Guide
What is the g factor?
People who do well on one kind of mental test tend to do well on others. In 1904 Charles Spearman gave that pattern a name: the general intelligence factor, or g. It is the reason a single IQ number means anything at all, and also the reason to be careful about what that number claims.
Want your own number? The test is 25 visual puzzles, about 12 minutes, no sign-up.
The observation behind g
Give a large group of people several quite different mental tests: vocabulary, arithmetic, matrix puzzles, memory for digits. You might expect each test to measure its own separate skill, so that being good at one says nothing about the others. That is not what happens. Across people, scores on different mental tests tend to go together. Those who do well on one tend, on average, to do well on the rest.
Charles Spearman noticed this pattern and, in 1904, proposed that a general factor runs through all such tests. He published the idea in the American Journal of Psychology under the title “General Intelligence, Objectively Determined and Measured” (Wikipedia: g factor). He called the general factor g.
What g is, in plain terms
g is a statistical summary. When many test scores all correlate with each other, a technique called factor analysis can extract the part they share. That shared part is g. Each test also has its own specific part, which is why someone can be stronger at matrices than at vocabulary while still scoring high on both.
A useful picture: think of each test score as made of two ingredients, a general ingredient that every test shares and a specific ingredient that belongs to that test alone. g is the first ingredient. A test that leans heavily on the general ingredient is called highly g-loaded.
Why g matters for IQ scores
g is the reason a single summary number is useful. If mental abilities were completely separate, adding them into one score would be like adding your height to your shoe size. Because they overlap, one number can say something real about performance across many kinds of task.
That is also why a test made of several item types tends to give a more stable estimate than one made of a single type. Each item type brings its own specific noise; combining them lets the shared part stand out. IQ Octopus uses five puzzle types for this reason, though all five are visual or numerical.
Try one · Pattern Recognition
Which tile comes next?
Tap the option you think is right, or show the answer.
Show the answer
The missing tile shows a tiny outlined square (option 4).
How to solve it
- Compare neighboring tiles and name what changes each time.
- From tile to tile, the square gets smaller.
- The fill cycles outline, hatched and solid, then starts again.
- Everything else stays the same from tile to tile.
- The last tile shows a small solid square. Apply each change once more.
- So the missing tile shows a tiny outlined square.
That was 1 puzzle. The full test has 25.
g and the outcomes research
Much of the research on what intelligence predicts uses measures of general ability. In about 70,000 English children, general intelligence at age 11 correlated 0.81 with total GCSE exam points at age 16 (Deary, Strand, Smith & Fernandes, 2007). A review of 85 years of personnel research estimated the validity of general mental ability for predicting job performance at 0.51 (Schmidt & Hunter, 1998). These are group-level findings. They describe tendencies across many people, not what will happen to any one person. Our guide to what IQ predicts, and what it does not explains how to read them.
g and its parts
g is not the end of the story. In 1963 Raymond Cattell distinguished two broad parts of general ability: fluid intelligence, reasoning with novel material, and crystallized intelligence, acquired knowledge (Wikipedia: Fluid and crystallized intelligence). You can think of g as the top of a hierarchy, with broader abilities such as these beneath it and the specific skills each test draws on beneath those. The guide to fluid vs crystallized intelligence covers the middle layer.
The layers under g
John Carroll's 1993 reanalysis of more than 460 data sets organized abilities into three strata with g at the top (Carroll, 1993). The Cattell-Horn-Carroll (CHC) family of theories that grew from it lists broad abilities such as fluid reasoning, comprehension-knowledge, visual processing, short-term working memory and processing speed beneath g (McGrew et al., 2023). Modern batteries like the WAIS-5 report index scores that map onto several of these. See CHC theory.
What g is not
- Not the same as an IQ score. An IQ is a result from one test, with measurement error. g is the shared factor behind many tests. A broad, well-built test is designed to be a good estimate of g.
- Not a location in the brain. g is found in test data. Whether it reflects one underlying process or many overlapping ones is a research question, not something the statistics alone can answer.
- Not the whole of intelligence. Creativity, practical judgment, specific talents, knowledge, character and motivation all matter in life, and a test samples only some of what people mean by intelligence.
- Not fixed or destiny. A person's measured score is a snapshot with error, compared with a particular group at a particular time.
How this connects to the IQ scale
Modern tests turn general ability into a deviation IQ: scores standardized to a mean of 100 and a standard deviation of 15 within each age group (Wikipedia: Intelligence quotient). That scale is just a way of expressing rank. An IQ of 115 sits one standard deviation above the mean of your age group, whatever mix of test items produced it. See deviation IQ and the IQ bell curve, explained.
Where IQ Octopus fits
IQ Octopus estimates general reasoning from 25 puzzles across five types: pattern sequences, matrices, spatial items, number series and figure analogies. Because the estimate draws on several formats, it leans toward what they share. Because it has no verbal or knowledge items, it does not capture the crystallized side of general ability. The result is an estimated IQ with a confidence range, not a measurement of g itself. Read the short definition of general intelligence, see what an IQ test measures, or try a few practice puzzles to see the formats.
Sources
- g factor (psychometrics). Wikipedia. en.wikipedia.org/wiki/G_factor_(psychometrics)
- Fluid and crystallized intelligence. Wikipedia. en.wikipedia.org/wiki/Fluid_and_crystallized_intelligence
- Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence, 35(1), 13-21. doi.org/10.1016/j.intell.2006.02.001
- Schmidt, F. L., & Hunter, J. E. (1998). Psychological Bulletin, 124(2), 262-274. doi.org/10.1037/0033-2909.124.2.262
- Intelligence quotient. Wikipedia. en.wikipedia.org/wiki/Intelligence_quotient
- Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge University Press. doi.org/10.1017/CBO9780511571312
- McGrew, K. S., Schneider, W. J., Decker, S. L., & Bulut, O. (2023). A psychometric network analysis of CHC intelligence measures. Journal of Intelligence, 11(1), 19. doi.org/10.3390/jintelligence11010019