Intelligence

In short

Intelligence is the global capacity to learn, reason abstractly, adapt to new situations and solve problems effectively. Psychology studies it through theories of its structure (one general ability, several factors, or many independent intelligences) and through psychometric tests that yield an IQ. It is examined in every BA/B.Sc Foundations paper, in UGC-NET Psychology (Psychology is a UGC-NET subject, not a CSIR-NET one), UPSC Psychology optional and M.Phil Clinical Psychology entrances (NIMHANS, RCI-recognised institutes). Applied in India in clinical diagnosis and disability certification under the RPwD Act, 2016, in special education, in school guidance, and in recruitment aptitude testing (TCS NQT, AMCAT, CoCubes, eLitmus, Mettl, SHL), where reasoning tests are essentially group intelligence tests.

Explained

What intelligence is - definitions and the nature of the construct

• Binet (1905): the ability to judge well, comprehend well and reason well - he stressed judgement, not sensory speed.
• Terman: the ability to carry on abstract thinking; Stern: general adaptability to new problems; Wechsler (1944): the "global capacity of the individual to act purposefully, think rationally and deal effectively with his environment" - the most quoted exam definition.
• Operational definition: intelligence is what intelligence tests measure - circular, but it keeps the construct testable.
• Positive manifold: scores on almost all cognitive tests correlate positively; this empirical fact is the starting point of all factor theories.
• Intelligence is an inferred capacity (a hypothetical construct), not a directly observable trait, and it differs from achievement (what has been learnt) and aptitude (potential for a specific skill).

Theories of structure

• Spearman's two-factor theory (1904, 1927): every performance involves a general factor g (common to all tasks, "mental energy") plus a specific factor s unique to that task. Shown by tetrad differences / hierarchy of correlations.
• Thurstone's multiple-factor theory (1938): no single g; seven Primary Mental Abilities - Verbal comprehension (V), Word fluency (W), Number (N), Space (S), Memory (M), Perceptual speed (P), Reasoning (R). Basis of the Primary Mental Abilities Test and of most placement test sections.
• Guilford's Structure of Intellect (SOI, 1967): a three-dimensional model - 5 operations (cognition, memory, divergent production, convergent production, evaluation) × 4-6 contents (figural, symbolic, semantic, behavioural) × 6 products (units, classes, relations, systems, transformations, implications) = 120 cells in the original model, extended to 150/180 later. Introduced divergent thinking - the link to creativity.
• Vernon's hierarchical model: g at the top, then two major group factors v:ed (verbal-educational) and k:m (practical-mechanical-spatial), then minor group factors, then specifics.
• Cattell-Horn: fluid intelligence (Gf) - on-the-spot reasoning, largely non-verbal, peaks in the early twenties and declines; crystallised intelligence (Gc) - stored knowledge and skill, rises or holds till late middle age.
• Carroll's three-stratum theory / CHC (Cattell-Horn-Carroll): stratum III = g, stratum II = about 8-10 broad abilities (Gf, Gc, Gv, Gsm short-term memory, Glr long-term retrieval, Ga auditory, Gs processing speed), stratum I = narrow abilities. The theory behind modern WISC-V and Woodcock-Johnson.
• Thomson's sampling theory and Thorndike's multifactor theory (intelligence as many independent bonds; abstract, mechanical and social intelligence) are the classical alternatives to g.
• Sternberg's triarchic theory (1985): componential (analytical - metacomponents, performance components, knowledge-acquisition components), experiential (creative - dealing with novelty and automatisation), contextual (practical - adaptation, shaping, selection); later the theory of successful intelligence.
• Gardner's multiple intelligences (1983): linguistic, logical-mathematical, spatial, musical, bodily-kinaesthetic, interpersonal, intrapersonal, naturalistic (and existential as a tentative ninth). Criteria include brain damage isolating an ability, prodigies/savants, and a distinct developmental course. Criticism: the "intelligences" are talents and are not independent (they correlate).
• PASS theory (Das, Naglieri, Kirby, 1994) - developed partly by the Indian psychologist J. P. Das from Luria's neuropsychology: Planning, Attention-Arousal, Simultaneous and Successive processing; operationalised in the Cognitive Assessment System (CAS/CAS2).
• Emotional intelligence: Salovey and Mayer (1990) - perceiving, using, understanding and managing emotions (ability model); Goleman (1995) - self-awareness, self-regulation, motivation, empathy, social skill (mixed model, popular in HR training). EQ is not measured on the IQ scale.

