Home › Personality Types › What an IQ Test Measures and What It Includes › Cattell-Horn-Carroll Theory of Intelligence

Cattell-Horn-Carroll Theory: The Three-Level Model Behind Modern Tests

Cattell-Horn-Carroll theory is the map almost every modern intelligence battery is built from: narrow skills at the bottom, a set of broad abilities in the middle, and a general factor at the top. This is what each level means and what is still argued about.

6 sources cited10 min read
Cattell-Horn-Carroll Theory of Intelligence: A woman in a cream jacket stacks wooden blocks into a pyramid beside an hourglass at a table of tasks.

In short

Cattell-Horn-Carroll theory, usually shortened to CHC, describes cognitive abilities as three stacked levels: dozens of narrow skills at the bottom, each measurable by a specific task, broad abilities in the middle such as fluid reasoning, comprehension-knowledge, working memory capacity, visual processing and processing speed, and a single general factor, g, at the top. It came from merging two research traditions: Raymond Cattell and John Horn's work on fluid and crystallized ability, and John Carroll's re-analysis of hundreds of factor-analytic data sets. The model is why a modern score report shows several domain scores rather than one number, and what counts as g remains genuinely disputed.

What it is
A three-level map of measured cognitive abilities
The three levels
Narrow skills, broad abilities, a general factor
Where it came from
Cattell and Horn's fluid-crystallized work plus Carroll's re-analysis
Why it matters
It decided which subtests go in modern batteries
Still argued about
What g is, and how many broad abilities there are
Test
IQ Test: Free Reasoning Quiz

The model in one picture

CHC theory answers a structural question rather than a philosophical one. Given that people's scores on many different mental tasks correlate with each other, but not perfectly, what is the simplest arrangement that accounts for the pattern? The answer the evidence kept producing has three layers.

At the bottom, stratum one, sit dozens of narrow abilities, each tied closely to a particular kind of task: vocabulary knowledge, spatial scanning, memory span, numerical facility, perceptual speed. These are the things an individual subtest actually measures, and there are many more of them than any test battery samples.

In the middle, stratum two, sit broad abilities. Each one groups narrow skills that rise and fall together, such as the cluster of reasoning tasks that forms fluid reasoning or the cluster of knowledge tasks that forms comprehension-knowledge. At the top, stratum three, sits a single general factor, g, accounting for the fact that all of the broad abilities are positively correlated with one another.

The crucial thing about the picture is that it is descriptive. It says the measurements arrange themselves this way. It does not, by itself, claim that there are three kinds of thing inside a head.

Where it came from

The first thread is Raymond Cattell's distinction between fluid and crystallized ability, which he drew in the 1940s and put to an explicit test in a 1963 paper in the Journal of Educational Psychology. Fluid ability is reasoning applied to something unfamiliar. Crystallized ability is the accumulated knowledge and verbal skill a person has built up. Cattell argued these behave differently enough that treating them as one quantity loses information.

John Horn, Cattell's student, spent decades extending that two-part model. Working through large data sets, he kept finding additional broad abilities that were not reducible to either Gf or Gc: visual processing, auditory processing, short-term memory, long-term retrieval, processing speed and quantitative knowledge among them. By the 1990s the extended Gf-Gc model described something closer to nine or ten broad abilities.

The second thread is John Carroll's 1993 book Human Cognitive Abilities, which re-analyzed hundreds of factor-analytic data sets collected across most of the twentieth century using a consistent method. Carroll's conclusion was the three-stratum model: a large number of narrow abilities, a smaller number of broad ones, and a general factor above them.

The merge happened in the late 1990s. Horn's extended model and Carroll's hierarchy described the same middle layer with different names and slightly different boundaries, and test developers needed one vocabulary. Kevin McGrew's 2009 editorial in Intelligence set out the combined framework and argued explicitly for adopting the CHC taxonomy as a common nomenclature, so that researchers using different batteries would at least be naming the same things.

