Guide
A short history of IQ tests
Intelligence testing is a little over a century old. In that time it went from a practical school test in Paris to a family of standardized clinical scales, picked up the term IQ, swapped one way of scoring for another, and produced the nonverbal puzzle formats that most online tests use today.
Want your own number? The test is 25 visual puzzles, about 12 minutes, no sign-up.
Timeline
| Year | Milestone |
|---|---|
| 1904 | Charles Spearman proposes the general intelligence factor, g. |
| 1905 | Alfred Binet and Theodore Simon publish the first practical intelligence test. |
| 1912 | William Stern proposes the term "intelligence quotient". |
| 1916 | Lewis Terman publishes the Stanford revision of the test. |
| 1938 | John C. Raven first publishes Raven's Progressive Matrices. |
| 1939 | David Wechsler introduces the deviation IQ with the Wechsler-Bellevue scale. |
| 1963 | Raymond Cattell distinguishes fluid from crystallized intelligence. |
| 1984 | James Flynn documents rising IQ scores in the US, about 3 points per decade. |
| 2003 | The Stanford-Binet Fifth Edition (SB5), by Gale Roid, is published. |
| 2008 | WAIS-IV is published. |
| 2024 | WAIS-5 is published. |
1904: a general factor
The story begins slightly before the first test. In 1904 Charles Spearman published “General Intelligence, Objectively Determined and Measured” in the American Journal of Psychology, proposing that a general factor, g, runs through performance on many different mental tasks (Wikipedia: g factor). That idea, that different tests share a common core, later justified summarizing a test in one number. The guide to the g factor explains it in detail.
1905: the first practical test
In 1905 Alfred Binet and Theodore Simon published the first practical intelligence test (Wikipedia: Intelligence quotient). It set the template for what followed: a standard set of tasks, given in a standard way, with results compared against what is typical.
1912 and 1916: a name and an American revision
In 1912 the psychologist William Stern proposed the term “intelligence quotient”, and in 1916 Lewis Terman published the Stanford revision of the test (Wikipedia: Intelligence quotient). The Stanford-Binet line continues to this day; its fifth edition, SB5, by Gale Roid, was published in 2003 (Wikipedia: Stanford-Binet).
The word quotient survives even though modern tests no longer divide anything. Today's scores are standard scores, the change that came next.
How the first IQ was calculated
The intelligence quotient was first suggested by William Stern and adopted by Lewis Terman for the Stanford-Binet (Britannica, intelligence test). It was a ratio: mental age divided by chronological age, times 100, so a 10-year-old performing like an average 12-year-old scored 120 (Britannica, IQ). Few tests compute mental ages today. See the original IQ formula.
Try one · Abstract Reasoning
Which tile completes the grid?
Tap the option you think is right, or show the answer.
Show the answer
The missing tile has three solid stars (option 2).
How to solve it
- Start with the first row and name what changes from tile to tile.
- Every row and every column contains each shape (cross, star and square) exactly once.
- Every row and every column contains each number of shapes (three, two and one) exactly once.
- The fill (solid) stays the same in every tile.
- Check the rules on the second row, then apply them to the third.
- So the missing tile has three solid stars.
That was 1 puzzle. The full test has 25.
1938: matrices without words
John C. Raven developed his Progressive Matrices in the 1930s and first published them in 1938. They are non-verbal matrix puzzles, and they are widely used as a measure of fluid, culture-reduced reasoning (Wikipedia: Raven's Progressive Matrices). The format, a grid of figures with one missing, is now the backbone of nonverbal testing, including the Abstract Reasoning section of IQ Octopus. See matrix reasoning, explained.
1939: the deviation IQ
David Wechsler introduced the deviation IQ, based on standard scores, with the Wechsler-Bellevue scale in 1939 (Wikipedia: WAIS). Instead of a quotient, a person's result is expressed as a distance from the average of their age group, measured in standard deviations, and rescaled so the mean is 100. Modern tests set the standard deviation to 15 (Wikipedia: Intelligence quotient).
This is the most important change in the history of IQ scoring. It is why 100 is always average, why a score keeps its meaning across ages, and why every IQ maps to a percentile. Older Stanford-Binet forms used a standard deviation of 16 instead, which is why old and new scores can differ for the same rank. See deviation IQ.
Wechsler's adult scale has been revised since. WAIS-IV was published in 2008 and WAIS-5 in 2024 (Wikipedia: WAIS).
1963: two kinds of ability
In 1963 Raymond Cattell distinguished fluid intelligence, reasoning with novel material, from crystallized intelligence, acquired knowledge (Wikipedia: Fluid and crystallized intelligence). The split helped explain why vocabulary tests and puzzle tests do not always agree, and why different abilities follow different paths with age. Read fluid vs crystallized intelligence.
1984 onward: scores on the move
In 1984 James Flynn documented that average measured IQ in the US had risen by about 3 points per decade. A 2015 meta-analysis of 271 samples, about 4 million people across 31 countries from 1909 to 2013, estimated about 2.8 points per decade and found the gains slowing (Pietschnig & Voracek, 2015). A reversal has since been reported in Norwegian conscript data for cohorts born after about 1975 (Bratsberg & Rogeberg, 2018).
These shifts are a big reason tests are renormed. If the average changes, an old norm table no longer says what 100 means today, so new editions are standardized against new samples. See the Flynn effect.
What the history means for a score today
- An IQ is a deviation score: a rank among same-age peers, with mean 100 and standard deviation 15 on most tests.
- Scale matters. A number from an SD 16 or SD 24 test is not directly comparable with one from an SD 15 test.
- Norms age. A score is a comparison with a particular reference group at a particular time.
- Nonverbal matrix-style puzzles, descended from Raven's work, are a standard way to test reasoning.
Where IQ Octopus fits
IQ Octopus borrows the modern conventions: an estimated IQ on the mean 100, standard deviation 15 scale, shown with a confidence range. Its puzzles follow the nonverbal tradition, with matrices, sequences, rotation, odd one out, number series and figure analogies, generated fresh for every attempt. It is not a clinical test and does not replace one. For how it compares with the tests in this history, read WAIS, Stanford-Binet, Raven's and Mensa tests compared, or the short answer to who invented the IQ test?
Sources
- Intelligence quotient. Wikipedia. en.wikipedia.org/wiki/Intelligence_quotient
- g factor (psychometrics). Wikipedia. en.wikipedia.org/wiki/G_factor_(psychometrics)
- Wechsler Adult Intelligence Scale. Wikipedia. en.wikipedia.org/wiki/Wechsler_Adult_Intelligence_Scale
- Stanford-Binet Intelligence Scales. Wikipedia. en.wikipedia.org/wiki/Stanford%E2%80%93Binet_Intelligence_Scales
- Raven's Progressive Matrices. Wikipedia. en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
- Fluid and crystallized intelligence. Wikipedia. en.wikipedia.org/wiki/Fluid_and_crystallized_intelligence
- Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains. Perspectives on Psychological Science, 10(3), 282-306. doi.org/10.1177/1745691615577701
- Bratsberg, B., & Rogeberg, O. (2018). Proceedings of the National Academy of Sciences, 115(26), 6674-6678. doi.org/10.1073/pnas.1718793115
- Encyclopaedia Britannica. intelligence test (read via Wayback Machine, January 2025) www.britannica.com/science/intelligence-test
- Encyclopaedia Britannica. IQ, revised by Jeannette L. Nolen (read via Wayback Machine, December 2024) www.britannica.com/science/IQ