From Ratios to a Bell Curve
Modern intelligence testing largely uses the Wechsler scales — the WAIS (Wechsler Adult Intelligence Scale) for adults and the WISC (Wechsler Intelligence Scale for Children) for children — which replaced Terman's old ratio IQ formula with deviation IQ: instead of comparing mental age to chronological age, a person's raw score is compared to the scores of a large, representative sample of SAME-AGE peers, converted onto a standardized scale with a mean of 100 and a standard deviation of 15. This solves the ratio formula's breakdown for adults, since it no longer depends on mental age continuing to climb.
Deviation IQ scores form a normal distribution (a symmetric bell curve): about 68% of people score within one standard deviation of the mean (between 85 and 115), and about 95% score within two standard deviations (between 70 and 130). This standardized, peer-relative approach also makes IQ scores directly comparable across different age groups, since everyone is compared only to others their own age.
| IQ range | Approximate population share |
|---|---|
| 85-115 (±1 SD) | About 68% |
| 70-130 (±2 SD) | About 95% |
A striking pattern discovered by researcher James Flynn, now called the Flynn effect, is that raw intelligence test scores have risen substantially across successive generations throughout the 20th century — commonly cited at roughly 3 points per decade in many countries — even though genes change far too slowly to explain a shift this fast. Because IQ tests are periodically RE-NORMED to keep the population mean at 100, this generational rise is easy to miss unless you compare raw scores directly across test versions; likely explanations include improved nutrition, more years of schooling, and increased familiarity with abstract test-taking and reasoning tasks in modern life.
Common pitfall: assuming a person's current deviation IQ score is meaningful in isolation, without reference to which NORM GROUP it was calculated against. Because tests are periodically re-normed to account for the Flynn effect, a raw test score from decades ago cannot be directly compared to a score from today without adjustment.
A bell curve: a bare symmetric curve peaking at 100; the ±1 SD band (85-115) shades in with a '68%' label; the ±2 SD band (70-130) shades in wider with a '95%' label, both bands visible together at the end.