative Reasoning (3), Logical Reasoning (3). There is no vocabulary, no general knowledge and no reading beyond the instructions. Puzzles are generated fresh for every attempt from rule templates, so remembering answers does not help.
That design is a deliberate trade. It keeps the test short and less tied to language and schooling, but it means the estimated IQ describes reasoning with new visual and numerical material, not everything a full clinical test samples. The how it works page explains the method and its limits.
Can either kind be trained?
Knowledge plainly grows with learning; that is what crystallized means. For fluid ability, targeted training has a weak record. A 2016 meta-analysis of 87 publications found no convincing evidence that working memory training improves far-transfer measures such as nonverbal reasoning, compared with an active control group (Melby-Lervåg et al., 2016). Schooling has a better record: about 1 to 5 IQ points per additional year of education (Ritchie & Tucker-Drob, 2018). Also well established: retaking cognitive ability tests raises scores by about a quarter of a standard deviation on average, roughly 4 IQ points, and more when the same form is reused (Hausknecht et al., 2007). Some of that is familiarity with the format rather than a change in ability. Our guide on how to prepare for an IQ test covers what practice can and cannot do.
Try both ideas on real puzzles
If you want to feel the difference, work through a few matrix puzzles. Notice that none of them can be solved by remembering anything; each needs a rule found on the spot. Then compare that with how you would answer a vocabulary question: you either know the word or you do not. The glossary entries for fluid intelligence and crystallized intelligence give short definitions to keep.
Sources
- Fluid and crystallized intelligence. Wikipedia. en.wikipedia.org/wiki/Fluid_and_crystallized_intelligence
- g factor (psychometrics). Wikipedia. en.wikipedia.org/wiki/G_factor_(psychometrics)
- Hartshorne, J. K., & Germine, L. T. (2015). Psychological Science, 26(4), 433-443. doi.org/10.1177/0956797614567339
- Salthouse, T. A. (2009). Neurobiology of Aging, 30(4), 507-514.
- Raven's Progressive Matrices. Wikipedia. en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
- Hausknecht, J. P., Halpert, J. A., Di Paolo, N. T., & Moriarty Gerrard, M. O. (2007). Journal of Applied Psychology, 92(2), 373-385. doi.org/10.1037/0021-9010.92.2.373
- Intelligence quotient. Wikipedia. en.wikipedia.org/wiki/Intelligence_quotient
- Melby-Lervåg, M., Redick, T. S., & Hulme, C. (2016). Working memory training does not improve performance on measures of intelligence or other measures of "far transfer." Perspectives on Psychological Science, 11(4), 512-534. doi.org/10.1177/1745691616635612
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358-1369. doi.org/10.1177/0956797618774253