What Is the G Factor? General Intelligence Explained

Written and Reviewed by the Cognitive Train Team
← Intelligence & IQ

Vocabulary, matrix puzzles, memory spans, and symbol searches look like different mental jobs. Yet the scores tend to move together. The g factor is the statistical pattern hiding inside that overlap.

Give a large group of people a broad battery of cognitive tests and something stubborn keeps happening. The person who is strong on one demanding task is somewhat more likely to be strong on others too. The relationship is far from perfect — a great verbal reasoner can still have ordinary processing speed — but the correlations keep leaning positive.

Psychologists call that pattern the positive manifold. The broad factor extracted from the shared part of those performances is general intelligence, usually shortened to g. In plain English, g represents what many different cognitive tasks appear to have in common. Our article on what IQ tests actually measure explains the battery itself; here the interesting question is why such different tasks produce a common signal at all.

The Pattern Came Before the Letter

Charles Spearman's famous move in 1904 was to look across mental performances instead of treating every ability as an isolated island. In his original paper, “General Intelligence,” Objectively Determined and Measured, he argued that diverse intellectual activities share a common element alongside task-specific elements.

The modern terminology came later, but the puzzle is still the same. Why should vocabulary have anything to do with spatial reasoning? Why should memory span relate to pattern solving? If the mind were a collection of completely independent abilities, a person's rank on one test would tell you virtually nothing about another. Real test batteries do not behave that way.

That recurring web of positive relationships is the raw phenomenon. g is the compact statistical description researchers build from it.

How Researchers Find g Without a “g Test”

There is no special question where the correct answer reveals your general intelligence. Researchers start with many ordinary test scores and examine how they correlate.

Imagine a table containing verbal reasoning, matrix reasoning, working memory, spatial tasks, and processing speed. Each pair gets a correlation. Factor analysis then asks whether a smaller number of underlying dimensions can account for much of that shared movement. In intelligence batteries, a broad first factor usually appears. That is the factor labeled g.

How different cognitive tasks contribute to the g factor Five different cognitive task areas surround a central circle labeled g, representing the shared factor found in their positive correlations. A note below explains that each task also retains ability-specific variance. Different tasks, shared movement Positive correlations reveal a common factor Verbal words · concepts Reasoning rules · patterns Memory hold · update Spatial shapes · space Speed scan · respond g shared factor Each ability still keeps its own specific variance

A task with a stronger g loading is simply a task whose scores share more variance with that general factor. A Matrix Reasoning Test may load strongly on g, for example, while still measuring something more specific about abstract visual reasoning. The common factor never makes the individual skill disappear.

Why Researchers Take g Seriously

A good way to challenge the idea is to swap out the tests. If g were mostly a quirk of one particular battery, changing the questions should change the general factor substantially.

Wendy Johnson and colleagues ran that test in a 2004 study of 436 adults. The participants completed three different mental-ability batteries built from different tasks and different ideas about intellectual ability. The general factors extracted from the three batteries correlated at .99, .99, and 1.00. The furniture changed; the room stayed remarkably recognizable.

A broader review by Ian Deary, Lars Penke, and Wendy Johnson describes general intelligence as one of the most robust findings in the psychology of individual differences. That persistence is why researchers can use different batteries and still talk meaningfully about a common general ability.

g Is Closely Related to IQ, Yet They Are Different Things

g is a latent statistical factor. IQ is a standardized score produced by a particular test. A well-designed IQ battery is usually built so that its overall score reflects general cognitive ability strongly, which makes the two highly related in practice.

The distinction becomes useful when the words get slippery. You can have one g factor estimated from several batteries, while each battery produces its own IQ scores and index scores. Our IQ vs intelligence guide separates the score from the broader construct, and What Is IQ? explains how the standardized number is interpreted.

g Does Not Flatten the Mind Into One Ability

A strong general factor can coexist with a lopsided cognitive profile. One person may be exceptional at novel reasoning and merely average at speed. Another may have unusually strong verbal knowledge alongside weaker working memory. Both the shared factor and those differences can be real at the same time.

That is why modern intelligence models are usually hierarchical. General intelligence sits at a broad level, with more specific abilities beneath it. The familiar distinction between fluid and crystallized intelligence is one example of structure below the general factor.

A 2025 meta-analysis of Wechsler standardization samples reinforces that point. It found a general intelligence factor across cognitive performances while also showing that the strength of g's contribution varies with ability level and across the lifespan. General does not mean uniform.

Does g Cause Intelligence?

This is where an important line appears. The positive manifold is an empirical pattern. Factor analysis summarizes that pattern extremely well. Neither fact, by itself, tells us that one hidden mental engine literally causes every cognitive ability.

Researchers have proposed several explanations for why the positive manifold exists. One influential alternative is the mutualism model developed by Han van der Maas and colleagues. In that model, abilities such as reasoning, memory, and knowledge strengthen one another during development. Their positive interactions can eventually produce the same kind of correlated test pattern that gives rise to g, without requiring a single underlying causal capacity.

That distinction is easy to lose in popular discussions. The existence of a strong general factor is well supported. The deeper question — why so many abilities become positively related — is a theory problem, and researchers can disagree about the mechanism while agreeing on the statistical pattern.

What the G Factor Tells You — and What It Leaves Behind

g is useful because it captures something genuinely broad. It explains why a person can carry a general advantage across tasks that look quite different on the surface, and why an overall score such as Full Scale IQ can contain meaningful information rather than being a random average of unrelated skills.

It also leaves information behind. A single factor cannot tell you whether someone's strength lies in verbal knowledge, spatial reasoning, memory, or speed. It does not measure motivation, creativity, social judgment, expertise, or persistence. General intelligence is broad within cognition; a whole person is broader still.

See the Shared-Skill Idea in Practice

Cognitive Train’s free IQ-Style Test mixes verbal, numerical, logical, and visual-spatial problems, then shows both an overall IQ-style result and a six-skill breakdown. That combination makes the central idea of g easier to feel: several distinct tasks can contribute to an overall pattern while still telling you different things about your performance.

You can also explore individual abilities through our brain tests and cognitive assessments instead of looking only at a composite.

The Pattern Is the Starting Point

The best way to remember g is to begin where Spearman did: with the correlations. People do not carry exactly the same strength into every mental task, yet cognitive performances refuse to behave like strangers. They overlap.

g is the name given to the broad common factor inside that overlap. It is powerful because the pattern survives across very different tests. It is limited because the shared part is only part of the story. The most interesting picture of intelligence includes both: the current running through many abilities, and the particular abilities that give the current somewhere to go.

Continue through the Brain Articles collection, or explore the skills directly with Cognitive Train’s free cognitive training and brain training tools.