Updated on
August 25, 2026
CHC Theory: The Cattell-Horn-Carroll Model Explained
CHC theory explained: Cattell and Horn's Gf-Gc model, Carroll's three-stratum theory, later synthesis, broad abilities, evidence and assessment limits.
What is CHC theory?
CHC theory is an influential, evolving psychometric taxonomy of cognitive abilities. It brings Cattell and Horn's extended Gf-Gc tradition into dialogue with Carroll's three-stratum theory. It organises patterns in cognitive-test performance, but versions differ in their ability labels and treatment of a general factor. CHC is not a brain map, diagnosis, learning-style system or teaching method.
CHC theory maps patterns in results from cognitive tests. Its full name is Cattell-Horn-Carroll theory. Cattell and Horn studied fluid, crystallised and other broad abilities.
Carroll (1993) based his three-stratum account on more than 460 factor-analytic data sets. Later scholars brought these traditions under the CHC label, although the earlier models were not the same (McGrew, 2009).
CHC is mainly a psychometric taxonomy. It gives researchers and qualified assessors a shared set of names for patterns across tasks. It is not a map of brain regions, a diagnosis, a system of learner types or a guide to how someone should be taught.
Key Takeaways
- CHC is a family of related versions. Cattell, Horn, Carroll and later CHC writers did not all propose the same hierarchy.
- The three-stratum pyramid comes from Carroll's work. It puts narrow abilities below broad abilities, with a general factor, g, at the top.
- The taxonomy changes. Names, codes and boundaries vary by version and test battery.
- Scores are estimates. Sound use requires the test manual, norms, confidence intervals, base rates and purpose of the assessment.
- A profile is not a prescription. One score cannot diagnose SEND, explain behaviour or identify an effective intervention.
What Is CHC Theory?
CHC theory is a hierarchical taxonomy of human cognitive abilities. It grew from factor-analytic research. It gives scholars shared terms for narrow task results, broad abilities and, in versions based on Carroll, a general factor.
It describes a statistical structure. It does not explain the causes of learning or intelligence.
A taxonomy puts related observations under common labels. CHC researchers ask which test results tend to rise and fall together. They infer factors from those patterns. These factors are latent constructs, so they cannot be observed directly.
This makes CHC different from an instructional learning theory. It does not set out stages through which people gain knowledge. Nor does it show which task or intervention will improve a person's attainment. Its main role is to describe structure and supply terms.
The phrase “CHC theory” can hide real disagreement. Some writers use it for a broad synthesis of the Cattell-Horn and Carroll traditions. Others separate Carroll's model, later 3S-CHC revisions and extended Gf-Gc versions that do not put one g above the broad abilities (McGrew, 2023).
The Lineage: Cattell, Horn and Carroll
CHC grew from several linked research programmes. It did not come from one experiment or one complete model. Cattell tested a split between fluid and crystallised intelligence. Horn added more broad abilities.
Carroll built a three-stratum theory. McGrew then played a central part in the later CHC synthesis and its terms.
The fluid-crystallised split has a complex origin. Letters between Donald Hebb and Raymond Cattell helped shape it (Brown, 2016). Cattell then tested fluid and crystallised factors in his 1963 study.
Fluid reasoning, Gf, covers induction and deduction with fairly new information. It is often called reasoning without prior knowledge. Yet no human test is wholly free from language, task familiarity or past experience.
Crystallised ability, Gc, covers learned comprehension and knowledge that a culture values. Our separate account of fluid intelligence explores this split in more depth.
Horn and Cattell went beyond the first two-factor account. They found further broad dimensions (Horn and Cattell, 1966). Horn's later extended Gf-Gc model did not adopt one top-level g in the same way as Carroll. A diagram should not imply that all three scholars proposed the same pyramid.
Carroll took a different route through the evidence. He reanalysed more than 460 data sets from much of the twentieth century. His 1993 book grouped abilities into three strata.
