Binet and the first practical test
In 1905, French psychologist Alfred Binet and his colleague Théodore Simon designed the first practical intelligence test, on commission from the French Ministry of Education. Their goal was modest: identify children who needed extra educational support. They built a series of age-graded tasks — naming objects, repeating sentences, comparing weights — and ranked children by the most advanced tasks they could complete.
The 'mental age' produced by this test was later divided by chronological age and multiplied by 100, giving the original Intelligence Quotient.
Wechsler and the modern battery
In the 1930s, David Wechsler developed a new family of tests for adults and children: the Wechsler Adult Intelligence Scale (WAIS) and the Wechsler Intelligence Scale for Children (WISC). These tests broke intelligence into multiple subscales — verbal comprehension, perceptual reasoning, working memory, and processing speed — and combined them into a Full-Scale IQ.
Wechsler also abandoned the original ratio formula. Modern scores use a normal distribution with a mean of 100 and a standard deviation of 15, so an IQ is now defined by how far you fall from average, in standard deviations.
Raven's Progressive Matrices
In 1936, John C. Raven introduced the Progressive Matrices, a set of visual analogy puzzles arranged in increasing difficulty. Each item shows a grid of geometric shapes with one cell missing; the test-taker must pick the answer that completes the pattern.
Raven's test was revolutionary because it required no language and minimal cultural knowledge. It became the standard tool for measuring fluid intelligence — the part of cognition most associated with novel problem solving.
Online assessments today
Modern online tests like the one on this site typically combine matrix items with shorter tasks on number series, odd-one-out reasoning, and analogies. The best online tests calibrate item difficulty against a reference sample, control for guessing, and avoid items that depend on culture-specific knowledge.
They cannot replicate the supervised, hour-long batteries used by clinical psychologists. But for self-exploration, they are a reasonable estimate — usually accurate to within ten points of a full clinical assessment.