Metlivi Blog

Understand fluid reasoning without ranking people

Fluid intelligence is the ability to reason through unfamiliar problems: noticing patterns, drawing inferences, and adapting when a task does not have a ready-made answer from memory. Learned knowledge—often called crystallized intelligence in the research literature—includes information and skills acquired through education and experience. The distinction is useful for understanding different kinds of thinking, but it is not a verdict about a person: one test or online quiz cannot define someone’s intelligence.

September 24, 20264 min readTime Management & Personal GrowthBy Metlivi Editorial Team
Section 1

What does fluid intelligence mean?

Psychologists use fluid intelligence to describe reasoning in situations where a solution is not simply recalled from prior learning. A person might compare unfamiliar shapes, work out a rule from examples, or figure out how a new object works. Research often studies fluid reasoning with problems such as visual matrices, but those tasks also rely on supporting processes. A primary study published in PNAS, for example, describes fluid-intelligence tests as involving complex problems and examines how their complexity and component parts affect performance. Duncan and colleagues, “Complexity and compositionality in fluid intelligence,” PNAS

The term does not mean thinking with no influence from experience, education, or familiar problem-solving strategies. People bring language, strategies, attention, and prior familiarity to new tasks. “Unfamiliar” means that a person cannot solve the problem merely by retrieving a specific answer or rehearsed procedure; it does not mean the mind starts from a blank slate.

Section 2

How is it different from learned knowledge?

A familiar distinction in intelligence research contrasts fluid ability (often abbreviated Gf) with crystallized ability (Gc). In broad terms, fluid reasoning helps someone work out a new pattern; crystallized knowledge helps them use facts, vocabulary, and practiced skills acquired over time. Cattell’s paper proposing a theory of fluid and crystallized intelligence is a foundational source for this distinction. Cattell, “Theory of Fluid and Crystallized Intelligence: A Critical Experiment” (1963)

Consider a simple example. If you are given an unfamiliar sequence—2, 5, 8, 11—and asked what comes next, you can infer a repeating addition rule and propose 14. That is an illustration of reasoning from the examples. If you already memorized the sequence or were taught the rule, recalling it draws more directly on learned knowledge. Real tasks can involve both: knowing what “sequence” means and keeping the numbers in mind are learned or supporting resources, while identifying the rule is the reasoning challenge.

These are examples, not pure categories. A new route may be easier for someone who has used maps before; a puzzle can become familiar after practice. The point is to ask what the task requires, rather than to label an entire person as “fluid” or “knowledge-based.”

An unfamiliar shape puzzle calls for reasoning about a pattern alongside attention and learned skills.
Recalling a country's capital calls mainly on learned knowledge.
Following a new map route calls on both learned map conventions and reasoning about unfamiliar paths.
Using a familiar shortcut draws on knowledge of that route.
Section 3

What can a fluid-reasoning task tell you?

A structured assessment can sample how someone handles particular reasoning problems under specified conditions. But performance on any one task reflects more than the ability a test is designed to measure. In a study of children and adults, Cochrane, Simmering, and Green examined fluid intelligence in relation to memory and attention, underscoring that these cognitive processes can be involved in test performance. Cochrane, Simmering, and Green, “Fluid intelligence is related to capacity in memory as well as attention,” PLOS ONE (2019)

Results therefore need context: what the task measured, how it was administered, and how the score should be interpreted. A short online quiz may be designed for entertainment, practice, or a narrow skill; without suitable information about its construction and interpretation, its result is not a comprehensive account of a person’s abilities. Even a formal test provides a sample of performance, not a complete description of someone’s knowledge, creativity, judgment, or potential.

Section 4

How to use the distinction in everyday life

When facing a new task, separate the parts. Ask: What do I already know that helps? What rule or relationship do I need to infer? What information must I hold in mind while I try? For a new route, prior map-reading knowledge may help with symbols, while fluid reasoning may help compare unfamiliar paths and adapt if one is blocked. This breakdown can suggest what kind of practice would help: learn the conventions, try varied examples, or slow down and check the inferred rule.

Keep the conclusion modest. Getting a puzzle right shows that you solved that puzzle under those conditions. Finding it difficult may reflect unfamiliarity, distractions, unclear instructions, or the particular task. Neither outcome, by itself, defines a person or fixes what they can learn.

Section 5

In brief

Fluid intelligence refers to reasoning through unfamiliar problems; crystallized intelligence refers broadly to knowledge and skills built through learning. Most real tasks draw on a mixture of both, along with other cognitive processes. The distinction can help explain what a task asks you to do, but a single test or online quiz cannot sum up who you are.

Section 6

Sources and scope

The listed sources support the factual definitions and form or product details. Scenarios, examples and choice steps are original editorial illustrations.

Duncan et al, PNAS fluid reasoning study: https://pubmed.ncbi.nlm.nih.gov/28461462/
Cattell, original fluid and crystallized distinction: https://psycnet.apa.org/record/1963-07991-001
Cochrane et al, PLOS ONE study: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0221353
Related reading

Keep exploring this topic