McGrew (2009) later set out the many links between Carroll's broad factors and the extended Cattell-Horn taxonomy. He also noted their differences.
The Three Strata: Narrow, Broad and General Ability
Carroll's theory has three strata. Fairly specific abilities sit at stratum I. Broad factors sit at stratum II, and general ability sits at stratum III. Later writers often use this hierarchy to explain CHC.
They should label it as Carroll-derived or 3S-CHC, not as an unchanged part of the earlier Cattell-Horn model.
Stratum I contains narrow abilities inferred from specific kinds of performance. A subtest may sample one or more of them. It is not a pure measure of one mental module. Task content, method, time limits and learned knowledge can all affect a score.
Stratum II groups linked narrow abilities into broad constructs. Examples include fluid reasoning, learned comprehension-knowledge, processing speed, visual processing and auditory processing. These names give scholars a shared vocabulary. Different batteries still measure them through different tasks and score groups.
Stratum III is general ability, or g. It reflects reliable variance shared across many cognitive tasks. Scholars still debate the weight of g and the broad factors. In one large cross-battery model, g and Gf could not be told apart in statistical terms (Caemmerer et al., 2020).
The strata form a statistical hierarchy, not a set of stages in human growth. A person does not pass through them. The pyramid does not place abilities in separate parts of the brain. It sums up how measured results vary together in specified data sets.
Text version of the study notes
Core definition: CHC theory is an influential and evolving psychometric taxonomy. It groups patterns in test results into narrow and broad abilities. Versions based on Carroll also include a general factor called g. CHC is not a brain map or teaching method.
Lineage: Cattell tested the Gf-Gc split in 1963. Horn added broad abilities but did not adopt Carroll's top g in the same form. Carroll built a three-stratum theory from more than 460 data sets. McGrew helped organise the later umbrella synthesis.
Three-stratum version: stratum I contains narrow abilities sampled through tasks. Stratum II contains broad abilities. Stratum III contains general ability.
Selected broad labels include Gf, Gc, Gs, Gv and Ga. Labels for memory, learning and retrieval vary by version.
Limits on use: ask which test, edition, norms and CHC version were used. Check confidence intervals, reliability and base rates. Add evidence about attainment, language, culture and observed need. Follow the qualified assessor's account of what the score can support.
Do not infer from one score: a low index cannot diagnose SEND, explain behaviour or prescribe an intervention. A high visual score does not make someone a visual learner. Scores with the same name in different batteries may not mean the same thing.
The Broad Abilities in CHC Theory
CHC ability lists depend on the version and date. This table gives common labels, not a fixed or full list. Each construct is broader than one subtest. Different batteries may sample the same named construct in different ways.
Swipe across to read all columns on a smaller screen.
| Label | Broad construct | Boundary |
|---|---|---|
| Gf | Induction and deduction with fairly new information. | Performance still draws on learned knowledge and task familiarity. |
| Gc | Acquired comprehension and culturally valued knowledge. | Language, education, culture and opportunity affect the sampled knowledge. |
| Gwm or Gsm | Working-memory capacity or short-term memory, based on the version. | Do not swap the labels without the right taxonomy and test manual. |
| Gs | Speed and fluency on simple, familiar cognitive tasks. | A score does not prove how quickly someone understands authentic school work. |
| Gv | Seeing, analysing and changing visual patterns. | It does not show a visual learning style or preferred way to teach. |
| Ga | Auditory discrimination and processing. | It is not equivalent to phonics skill, reading ability or a diagnosis. |
| Gl and Gr, or Glr | Learning efficiency and retrieval fluency, or an older joint storage-retrieval construct. | New and old lists divide this area in different ways. |
| Gq, Grw and Gt | Quantitative knowledge, reading-writing ability and reaction-decision speed. | Some formulations place acquired achievement domains or speed constructs differently. |
An old paper and a new report may use different codes without either being wrong. For example, some current accounts split working-memory capacity from older short-term-memory terms. Readers should check the cited CHC version and the test manual. They should not silently translate each label into one preferred list.