Measurement - from Binet to current Indian tests

• History: Galton (sensory tests, individual differences), Binet-Simon scale 1905 (first practical test, for Paris school placement), 1908 revision introduced mental age; Stern 1912 proposed the IQ ratio; Terman's Stanford-Binet 1916 popularised IQ in the USA; Army Alpha (verbal) and Army Beta (non-verbal/illiterate) 1917 began group testing; Wechsler-Bellevue 1939 introduced the deviation IQ and separate verbal and performance scales.
• Wechsler family today: WAIS-IV/V (adults), WISC-V (6-16 yrs), WPPSI-IV (pre-school). WISC-V gives a Full Scale IQ plus five indices - Verbal Comprehension, Visual Spatial, Fluid Reasoning, Working Memory, Processing Speed. Mean 100, SD 15.
• Classification of tests: individual vs group; verbal vs non-verbal vs performance; speed vs power; paper-pencil vs apparatus; culture-loaded vs culture-fair.
• Culture-fair / non-verbal tests: Raven's Progressive Matrices (SPM, CPM, APM) - the purest common measure of Gf; Cattell's Culture Fair Test; Seguin Form Board; Koh's Block Design; Alexander's Pass-along.
• Indian tests and adaptations (commonly asked):
  • Bhatia's Battery of Performance Tests of Intelligence (C. M. Bhatia, 1955) - five subtests: Koh's Block Design, Alexander's Pass-along, Pattern Drawing, Immediate Memory (digit and sentence), Picture Construction; standardised on Indian boys, usable with illiterate testees.
  • MISIC - Malin's Intelligence Scale for Indian Children (Arthur J. Malin, 1969) - Indian adaptation of WISC; verbal and performance scales, gives VIQ, PIQ, FSIQ.
  • Kamat's Measuring Intelligence of Indian Children (V. V. Kamat, 1934) - Binet-type scale in Marathi and Kannada; Kulshrestha's Hindi adaptation of Stanford-Binet.
  • Vineland Social Maturity Scale (VSMS), Indian adaptation by Malin - measures social/adaptive age, used alongside IQ in disability assessment; BASIC-MR and MDPS/FACP from NIEPID (formerly NIMH, Secunderabad) for programme planning.
  • Gessell's Drawing Test, Draw-a-Man test (Goodenough-Harris; Indian adaptation by Pramila Pathak), Binet-Kamat, DAT and group tests by NCERT and state SCERTs for school guidance.
• Requirements of a good test: reliability (test-retest, split-half, alternate form, inter-scorer), validity (content, criterion - concurrent and predictive, construct), standardisation (uniform administration and scoring), norms (age, grade, percentile, standard-score norms), practicability and freedom from bias. Typical FSIQ reliability is about 0.95-0.97, and FSIQ-school achievement correlations run about 0.5-0.6.
• Ethics and regulation in India: psychological testing and reporting for clinical purposes is the domain of clinical psychologists registered under the Rehabilitation Council of India Act, 1992; test materials are restricted, informed consent is required, and the DPDP Act, 2023 governs the personal data in test records.

Distribution, extremes and classification

• IQ is scaled to a normal distribution, mean 100, SD 15 (Wechsler, Raven's standard scores, MISIC); older Stanford-Binet used SD 16, so an SD-16 score of 132 equals an SD-15 score of 130 - always state the SD.
• Conventional Wechsler bands: 130 and above extremely high/very superior; 120-129 very high; 110-119 high average; 90-109 average; 80-89 low average; 70-79 borderline; 69 and below extremely low.
• Intellectual disability (ID): three criteria - (1) significant deficits in intellectual functioning (about IQ 70 or below, i.e. roughly 2 SD below the mean, allowing for measurement error), (2) deficits in adaptive functioning (conceptual, social, practical), (3) onset during the developmental period. DSM-5/DSM-5-TR and ICD-11 grade severity by adaptive functioning, not by IQ bands; the older ICD-10 IQ bands - mild 50-69, moderate 35-49, severe 20-34, profound below 20 - are still what most Indian university papers ask for.
• Causes of ID: genetic/chromosomal (Down syndrome, fragile X, PKU), prenatal (rubella, alcohol, maternal iodine deficiency, radiation), perinatal (birth asphyxia, prematurity, kernicterus), postnatal (meningitis, head injury, lead, severe protein-energy malnutrition, extreme deprivation).
• Indian law and practice: the Rights of Persons with Disabilities Act, 2016 lists intellectual disability (including specific learning disabilities and autism spectrum disorder) among the 21 specified disabilities; benchmark disability = 40% or more; Section 34 reserves 4% of posts in government establishments for persons with benchmark disabilities, with 1% for the cluster of autism, intellectual disability, specific learning disability and mental illness, and Section 32 reserves 5% of seats in government-funded higher education. Disability percentage for ID is certified by a notified medical authority using an IQ test together with a social/adaptive measure under the Ministry of Social Justice and Empowerment's notified assessment guidelines (the guidelines have been revised more than once - check the current gazette notification before quoting a table).
• Giftedness: conventionally IQ 130+ (top ~2%); Terman's longitudinal study of 1,528 gifted children found them healthier and more successful, disproving the "genius-insanity" stereotype; Renzulli's three-ring model - above-average ability + task commitment + creativity. NEP 2020 calls for identifying and nurturing gifted students (olympiads, NTSE-type schemes, enrichment) rather than only acceleration.

