There is no single agreed definition
Ask a group of psychologists to define intelligence and the answers overlap without matching. Common themes include the ability to learn from experience, reason, solve problems, understand complex ideas and adapt to surroundings. Some definitions are narrow, focusing on abstract reasoning. Others are broad, including creativity, practical skills or social understanding. Because the concept is broad, how it is measured depends heavily on which definition one adopts.
A short history of testing
In the early 1900s, the French psychologist Alfred Binet and his colleague Théodore Simon developed a test to identify schoolchildren who might need extra help. Binet cautioned that scores were rough and that intelligence was not a fixed quantity. Others adapted the approach, and the idea of an “intelligence quotient,” or IQ, took hold. Early versions compared a child’s “mental age” with their actual age. Modern tests, such as the Wechsler scales and the Stanford-Binet, instead compare a person’s performance with that of a large reference group of the same age.
The history also includes misuse. Tests were applied in the early 20th century to support discriminatory policies, including restrictive immigration views and forced sterilization programs in some countries, using claims that modern researchers regard as unscientific. That legacy is one reason the subject remains sensitive.
How scores work
Most modern IQ tests are scaled so the average score is 100 and the standard deviation is 15. That means roughly two-thirds of people score between 85 and 115, and roughly 95 percent between 70 and 130. Scores are not a measure of a fixed amount of something, such as height; they describe where someone’s performance falls compared with others who took the same test. Every score also carries a margin of error, so a difference of a few points between two people is usually not meaningful.
Typical tests include a mix of tasks: vocabulary and general knowledge, pattern and puzzle solving, working memory tasks such as repeating strings of digits, and processing speed. The way memory works affects performance on several of these.
General intelligence
Charles Spearman noticed in the early 1900s that people who did well on one mental test tended to do well on others. He proposed an underlying general factor, called g. The pattern, a positive correlation among diverse cognitive tests, is one of the more robust findings in psychology. Later models such as the Cattell-Horn-Carroll framework divide abilities into layers, with broad abilities such as fluid reasoning (solving novel problems) and crystallized knowledge (accumulated facts and vocabulary) sitting under a general factor.
What tests do capture
Intelligence test scores are statistically associated with things like school performance, years of education and, to a lesser extent, job performance and some health outcomes. These are group-level associations with plenty of exceptions, and the relationship between test scores and education runs both ways, since schooling itself raises scores. Studies of education and growth show how hard it is to separate cause from correlation.
Scores also tend to be fairly stable over years, especially from adolescence on, but they are not fixed. Training, health, schooling and circumstances can all shift performance.
What tests miss
- Creativity and originality. Tests reward correct answers to defined problems, not new ideas.
- Practical and social skills. Understanding people, handling daily problems and making wise decisions overlap only partly with test scores.
- Motivation and persistence. Effort and interest affect results and outcomes alike, as discussed in motivation psychology.
- Judgment. Smart people are also subject to cognitive bias.
- Wisdom, character and values. These are not intended to be captured at all.
Debates over the concept
Multiple intelligences. Howard Gardner proposed in 1983 that there are several relatively independent intelligences, such as musical, bodily-kinesthetic, interpersonal and linguistic. The idea is popular in education. Many psychologists criticize it for lacking solid testing, noting that the different abilities tend to correlate and that some of Gardner’s categories look more like talents or skills. Robert Sternberg suggested analytical, creative and practical aspects. These views raise valid points about breadth, though they have attracted less empirical support than the standard psychometric approach.
Nature and nurture. Studies of twins and adoptees suggest that genetic differences account for a substantial share of variation in IQ scores in the populations studied, and that the share tends to appear larger in adulthood than in childhood. But heritability describes populations, not individuals, and it does not mean scores are unchangeable. Environment matters, and the role of inherited traits is complex.
The Flynn effect. James Flynn documented that average raw scores rose substantially across many countries during much of the 20th century, which is too fast to be explained by genes. Possible contributors include better nutrition, health, schooling and more exposure to abstract thinking. In some countries, the trend has recently stalled or reversed, and the reasons are debated.
Fairness and culture. Tests developed in one cultural and linguistic setting may favor people familiar with it. Test designers try to reduce this with careful item testing and non-verbal tasks, but concerns persist. Group differences in average scores have been reported, and the causes are discussed in research without a simple consensus; differences between groups say little about any individual.
Artificial systems. Computers can outperform people on many narrow tasks, which prompts the question of whether those tasks measure what we mean by intelligence, a question touched on in artificial intelligence.
A sensible way to think about it
IQ tests are among psychology’s best-validated measuring tools, at least for the narrow range of abilities they cover, and they can be useful in education and clinical assessment when administered by trained professionals. They are not a full measure of a person’s worth, potential or capability. A score is a snapshot with a margin of error, influenced by many factors. Nothing in this article is intended as an assessment of any individual, and anyone with questions about learning or cognitive abilities should speak with a qualified psychologist or educator.