The broad abilities, in plain words

The middle layer is where the model earns its keep, because this is the level at which differences between people become recognizable in daily life. The descriptions below use the everyday version of each ability rather than the abbreviations.

  • Fluid reasoning: working out a rule nobody taught you. Figuring out how an unfamiliar appliance works without the manual.
  • Comprehension-knowledge: what you know and how precisely you can say it. Explaining a complicated situation so a stranger follows it.
  • Quantitative knowledge: acquired mathematical skill and fact. Checking whether a quoted discount is actually the better deal.
  • Reading and writing: acquired literacy skill. Getting through a dense contract and noticing the clause that matters.
  • Visual processing: handling shape, form and space in your head. Judging whether the sofa will fit through the doorway before you lift it.
  • Auditory processing: discriminating patterns in sound. Picking one voice out of a noisy room, or hearing that an instrument is slightly flat.
  • Short-term and working memory: holding and rearranging information while you use it. Keeping a phone number in mind while someone keeps talking.
  • Long-term storage and retrieval: getting stored information back fluently. Producing the name of the person walking toward you, on time.
  • Processing speed: doing easy things quickly and accurately. Scanning a spreadsheet column for the one wrong figure.
  • Reaction and decision speed: how fast a simple decision gets made. Braking when a cyclist appears.

The catalog has grown over the years, and recent presentations of CHC list roughly sixteen broad abilities, including motor and sensory domains that no intelligence battery measures. That growth is a feature of the research program rather than a sign of confusion: the model was designed to be extended as new data sets arrived, which is exactly what McGrew argued for.

How the model shaped the tests

If you have a score report from the last twenty-five years, you are holding CHC theory in applied form. The Stanford-Binet fifth edition reports five factors, which are fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory, and those labels are CHC labels. The current Wechsler adult scale reports verbal comprehension, visual spatial, fluid reasoning, working memory and processing speed for the same reason.

That was not always the arrangement. Earlier Wechsler editions split subtests into verbal and performance halves, a division based on how the task was presented rather than on what it measured. CHC research showed that the presentation split cut across the ability structure, which is why modern editions separate visual-spatial processing from fluid reasoning even though both are mostly nonverbal.

The practical result is that a report now tells you more than a single number does. Two people with identical full-scale scores can have opposite profiles underneath, and the profile is usually what the referral question was actually about.

Fluid and crystallized across a lifetime

The distinction Cattell drew matters most when you look at how abilities move with age. Crystallized ability, the accumulated knowledge and vocabulary, holds up well and can keep rising into later adulthood, because it is built from everything a person has read, done and learned. Fluid reasoning, working memory and processing speed behave differently: they peak earlier and decline gradually.

How early is genuinely contested. Timothy Salthouse's 2009 analysis argued that cross-sectional data show age-related declines in reasoning, memory and speed beginning in the twenties and thirties, much earlier than most people assume. The same journal issue carried a direct rebuttal from K. Warner Schaie, who argued that longitudinal designs put the onset considerably later and that cross-sectional comparisons confuse aging with differences between birth cohorts. Both arguments are still live.

The useful takeaway is practical rather than statistical. Someone in their sixties who is slower on a timed matrix task but sharper than ever on vocabulary and judgment is showing the normal shape of adult cognitive change, not a decline in intelligence. Which is also why a single full-scale number is a poor instrument for tracking anyone over decades.

What is still argued about

The biggest open question is what sits at the top. Everyone agrees on the data: scores on diverse mental tasks are positively correlated, which is why a general factor appears when you analyze them. What that factor is remains disputed. One reading treats g as a real underlying capacity that feeds every task. Another treats it as a statistical summary of many partly independent processes that develop together, in which case extracting it tells you about the correlations rather than about a thing.

Horn himself never accepted a single general factor as a meaningful entity, which is why the merged model is sometimes presented with g at the top and sometimes without it. That disagreement sits inside the theory's own name.

