The Stanford–Binet Test: Structure, Scores and the Scale That Changed
- Written by
- IQCognify Editorial Team
- Reviewed for accuracy
- IQCognify Research Review Process
- Last updated
- Reading time
- 6 min
Quick answer
The Stanford–Binet Intelligence Scales is the oldest IQ test still in use, and the reason the phrase “IQ” exists at all. Its fifth edition assesses five cognitive factors in both verbal and nonverbal form across ten subtests, for ages two to eighty-five and beyond. It also did something in 2003 that quietly invalidated every comparison with its own history: it changed the standard deviation from 16 to 15.
What the SB5 measures
The fifth edition, published by Gale Roid in 2003, was rebuilt around the Cattell–Horn–Carroll model — the same taxonomy that underpins most contemporary ability testing. Its design is unusually symmetrical: five cognitive factors, each assessed twice, once verbally and once nonverbally.
| Factor | What it assesses |
|---|---|
| Fluid Reasoning | Solving novel problems; inferring rules from unfamiliar material |
| Knowledge | Accumulated information, vocabulary, learned concepts |
| Quantitative Reasoning | Numerical relationships and mathematical problem-solving |
| Visual-Spatial Processing | Perceiving and manipulating spatial relations and patterns |
| Working Memory | Holding and transforming information over short intervals |
Five factors × two domains gives ten subtests. From those the test reports a Full Scale IQ, a Verbal IQ, a Nonverbal IQ, and an index for each of the five factors.
Why the verbal/nonverbal symmetry matters
Most batteries confound ability with modality — quantitative reasoning gets tested verbally, spatial reasoning nonverbally. The SB5 tests each factor both ways, which makes it possible to separate a genuine weakness in a cognitive factor from a weakness in expressing it through language.
How the test adapts to you
The Stanford–Binet does not administer the same items to everyone. It begins with two routing subtests: Nonverbal Fluid Reasoning and Verbal Knowledge. Performance on those determines where in each subsequent subtest the examiner starts.
The purpose is efficiency and precision at once. A capable adult is not made to work through items designed for a six-year-old, and a struggling child is not confronted with material far beyond them. Every examinee spends most of their time on items near their own level, which is where the measurement information actually is.
It is also the reason the test spans ages two through eighty-five in a single instrument, where the Wechsler family requires separate scales for preschoolers, children and adults.
The WAIS intelligence test — the other major individually administered battery, and how its five indices compare.The standard deviation changed — and almost nobody says so
This is the single most consequential fact about Stanford–Binet scores, and it is missing from nearly every page that lists them.
An IQ score is a position on a normal distribution. Fixing the mean at 100 is not enough to define it; you also have to fix the spread. Wechsler tests have always used a standard deviation of 15. The Stanford–Binet, through every edition up to and including the famous Form L-M, used 16. The SB5, in 2003, switched to 15.
| Percentile | SB5 and Wechsler (SD 15) | Older Stanford–Binet (SD 16) |
|---|---|---|
| 98th (Mensa's threshold) | 130 | 132 |
| 99.9th | ~146 | ~149 |
| Two SDs above the mean | 130 | 132 |
| Three SDs above the mean | 145 | 148 |
What this means in practice
A 132 obtained on Form L-M and a 130 obtained on the SB5 describe the same person in the same place. Neither is higher than the other. High-IQ societies publish different qualifying numbers for different editions for exactly this reason — the numbers are not on the same scale.
This is also where a great many inflated historical scores come from. A gifted child assessed on Form L-M in 1975 and a gifted child assessed on the SB5 in 2015 may be identically able, and the older one will appear to have scored higher. Any comparison across those editions that does not convert first is meaningless.
IQ percentiles explained — why the percentile, not the score, is the quantity that transfers between tests.Where “IQ” came from
Alfred Binet and Théodore Simon built the original scale in Paris in 1905, to identify schoolchildren needing additional support. It produced a mental age, not a quotient, and Binet was explicit that it measured current performance rather than fixed capacity.
Lewis Terman at Stanford translated, extended and renormed it in 1916. His revision adopted the intelligence quotient — mental age divided by chronological age, multiplied by 100 — and it was that publication which fixed the term in the language.
Ratio IQ is obsolete
Dividing mental age by chronological age works reasonably for children and collapses entirely for adults, whose mental age stops tracking their birthday. David Wechsler's 1939 deviation IQ replaced it. Every modern Stanford–Binet score, including every SB5 score, is a deviation IQ. The ratio formula survives only in historical estimates — which is why figures like “Goethe had an IQ of 210” cannot be compared with a modern score.
Terman's revision also carried assumptions that the field has since repudiated, and the history of American intelligence testing in the first half of the twentieth century is not a comfortable one.
The history of IQ testing — from Binet's schoolchildren to the modern batteries, including the parts that are not flattering.Stanford–Binet or Wechsler?
Both are individually administered by a qualified psychologist, both take one to two hours, both are well normed and both are built on the CHC framework. For most purposes they are interchangeable, and a person's scores on the two will usually agree closely.
| Stanford–Binet 5 | Wechsler (WPPSI/WISC/WAIS) | |
|---|---|---|
| Age range | 2 to 85+, one instrument | Three separate scales by age band |
| Structure | 5 factors × verbal and nonverbal | 5 indices, not symmetrically crossed |
| Adaptive routing | Yes, via two routing subtests | Start points by age, with reversal rules |
| Standard deviation | 15 (since 2003) | 15 throughout |
| Common use | Preschool assessment, gifted identification, wide age spans | The default adult and school-age battery |
The Stanford–Binet is often preferred at the extremes. Its nonverbal domain is fully developed rather than an afterthought, which helps when assessing children with limited language, and its structure holds up better at the top of the range where the Wechsler scales begin to run out of difficulty.
