The g factor, short for “general factor,” is a single broad mental ability that influences performance across all cognitive tasks. First proposed by Charles Spearman in 1904, it remains one of the most studied and debated concepts in psychology. Spearman noticed that people who scored well on one type of mental test tended to score well on others, and he proposed that a common underlying source of mental ability explained this pattern.
How Spearman Discovered It
Spearman’s key observation was something called the “positive manifold”: scores on different cognitive tests are always positively correlated with each other. Someone who excels at vocabulary tasks tends to do better than average on spatial reasoning, memory, and processing speed tests too. Rather than treating each ability as completely separate, Spearman argued that a single general factor, g, acts as a common source powering all of them. He described it loosely as a kind of “mental energy.” In his two-factor model, every cognitive test result reflects two things: this shared general factor and a specific ability unique to that particular task.
Where g Sits in Modern Intelligence Models
Today’s most widely used framework for understanding intelligence is the Cattell-Horn-Carroll (CHC) model, which arranges cognitive abilities in a three-level hierarchy. At the bottom are dozens of narrow skills like spelling ability or spatial scanning. These cluster into broad abilities at the middle level, including fluid reasoning (solving novel problems), crystallized knowledge (accumulated knowledge and vocabulary), visual processing, short-term memory, and processing speed. At the top of the hierarchy sits g, the single general factor that emerges because all of those broad abilities are positively correlated with each other.
In intelligence testing, different types of tasks “load” onto g to different degrees. Fluid reasoning and visual processing tasks tend to have the strongest connections to g, with loading values reaching 0.86 and 0.82 in school-age children. Processing speed, by contrast, has a weaker link to g, with loadings around 0.44 in teenagers. The more varied the mix of subtests included in an IQ composite score, the more that score reflects g rather than any single narrow ability.
How Stable Is g Over a Lifetime?
Remarkably stable. The Lothian Birth Cohort study in Scotland tested the same people at age 11 and again at age 90, finding a correlation of 0.54 between their scores, nearly eight decades apart. After adjusting for statistical limitations, that correlation rose to 0.67. This means that a substantial portion of the differences between people in cognitive ability in very old age were already visible in childhood. That doesn’t mean intelligence is fixed or that learning doesn’t matter, but the rank order among individuals holds up to a surprising degree across the lifespan.
Genetics and Environment
Twin studies consistently show that g is highly heritable in adulthood, with estimates ranging from 75 to 85% of the variation between people attributable to genetic influences. A more detailed breakdown from one large analysis found that 44% of the variance came from additive genetic effects (the straightforward inheritance of many small genetic contributions), 27% from non-additive genetic effects (interactions between genes), 11% from the tendency of similar people to partner together, and 18% from individual environmental experiences not shared with siblings.
Heritability increases with age. In childhood, shared family environment plays a larger role, but by adulthood, genetic factors dominate. This doesn’t mean environment is irrelevant. Nutrition, education, and childhood conditions all shape cognitive development. But it does mean that within a population where most people have access to similar environments, genetic differences explain most of the remaining variation.
What g Predicts in Real Life
The practical significance of g extends well beyond test scores. Higher general cognitive ability in childhood is associated with a greater likelihood of exercising, eating well, being health literate, avoiding smoking, and living longer. It is also linked to lower rates of hypertension, obesity, and coronary heart disease.
In the workplace, the relationship between g and job performance has been extensively studied. Hundreds of early studies found raw correlations between IQ scores and job performance in the range of 0.2 to 0.3. After correcting for statistical artifacts like restricted ranges of ability within any single workplace, those correlations roughly doubled to around 0.5, a moderate-to-strong relationship. The strength of the link varies: in one study of salespeople, cognitive ability correlated 0.40 with supervisor ratings but only 0.04 with actual sales numbers, highlighting that what “performance” means matters a great deal. For training success, the correlation with general intelligence is around 0.54.
Social mobility is another area where g shows up. In a longitudinal study tracking people from childhood to their early fifties, higher childhood IQ predicted upward occupational mobility even after accounting for education level. Education was the stronger predictor of career advancement in early adulthood (up to age 25), but from the mid-twenties onward, IQ became the dominant factor. People with high childhood IQ scores tended to climb the occupational ladder regardless of their formal qualifications. Higher childhood IQ was also protective against downward mobility, while factors like adult height, childhood housing conditions, and adverse childhood events were not significant predictors.
The Brain Basis of g
Researchers have looked for a clear neural signature of g, but the picture is complicated. One study of older adults examined whether the structural quality of the brain’s white matter (the wiring that connects different regions) could explain g. A general factor of white matter integrity did emerge, meaning that if one brain connection is in good shape, others tend to be too. However, this wiring quality predicted processing speed, not general intelligence or memory. The correlation between white matter integrity and the general intelligence factor was only about 0.08, essentially zero. This suggests that g is not simply a matter of how well-connected the brain’s wiring is, and that its biological basis likely involves more complex mechanisms spread across multiple brain systems.
Criticisms and Rival Theories
Not everyone accepts that g is a meaningful real-world entity. The most prominent alternative comes from Howard Gardner, who proposed that intelligence is better understood as a set of independent abilities: linguistic, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal, intrapersonal, and naturalistic. Gardner argued that evidence for each intelligence comes from people who excel in one area while being average or impaired in others, and from distinct brain regions that support each ability.
The empirical evidence, however, has not been kind to the independence claim. When researchers actually measured multiple Gardner-style intelligences in the same individuals, they found that scores on most of the intelligences were intercorrelated, which is exactly the pattern that gives rise to g. This is a fundamental problem for the theory, because if the intelligences are not independent, they do not exist as Gardner described them. There are no standardized tests for the multiple intelligences, which means different researchers create their own measures and findings cannot be compared or accumulated. A recent review labeled the theory a “neuromyth,” though one factor analysis did find some evidence for independent linguistic, logical-mathematical, and spatial subscales.
Other critics take a more nuanced position. They accept that g is a robust statistical finding but question whether it represents a single cognitive mechanism or is instead an emergent property of many interacting abilities. The debate is not about whether g exists in the data. It clearly does. The question is what it means: whether it points to a single engine of cognition or is simply a useful summary of the fact that cognitive abilities tend to travel together.