The number of broad abilities is also unsettled, and it moves as new data sets are analyzed and as the boundaries between neighboring abilities are redrawn. There is further evidence that the structure itself is not constant across people: Elliot Tucker-Drob's 2009 analysis found that abilities are less differentiated in childhood and more differentiated later, meaning the map is somewhat different at different ages. A model that changes shape across the lifespan is harder to use and closer to the truth.

Why this matters when you read a score

Three things follow for anyone holding a result. First, read the domain scores before the total. If fluid reasoning and processing speed are far apart, the average of them describes nobody, and the gap is usually the finding.

Second, a low score in one broad ability is not a statement about intelligence. Working memory and processing speed are the two most sensitive to sleep, anxiety, pain and medication, so a dip there has a long list of ordinary explanations before anything about capability.

Third, the model explains why no short quiz can produce a full picture. Covering the broad abilities needs several different kinds of material, including spoken items, timed items and visual items, and the only way to get them all is a battery given in person. A text-only quiz samples a slice of the fluid reasoning and knowledge corner and leaves the rest of the map blank.

What the research says

By the standards of psychological theory, this one is unusually well supported. The layered structure has been recovered again and again from independent data sets, across decades, countries and instruments, and Carroll's survey established it using a consistent method rather than a single study. The broad abilities are not a guess about how thinking ought to work: they are the groupings the measurements keep producing.

What CHC does not do is explain where the structure comes from. It is a taxonomy of measured performance, not an account of brain mechanisms, and it says nothing about what makes one person score higher than another. Treating a broad-ability score as a quantity of a faculty is reading more into the model than it supports.

It is also silent on everything outside the cognitive domain. Motivation, persistence, creativity, judgment and character are not in the map, which is a limitation of the subject matter rather than of the theory. The honest summary is that CHC is the best available description of how cognitive test performance is organized, and nothing more than that.

How to find out where you land

A full CHC-style profile comes from an individually administered battery given by a qualified examiner, which is what the Stanford-Binet and the Wechsler scales produce. If that is what you need, the route is a professional appointment and a written report.

Psychology.com's IQ Test is a free reasoning quiz: 30 original text-only questions, untimed, reported as an estimated range with a five-category breakdown, and an optional in-depth report is available. It is self-administered, unproctored and unnormed, so it is not an IQ score and not a clinical assessment, and it samples only the kinds of reasoning that can be written down.

FeatureFluid reasoningComprehension-knowledge
What it isReasoning about something unfamiliarWhat you have learned and can express
Typical taskInferring the rule in a pattern or seriesDefining words, explaining how two things relate
Depends on schoolingMuch lessHeavily
Depends on languageCan be tested with pictures onlyStrongly, and on which language is strongest
Across adulthoodPeaks early, declines graduallyHolds, and can keep rising
Sensitive to sleep and stressYes, noticeablyMuch less
What a short written quiz can samplePart of it, through reasoning itemsPart of it, through vocabulary and knowledge items
Everyday versionWorking out a new system without a manualKnowing the field and explaining it well

Key takeaways

  • CHC theory arranges cognitive abilities in three levels: dozens of narrow skills, a middle layer of broad abilities, and a general factor on top.
  • It merges two traditions: Cattell and Horn's work separating fluid from crystallized ability, and Carroll's 1993 re-analysis of hundreds of factor-analytic data sets.
  • The broad abilities are the level that shows up in daily life, and they are why a modern score report gives five domain scores instead of one number.
  • Crystallized knowledge holds or rises into later adulthood while fluid reasoning, working memory and speed decline, though when that decline starts is still argued between cross-sectional and longitudinal researchers.
  • What g actually is remains disputed, Horn himself never accepted it as a real entity, and the number of broad abilities keeps moving as new data arrives.
  • The model is a description of measured performance, not an account of brain mechanisms, and it is silent on motivation, creativity and character.

