The g factor, short for general intelligence, is the name psychologists give to a broad mental ability that seems to show up in almost every kind of thinking test. Someone who scores well on vocabulary tends to score well on puzzles and arithmetic too, and g is the label for whatever they share. What g actually is remains an open question: some researchers see one underlying ability, others see many separate mental processes that simply overlap.1Psychological Inquiry. Process Overlap Theory: A Unified Account of the General Factor of Intelligence The idea matters outside the lab, because tests built around it help decide who gets hired, trained, or placed in military jobs.2Intelligence. Reducing adverse impact in high-stakes testing
The Pattern That Gave Rise to G
One of the best-established findings in intelligence research is that scores on very different intelligence tests go up and down together.3PubMed Central. Do Executive Attentional Processes Uniquely or Commonly Explain Psychometric g and Correlations in the Positive Manifold? A Structural Equation Modeling and Network-Analysis Approach to Investigate the Process Overlap Theory4Intelligence. Attention Control and Process Overlap Theory: Searching for Cognitive Processes Underpinning the Positive Manifold This overlap has long been taken as evidence of a single general ability.5PubMed. A single g factor is not necessary to simulate positive correlations between cognitive tests The reasoning is simple: if one shared ability feeds into every test, people strong in it would do better across the board.
A person’s ranking on one task usually predicts roughly where they rank on the others. The match is far from perfect, though, so people still have relative strengths and weaknesses. g captures the shared part, and the rest reflects specific skills.
g is estimated by a statistical method that looks for what many test scores have in common. It is not measured directly, which means an IQ score is a practical stand-in for g, not a direct reading of it.
IQ is a score from one test on one day, while g is the shared factor across many tests. A single IQ score mixes g with specific skills and testing conditions, so two people with the same IQ can differ in their specific abilities.
g is also narrower than everyday “smart.” It is about reasoning and learning on tests, not wisdom, creativity or social skill, which people often include when they call someone smart.
How G Fits With Other Kinds of Intelligence
An early split divided general intelligence into fluid intelligence, the ability to solve new reasoning problems, and crystallized intelligence, the ability to use what you’ve already learned.6Wikipedia. Fluid and crystallized intelligence Fluid intelligence is linked to comprehension, problem-solving and learning.6Wikipedia. Fluid and crystallized intelligence Solving an unfamiliar pattern puzzle draws mainly on it. Defining words or recalling history draws on crystallized intelligence, which builds up through schooling and reading. Both still overlap through g.
Today the dominant framework combines these ideas into a layered map of abilities.7Cogn-IQ.org. Fluid vs Crystallized Intelligence: Gf, Gc, and CHC Theory It has three tiers: g at the top, around a dozen or so broad abilities in the middle, and many narrow, specific skills at the bottom, such as spelling, mental rotation of shapes, or recalling a list of digits. The map is a model built from test data. It is a useful way of organizing abilities, not a diagram of separate organs in the brain.
In one common way of modelling the map, each test draws on g directly, while broad abilities like verbal or memory skill add their own separate contribution.8PubMed Central. The Bifactor Model Fits Better Than the Higher-Order Model in More Than 90% of Comparisons for Mental Abilities Test Batteries This kind of model usually fits test data better than one where g works only through the broad abilities.9PubMed Central. Why Do Bi-Factor Models Outperform Higher-Order g Factor Models? A Network Perspective
This helps when you read a test report. School psychologists’ reports often list broad abilities like verbal comprehension or processing speed. An overall score summarises g, while index scores show the broad abilities. Uneven index scores can matter more than the overall number for spotting a learning difficulty.
Is G One Ability or Many Working Together
Working memory, the mental scratchpad that holds information while you use it, is strongly linked to reasoning ability.10PubMed. Is working memory capacity a causal factor in fluid intelligence? Keeping a phone number in mind while you dial it, or following multi-step directions you were just given, relies on it. Reasoning problems often need several pieces in mind at once, so working memory and reasoning scores go together.
In some analyses g and working memory are very closely linked, more closely than g and any other ability.11Intelligence. Working memory is (almost) perfectly predicted by g What seems to matter is how many items you can hold at once, not how precisely you hold each one.12PubMed Central. Quantity not quality: The relationship between fluid intelligence and working memory capacity
Process overlap theory offers a different reading. It says each test draws on a mix of general control processes, such as focusing attention, and skills specific to that kind of test.13PubMed Central. In Search of the Executive Cognitive Processes Proposed by Process-Overlap Theory Tests overlap because they share some of those control processes, not because one ability drives them all.13PubMed Central. In Search of the Executive Cognitive Processes Proposed by Process-Overlap Theory Computer simulations show test scores can produce a convincing g pattern even when no single general ability exists.14Personality and Individual Differences. Simulating the process overlap theory of intelligence: A unified framework bridging psychometric and cognitive perspectives Finding g in test scores therefore does not by itself prove there’s one thing in the head called g.