Evidence for CHC and Its Revisions
Evidence supports CHC as a useful shared taxonomy across several major test families. It does not support every score group or use. Studies across batteries have found strong links among broad constructs sampled by different tests (Reynolds et al., 2013; Caemmerer et al., 2020).
This is evidence about structure. It asks whether a chosen latent model can account for score patterns at group level. It does not show that two scores with the same name mean the same thing. Nor does it show that one person's profile is stable or can diagnose a learning difficulty.
Research still changes the taxonomy. McGrew (2023) separates Carroll's three-stratum legacy from broader uses of CHC and sets out new boundaries. Scholars still debate whether g, Gf and other broad factors can be told apart in a sound and stable way. This is part of current work on the theory.
Studies of single batteries offer another kind of evidence. Some WISC-V studies found similar factor structures in several national norm samples (Wilson et al., 2023, 2024). These results apply to the named tests, samples and models. Factorial invariance alone does not prove that a test is free from cultural bias, predicts equally well or is fair in every use.
Knowledge factors need special care. Gc samples learned knowledge that a culture values. Language, schooling and access to knowledge therefore matter.
A cross-national study of declarative knowledge found that items can reflect their nation of origin (Watrin et al., 2023). This does not make Gc invalid. It does limit claims that the score shows an inborn verbal capacity.
Why CHC Theory Matters in Schools
CHC matters in schools because its terms shape cognitive tests and reports. It can help a school leader see which construct a qualified assessor meant to sample. It cannot make school staff independent test experts. Nor can it turn a number straight into support.
Several intelligence batteries group some scores around constructs that align with CHC. The match is not pure. Wechsler indices, for example, combine tasks chosen by the publisher.
The tasks may load on more than one factor and share method effects. A report should name the test, edition and sound use of its scores. A familiar label may not mean the same thing in every battery.
Group reasoning tests sit next to this work, but they have their own manuals and studies. Our overview of the Cognitive Abilities Test explains this type of test. The CAT4 guide covers that battery's own verbal, quantitative, non-verbal and spatial scores. Those strands are not pure CHC factors unless the manual gives evidence for that claim.
The wider set of thinking skills assessments contains tools with different aims and claims. CHC can supply terms for constructs. The named test must supply its own norms, precision, validity evidence and rules for use.
How to Read a Cognitive Assessment Report
A cognitive report is a qualified judgement based on several sources. Start with the referral question. Then check the test and edition, the norm group and the assessor's stated limits. Do not start by finding the lowest score and giving it a cause.
Check uncertainty. Standard scores are estimates, not exact facts. Confidence intervals show a range of plausible values under the test model.
Whether two scores differ in a useful way depends on the manual. Key points include its comparison rules, reliability, standard error, base rate and the purpose of the test.
Check rarity and stability. A reliable score gap is not always rare in the norm sample. It may not matter for education or stay stable over time.
Watkins et al. (2022) found weak long-term stability in differences within WISC-V profiles in one clinical sample. That study does not settle the issue for every battery. It does show why a lone pattern needs care.
Check other evidence. Read cognitive data beside attainment, development, language, culture, behaviour and teaching history. Include direct evidence of the problem in school.
The Standards for Educational and Psychological Testing and APA guidance both stress intended use, fairness and precision. They also stress the need to join several sources of evidence.
Check the proposed response on its own merits. An adjustment or intervention needs evidence that it meets an observed need. It must also meet the right professional or legal rules.
A CHC label may help frame a hypothesis. It does not prove that the proposed response works.
School staff can ask sound questions without doing their own test analysis. Which construct did this task sample? How precise is the estimate? Is the difference rare, and does other evidence support it?
What other causes did the assessor consider? What separate evidence supports the advice?
What CHC Theory Does Not Tell You
CHC does not reveal a preferred way to teach, diagnose SEND or set an intervention. It cannot explain why one person lost track of an instruction, worked slowly or found reading hard. Those are questions about causes and education. A latent ability label cannot answer them on its own.