Curious which reasoning comes easiest to you?

Our free IQ Test samples two corners of this map, reasoning and knowledge, and reports an estimated range with a five-category breakdown. It is not a full CHC profile.

Take the free IQ Test

Cite this source

Fontane Pennock, S. (2026, November 8). Cattell-Horn-Carroll Theory of Intelligence. Psychology.com. https://psychology.com/types/cattell-horn-carroll-theory

Frequently asked questions

What is the Cattell-Horn-Carroll theory of intelligence?

A three-level description of how cognitive abilities are organized. Dozens of narrow skills sit at the bottom, each measured by a specific kind of task. A middle layer of broad abilities groups the narrow skills that move together, such as fluid reasoning or processing speed. A single general factor, g, sits on top, accounting for the fact that all the broad abilities are positively correlated.

What is the difference between Cattell-Horn and Cattell-Horn-Carroll?

Cattell and Horn built an extended fluid-crystallized model with around nine or ten broad abilities and no general factor on top, because Horn did not accept g as a meaningful entity. Carroll's independent re-analysis produced a three-stratum hierarchy that did include g. The merged CHC framework, set out for common use by Kevin McGrew, combines the two vocabularies.

What is the difference between fluid and crystallized intelligence?

Fluid reasoning is working out something unfamiliar with no learned method to fall back on. Crystallized ability is the knowledge and verbal skill you have accumulated. The practical difference shows up with age: crystallized ability holds or rises into later adulthood, while fluid reasoning, working memory and processing speed peak earlier and decline gradually.

How many broad abilities are there in CHC theory?

It depends which presentation you read, which is itself part of the answer. The extended Cattell-Horn model described around nine or ten, and recent CHC presentations list roughly sixteen, including motor and sensory domains that no intelligence battery measures. The count moves as new data sets are analyzed, which the framework's authors treated as intended rather than a problem.

Why does CHC theory matter for my IQ score?

Because it decided what is in the test. The five domains on a modern score report are CHC broad abilities, and the model is the reason a battery reports a profile instead of one figure. When your domain scores are far apart, the total is the average of things that disagree, and the profile is usually what the assessment was for.

Is CHC theory the same as saying intelligence is one thing?

No. The model includes a general factor because diverse test scores correlate, but it treats the broad abilities as separately meaningful and separately measurable. Whether g is a real underlying capacity or a statistical summary of many partly independent processes is an open question, and the theory can be used either way.

Which tests are built on CHC theory?

Most current batteries, in practice. The Stanford-Binet fifth edition reports fluid reasoning, knowledge, quantitative reasoning, visual-spatial processing and working memory. The current Wechsler adult scale reports verbal comprehension, visual spatial, fluid reasoning, working memory and processing speed. The Woodcock-Johnson family, developed alongside the CHC work, is the most explicitly CHC-aligned of them.

Related types

References

  1. Cattell RB. Theory of fluid and crystallized intelligence: A critical experiment. Journal of Educational Psychology. 1963;54(1):1-22. doi.org
  2. Carroll JB. Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Cambridge University Press; 1993. doi.org
  3. McGrew KS. CHC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence. 2009;37(1):1-10. doi.org
  4. Flanagan DP. The Cattell-Horn-Carroll Theory of Cognitive Abilities. In: Encyclopedia of Special Education. Wiley; 2008:368-382. doi.org
  5. Salthouse TA. When does age-related cognitive decline begin? Neurobiology of Aging. 2009;30(4):507-514 (with K. Warner Schaie's reply in the same issue, 30(4):528-529). doi.org
  6. Tucker-Drob EM. Differentiation of cognitive abilities across the life span. Developmental Psychology. 2009;45(4):1097-1118. doi.org
Important: A personality type, attachment style, or result pattern is an educational way of describing tendencies, not a diagnosis, a fixed identity, or a substitute for professional care. If you are struggling, reach out to a licensed mental-health professional.