Psychology › Personality & intelligence
Intelligence and the g factor
1904 (Spearman)· Concept
Well replicatedStrong, repeated evidence.
The finding that people who do well on one kind of mental test tend to do well on others, summarized as a general factor, g.
See it on the Psychology map →What it says
In 1904 British psychologist Charles Spearman noticed that schoolchildren's scores in unrelated areas, from Classics to telling musical pitches apart, correlated. He proposed a general intelligence, g, plus specific abilities. A century of testing has confirmed this 'positive manifold': g typically accounts for around 40 to 50 percent of the differences on a test battery. Scores predict school results, job performance, income and even lifespan, on average, though far from perfectly.
Why it matters
g is among the most predictive and best replicated measures in psychology. Its history is also dark: it was misused to justify forced sterilization and immigration limits.
Think of it like
Like general athleticism: people fast at sprinting tend to jump and throw well too, though each has strengths and weak spots.
The evidence
Factor analysis of thousands of test datasets, including John Carroll's 1993 reanalysis of more than 460 of them, and long-term studies linking childhood scores to adult outcomes.
Open questions
What g is in the brain remains unknown: candidates include processing speed, working memory and how efficiently brain networks communicate. Some argue it is a statistical summary, not one thing.
Inside this (5)
Connections (4)
▸Working memory correlates strongly with Intelligence and the g factorWell replicated
The connection
Working memory is the mental workspace that holds a few items while you use them. People with larger capacity tend to score higher on reasoning tests; some analyses put the correlation with fluid intelligence around 0.7, suggesting both draw on the ability to control attention.
Who showed it and when
Key studies include Patrick Kyllonen and Raymond Christal's 1990 US Air Force research and a 2005 analysis by Michael Kane and colleagues.
The evidence
Large test batteries consistently show the correlation. But training that boosted working memory did not raise intelligence, according to 2013 and 2016 meta-analyses of brain-training studies.
Open questions
Is working memory a cause of intelligence, or simply another measure of the same underlying ability? Researchers still disagree.
▸IQ tests measure Intelligence and the g factorWell replicated
The connection
IQ tests are the main tool for estimating g. A battery like the Wechsler Adult Intelligence Scale combines vocabulary, matrix puzzles, memory span and speed tasks; because these scores correlate, their total gives a good estimate of general ability plus a profile of strengths and weaknesses.
Who showed it and when
From Binet's 1905 Paris test through David Wechsler's adult scales, first published in 1939 and revised regularly, most recently in 2024.
The evidence
Different IQ tests agree closely, and a 2004 study by Wendy Johnson and colleagues found g from three separate batteries correlated almost perfectly.
Open questions
Psychologist Edwin Boring quipped in 1923 that intelligence is 'what the tests test.' Whether tests capture creativity, wisdom or practical skill is still argued.
▸Gardner's multiple intelligences challenged the single-g view of Intelligence and the g factorWeak or failed to replicate
The connection
Gardner rejected a single general factor, arguing for separate intelligences that vary independently. Psychometricians responded that when his intelligences are actually tested, most correlate positively and overlap with g, though musical and bodily skills are less tied to it. The challenge popularized broader views of talent without overturning g.
Who showed it and when
Gardner's Frames of Mind, Harvard, 1983; responses included a 2006 Canadian study by Beth Visser, Michael Ashton and Philip Vernon.
How solid is it?
Visser's team tested about 200 adults and found the more academic intelligences overlapped strongly with g. Gardner offered little direct evidence for independence.
Open questions
Whether intelligence should include music, movement and social skill is partly a matter of definition, which data alone cannot settle.
▸Fluid vs crystallized intelligence splits Intelligence and the g factorWell replicated
The connection
Cattell argued g actually blends two abilities: fluid reasoning on new problems and crystallized knowledge from learning. Later models placed both beneath g alongside other broad abilities such as memory and processing speed. Most modern tests, built on the Cattell–Horn–Carroll model, report fluid and crystallized scores separately.
Who showed it and when
Cattell, 1940s to 1963; John Horn added more broad abilities; John Carroll's 1993 reanalysis of more than 460 datasets completed the hierarchy.
The evidence
Factor analyses consistently separate the two, and they age differently, which a single undivided ability would not predict.
Open questions
Some analyses find fluid intelligence nearly identical to g, and whether crystallized knowledge counts as intelligence or as its product is debated.