A high Gv score does not make someone a visual learner. Evidence on learning styles does not support matching teaching to a preferred sense. CHC does not provide an exception. A measured visual construct and a claimed teaching preference are different ideas.
A low memory score does not prove the cause of behaviour in class. Our article on working memory covers that construct and its evidence. Cognitive load theory deals with limits during learning. Neither source turns one cognitive index into a diagnosis.
CHC also differs from cognitivism. Cognitivist accounts ask how people represent, process and learn information. CHC mainly sorts the ways in which test results vary together. Shared terms such as memory or reasoning do not make the two frameworks the same.
Limitations and Critiques
CHC's main strength is its broad taxonomy. Its main risk is that readers may mistake a taxonomy for a cause. Factor analysis can show that results cluster under a chosen model. On its own, it cannot explain which mental or brain processes caused those links.
The hierarchy also depends on the model. Choices about tasks, samples, factor methods and rules can change which factors seem distinct. McGill and Dombrowski (2019) note that CHC has shaped test design. They also ask how closely some published tests match the model they claim to use.
General and broad scores both need evidence tied to their purpose. It is too strong to say that a full-scale score hides the “real profile”. General ability can account for substantial reliable common variance. The extra value of broad scores varies by battery and decision.
Methods based on profiles of learning disability face stronger criticism. Reviews and studies find low agreement between some patterns-of-strengths-and-weaknesses methods. They also find limits in external validity and weak support for matching interventions to profiles (Miciak et al., 2014, 2016; McGill et al., 2018).
These findings challenge some uses in diagnosis and treatment. They do not reject the whole CHC taxonomy.
Factor structure alone cannot prove fairness. A model may fit several groups while items, norms or language demands still differ. Links with later outcomes and the effects of test use may differ too. Sound use therefore needs evidence about the named test, population and decision.
CHC does not cover all forms of intelligence. Many cognitive batteries leave out some or all of motivation, creativity, adaptive skills and expertise. They may also miss cultural views of competent action. CHC should be judged by the question it answers, not used as a full list of human capability.
Frequently Asked Questions
What does CHC stand for?
CHC stands for Cattell-Horn-Carroll. The name refers to Cattell's Gf-Gc work, Horn's extended broad-ability model and Carroll's three-stratum theory. It also marks later scholarship that brought their many areas of agreement under one name.
Is CHC theory the same as IQ?
No. IQ is a score, or set of scores, from a named test. CHC is a taxonomy for cognitive constructs and their hierarchy. A general score may relate to stratum III g.
Broad and narrow scores may sample other levels. The exact link depends on the battery.
How many broad abilities are in CHC theory?
There is no fixed number for all time. Published lists have grown and changed. New versions may split constructs that old accounts joined. A reader should name the version and source instead of treating one count as permanent.
Does the WISC-V measure CHC abilities?
Several WISC-V constructs have a rough match with CHC classes. Researchers have also tested CHC-related models of its scores. Yet its indices are score groups set by the publisher for that battery. They are not pure CHC broad abilities that can be swapped across tests.
Can a CHC profile diagnose a learning disability?
Not on its own. Cognitive tests may help answer some referral questions. Profile methods still have limits in reliability, agreement and value for treatment.
Diagnosis needs the right professional framework. It also needs direct evidence of attainment and impairment, context, development and qualified judgement.
Should teaching be matched to a learner's strongest CHC ability?
No general score-to-teaching rule has sound support. A strength may help form a qualified hypothesis. A teaching decision still needs evidence about the real skill, task and intervention. CHC does not validate learning styles or show that teaching to a cognitive strength will improve attainment.