The WISC-V test — the Wechsler children's scale, and the factor structure its own manual could not defend.IQ testing for children — what parents should know before a child sits either battery.Myths and facts about the Stanford–Binet
| Myth | Fact |
|---|---|
| Stanford–Binet scores use a standard deviation of 16 | Editions up to Form L-M did. The SB5 has used 15 since 2003, making it directly comparable to Wechsler scores. |
| A 132 on Stanford–Binet beats a 130 on the WAIS | On the old SD-16 scale, 132 is the 98th percentile — exactly what 130 is on an SD-15 scale. They describe the same person. |
| The Stanford–Binet gives a ratio IQ | It has reported deviation IQs for decades. Mental age divided by chronological age has not been used in modern editions. |
| Binet invented the IQ | Binet produced a mental age and warned against treating it as fixed capacity. Terman's 1916 Stanford revision introduced the quotient. |
| It is just an older version of the WAIS | They are separate instruments from different publishers. The SB5 crosses all five factors with a verbal and nonverbal domain; the Wechsler scales do not. |
| You can take the Stanford–Binet online | It is a restricted instrument administered one-to-one by a qualified professional. No online version exists. |
Frequently asked questions
What is the Stanford–Binet test?+
It is an individually administered intelligence test, the oldest still in use, tracing to Binet and Simon's 1905 scale and Lewis Terman's 1916 Stanford revision. The current fifth edition measures five cognitive factors — fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory — in both verbal and nonverbal form, for ages 2 to 85 and beyond.
Does the Stanford–Binet use a standard deviation of 15 or 16?+
The fifth edition, published in 2003, uses 15. Every earlier edition, including Form L-M, used 16. This means historical Stanford–Binet scores sit on a different scale from modern ones and cannot be compared point-for-point.
Is a 132 on the Stanford–Binet the same as a 130 on the WAIS?+
If the 132 came from an older SD-16 edition, yes — both correspond to roughly the 98th percentile, which is Mensa's threshold. If the 132 came from the SB5, which uses SD 15, it is slightly higher than a WAIS 130. The edition matters more than the number.
What is the difference between the Stanford–Binet and the Wechsler tests?+
Both are one-to-one batteries built on the Cattell–Horn–Carroll framework and both now use a standard deviation of 15. The Stanford–Binet covers ages 2 to 85+ in a single instrument and crosses all five of its factors with a verbal and a nonverbal domain. The Wechsler family uses three separate scales by age band and does not cross its indices symmetrically.
How long does the Stanford–Binet take?+
Typically one to two hours for the full battery of ten subtests, though the adaptive routing procedure means the exact length depends on the examinee. Two routing subtests are given first, and their results set the starting difficulty for everything that follows.
Who invented the IQ score?+
Binet and Simon's 1905 scale produced a mental age rather than a quotient, and Binet cautioned against treating it as a fixed capacity. Lewis Terman's 1916 Stanford revision adopted the intelligence quotient — mental age divided by chronological age, multiplied by 100 — and popularised the term.
Is the Stanford–Binet a ratio IQ test?+
Not any more. Ratio IQ works only while mental age tracks chronological age, which it stops doing in adulthood. David Wechsler's deviation IQ, introduced in 1939, replaced it, and modern Stanford–Binet editions report deviation IQs. Historical genius estimates using ratio IQ cannot be compared with modern scores.
Can I take the Stanford–Binet online?+
No. It is a restricted psychological instrument administered one-to-one by a qualified professional, with adaptive routing that depends on a trained examiner's judgement. Online tests advertising themselves as the Stanford–Binet are not that test.
Sources
This guide draws on standard psychometric references and peer-reviewed research:
- 1.Roid, G. H. (2003). Stanford–Binet Intelligence Scales, Fifth Edition (SB5): Technical Manual. Riverside Publishing.
- 2.Riverside Insights — Stanford–Binet Intelligence Scales.
- 3.Terman, L. M. (1916). The Measurement of Intelligence: The Stanford Revision and Extension of the Binet–Simon Intelligence Scale. Houghton Mifflin.
- 4.Binet, A., & Simon, T. (1905). New methods for the diagnosis of the intellectual level of subnormals. L’Année Psychologique, 12.
- 5.Terman, L. M. (1925). Genetic Studies of Genius, Vol. 1. Stanford University Press.
- 6.Wechsler, D. (1939). The Measurement of Adult Intelligence. Williams & Wilkins.
- 7.McGrew, K. S. (2009). “CHC theory and the human cognitive abilities project.” Intelligence, 37(1).
- 8.Carroll, J. B. (1993). Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Cambridge University Press.
- 9.American Mensa — Qualifying Test Scores and evidence-of-eligibility policy.
- 10.Deary, I. J. (2020). Intelligence: A Very Short Introduction (2nd ed.). Oxford University Press.
- 11.American Psychological Association (APA)
Sources are provided for further reading. Organization links point to official sites; academic works are cited in full. See our research standards and editorial team.
Find out your IQ
Take the free IQ test and get your score, percentile, and a full cognitive breakdown in about 12 minutes.
Start Free Test