What G Looks Like in the Brain
People with larger brains tend to score somewhat higher on intelligence tests, a moderate but consistent link.15PubMed Central. Genes, Cells and Brain Areas of Intelligence Still, brain volume accounts for only about 12% of differences in intelligence, and all brain scan measures together about 18 to 21%.16PubMed Central. Beyond a bigger brain: Multivariable structural brain imaging and intelligence So many people with average-sized brains score high, and the reverse. The health of the brain’s white matter, the long connecting fibers, also matters, because damaged fibers slow information processing.15PubMed Central. Genes, Cells and Brain Areas of Intelligence
Speed shows up too. People who can correctly judge very briefly flashed images tend to score higher on general intelligence.17Intelligence. Placing inspection time, reaction time, and perceptual speed in the broader context of cognitive ability: The VPR model in the Lothian Birth Cohort 1936 Faster basic processing is linked to better performance on harder thinking tasks.18PubMed. A genetic investigation of the covariation among inspection time, choice reaction time, and IQ subtest scores
Brain imaging points to a network linking frontal and parietal areas, rather than a single “intelligence spot.”19Behavioral and Brain Sciences. The Parieto-Frontal Integration Theory (P-FIT) of intelligence: Converging neuroimaging evidence Frontal areas help plan and hold goals in mind. Parietal areas help with attention and handling spatial information. Good reasoning seems to need these regions to work together quickly.20Wikipedia. Parieto-frontal integration theory
Is G Inherited
Heritability is the share of differences between people in a group that traces to genetic differences. Twin and adoption studies show IQ heritability rises with age, reaching about 80% by around 18 to 20 and staying there through adulthood.21PubMed. The Wilson Effect: the increase in heritability of IQ with age DNA studies estimate that common genetic variants explain about 40 to 50% of differences in intelligence.22PubMed Central. Genome-wide association studies establish that human intelligence is highly heritable and polygenic
The two figures come from different methods. Twin and adoption studies compare relatives, while DNA studies look at measured genetic variants in unrelated people. The DNA method captures only common variants, so it is expected to give a lower figure.
A study of nearly 270,000 people found 205 regions of DNA and over 1,000 genes linked to intelligence.23PubMed. Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence Each gene has a tiny effect, and there is no single “intelligence gene.” These genes are active in the brain and involved in its development and connections.23PubMed. Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence
A common misunderstanding is that 80% heritable means 80% of your own intelligence comes from genes. It describes differences across a group in a given setting. High heritability doesn’t mean environment can’t change scores, and a nutritious, stimulating upbringing still matters.
Does G Change Over Your Life
Ability measured at age 1 to 2 only loosely predicts adult ability, but stability grows strongly through childhood.24PubMed Central. Stability of general cognitive ability from infancy to adulthood: A combined twin and genomic investigation Infants are tested on things like attention and responses to objects, which differ from the reasoning and language tasks used later.
By school age, scores at 7, 16 and 29 track each other closely.24PubMed Central. Stability of general cognitive ability from infancy to adulthood: A combined twin and genomic investigation In one group tested at 11 and again at 90, childhood scores still predicted old-age scores fairly well.25PubMed. The stability of intelligence from age 11 to age 90 years: the Lothian birth cohort of 1921 Stability here means your position relative to peers tends to hold, even though everyone’s raw abilities grow in childhood and shift in old age.
In later life, different abilities decline together, so g plays a larger role in test scores after 75 than in young adults.26CCAB Research. Factor Analysis of Comprehensive Cognitive Testing Reveals Age-related Cognitive Dedifferentiation Vocabulary holds up well into later adulthood, though in very late life it tends to decline alongside memory. Shared brain changes like blood vessel disease affect many abilities at once.
Across generations, IQ scores rose over the last half century.27Intelligence. Is the Flynn Effect on g?: A meta-analysis These gains were largest on the skills least tied to g, suggesting environment raised specific skills more than g itself.27Intelligence. Is the Flynn Effect on g?: A meta-analysis
Can You Increase Your G With Brain Training
People who train on working memory tasks get much better at those tasks, with gains lasting at least 6 months.28PubMed Central. Failure of Working Memory Training to Enhance Cognition or Intelligence But the improvement doesn’t spread to untrained tasks or to intelligence tests.28PubMed Central. Failure of Working Memory Training to Enhance Cognition or Intelligence Compared with people doing a placebo activity, effects on reasoning, verbal skill, reading and arithmetic are close to zero.29PubMed Central. Working Memory Training Does Not Improve Performance on Measures of Intelligence or Other Measures of “Far Transfer”
What helps thinking in practice is more ordinary. Learning real skills and knowledge builds crystallized ability. Sleep, physical activity and managing health problems support everyday mental performance. Practice on a specific test type can raise that test score without changing g.
What G Predicts in Real Life
Across many studies, general mental ability is one of the better predictors of job performance, stronger in complex jobs than simple ones.30OpenPsych. Intelligence Really Does Predict Job Performance: A Long-Needed Reply to Richardson and Norgate It also predicts success in job training well, as large military training records show.30OpenPsych. Intelligence Really Does Predict Job Performance: A Long-Needed Reply to Richardson and Norgate In some jobs, such as aviation roles, specific knowledge predicted performance better than g.31PubMed Central. The Role of General and Specific Cognitive Abilities in Predicting Performance of Three Occupations: Evidence from Bifactor Models Similar results appear outside the US, such as in Germany.32International Journal of Selection and Assessment. Validity of General Mental Ability for the Prediction of Job Performance and Training Success in Germany: A meta‐analysis
Tests like the Wonderlic for employers and the ASVAB for the military rely heavily on general ability.2Intelligence. Reducing adverse impact in high-stakes testing Scores can open doors to prestigious careers or limit options.2Intelligence. Reducing adverse impact in high-stakes testing Critics argue such tests can deepen social inequality and that their validity has been overstated.33Journal of Occupational and Organizational Psychology. A critical review of the use of cognitive ability testing for selection into graduate and higher professional occupations Meanwhile, the structure of abilities these tests measure holds up well across many countries and languages.34Intelligence. The cross-cultural generalizability of cognitive ability measures: A systematic literature review
If you’re facing an aptitude test, practice with sample questions to get used to the format and timing. Familiarity can raise your score without changing your ability. Rest and arrive calm, since fatigue and anxiety lower performance.