Further Reading
- CHC theory and the human cognitive abilities project
McGrew (2009). An integrative history of the Cattell-Horn and Carroll traditions. - Carroll's Three-Stratum Cognitive Ability Theory at 30 Years
McGrew (2023). A current clarification and revision of 3S-CHC. - Beyond individual intelligence tests
Caemmerer, Keith and Reynolds (2020). Cross-battery structural evidence and an important g-Gf boundary. - Cognitive profile analysis in school psychology
McGill, Dombrowski and Canivez (2018). A critical review of profile interpretation. - Standards for Educational and Psychological Testing
AERA, APA and NCME (2014). Professional standards for score validity, precision, fairness and use. - APA Guidelines for Psychological Assessment and Evaluation
APA (2020). Guidance on purpose, norms, culture, uncertainty and integration of evidence.
References
Brown, R. E. (2016). Hebb and Cattell: The genesis of the theory of fluid and crystallized intelligence. Frontiers in Human Neuroscience, 10, 606.
Caemmerer, J. M., Keith, T. Z., & Reynolds, M. R. (2020). Beyond individual intelligence tests: Application of Cattell-Horn-Carroll theory. Intelligence, 79, 101433.
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge University Press.
Cattell, R. B. (1963). Theory of fluid and crystallized intelligence: A critical experiment. Journal of Educational Psychology, 54(1), 1-22.
Horn, J. L., & Cattell, R. B. (1966). Refinement and test of the theory of fluid and crystallized general intelligences. Journal of Educational Psychology, 57(5), 253-270.
McGill, R. J., Dombrowski, S. C., & Canivez, G. L. (2018). Cognitive profile analysis in school psychology: History, issues, and continued concerns. Journal of School Psychology, 71, 108-121.
McGrew, K. S. (2009). CHC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence, 37(1), 1-10.
McGrew, K. S. (2023). Carroll's three-stratum cognitive ability theory at 30 years. Journal of Intelligence, 11(2), 32.
McGill, R. J., & Dombrowski, S. C. (2019). Critically reflecting on the origins, evolution, and impact of the Cattell-Horn-Carroll model. Applied Measurement in Education, 32(3), 216-231.
Miciak, J., Fletcher, J. M., Stuebing, K. K., Vaughn, S., & Tolar, T. D. (2014). Patterns of cognitive strengths and weaknesses: Identification rates, agreement, and validity for learning disabilities identification. School Psychology Quarterly, 29(1), 21-33.
Miciak, J., Williams, J. L., Taylor, W. P., Cirino, P. T., Fletcher, J. M., & Vaughn, S. (2016). Do processing patterns of strengths and weaknesses predict differential treatment response? Journal of Educational Psychology, 108(6), 898-909.
Reynolds, M. R., Keith, T. Z., Flanagan, D. P., & Alfonso, V. C. (2013). A cross-battery, reference variable, confirmatory factor analytic investigation of the CHC taxonomy. Journal of School Psychology, 51(4), 535-555.
Watkins, M. W., Canivez, G. L., Dombrowski, S. C., McGill, R. J., Pritchard, A. E., Holingue, C. B., & Jacobson, L. A. (2022). Long-term stability of Wechsler Intelligence Scale for Children-fifth edition scores in a clinical sample. Applied Neuropsychology: Child, 11(3), 422-431.
Watrin, L., Schroeders, U., & Wilhelm, O. (2023). Gc at its boundaries: A cross-national investigation of declarative knowledge. Learning and Individual Differences, 103, 102267.
Wilson, C. J., Bowden, S. C., Byrne, L. K., Vannier, L.-C., Hernandez, A., & Weiss, L. G. (2023). Cross-national generalizability of WISC-V and CHC broad ability constructs across France, Spain, and the US. Journal of Intelligence, 11(8), 159.
Wilson, C. J., Bowden, S. C., Batty, A. M., Byrne, L. K., & Weiss, L. G. (2024). Factorial invariance of the Wechsler Intelligence Scale for Children, Fifth Edition, across the UK, US and Australia and New Zealand. British Journal of Clinical Psychology, 63(2), 364-381.