Determinants - heredity and environment

• Twin and family evidence (typical reported correlations): MZ reared together ≈ 0.85; MZ reared apart ≈ 0.72; DZ together ≈ 0.60; siblings together ≈ 0.47; parent-child ≈ 0.40; adopted child with adoptive parent ≈ 0.20; unrelated reared apart ≈ 0.00.
• Heritability (h²) of IQ in adults is commonly estimated at about 0.5-0.8 in Western samples; it is a population statistic for variance within that population at that time, not the fixed proportion of any one person's intelligence, and it says nothing about between-group differences.
• Heritability rises with age and falls in deprived environments (Scarr-Rowe effect) - genes express themselves more fully when the environment is adequate.
• Environmental influences: nutrition (iron, iodine, protein), schooling quality and years, home stimulation and language exposure, parental education and income, toxins (lead), and stress. Indian programmes bearing on this: ICDS/anganwadi, PM POSHAN (mid-day meal), the National Iodine Deficiency Disorders Control Programme and universal salt iodisation, and Samagra Shiksha's foundational literacy and numeracy mission.
• Flynn effect: mean raw performance rose roughly 3 IQ points per decade through the 20th century, so old norms inflate current IQ - a strong argument for environmental influence and for periodic re-standardisation.
• Interactionist conclusion: heredity sets a range of reaction; environment decides where in that range the person falls. Gene-environment correlation and niche-picking blur the two.
• Group differences and bias: average score differences between groups reflect test content, language, schooling and socio-economic factors; a test is biased only if it predicts the criterion differently across groups. Indian testees are frequently assessed in a second language - a major source of invalidity.

Strengths and limitations of IQ testing

• Advantages: objective, quantitative and reliable; predicts school performance and job training success better than most single variables; essential for identifying disability, learning disorders and giftedness; group tests allow mass screening cheaply.
• Limitations: narrow coverage (ignores creativity, practical and emotional competence, motivation, persistence); culture, language and schooling loaded; test anxiety, rapport, fatigue and practice effects distort scores; a score is a band, not a point; labelling can produce self-fulfilling expectations; misuse in eugenics and in streaming children.
• Good practice: report IQ with its confidence interval and the test name, edition and norms used; never diagnose from a single score; always pair IQ with adaptive-behaviour data; retest with the same instrument only after the recommended interval.

Key points

• Intelligence - global capacity to think rationally, act purposefully and adapt to the environment (Wechsler).
• g factor - Spearman's general ability common to all cognitive tasks; s is task-specific.
• PMA - Thurstone's seven primary abilities: V, W, N, S, M, P, R.
• SOI - Guilford's operations × contents × products cube; source of divergent thinking.
• Gf vs Gc - fluid reasoning declines with age, crystallised knowledge is retained.
• CHC - three-stratum hierarchy (g → broad abilities → narrow abilities) behind modern test batteries.
• Triarchic theory - componential, experiential and contextual intelligence (Sternberg).
• Multiple intelligences - Gardner's eight-plus independent intelligences; criticised as talents.
• PASS - Planning, Attention, Simultaneous, Successive processing (Das); measured by CAS.
• Mental age (MA) - the age level of performance; CA is chronological age.
• Ratio IQ - MA/CA × 100; valid only for children, meaningless after mental growth ceases.
• Deviation IQ - a standard score with mean 100 and SD 15 (Wechsler) or 16 (old Stanford-Binet).
• Culture-fair test - minimises language and cultural content, e.g. Raven's Progressive Matrices, Seguin Form Board.
• Bhatia's Battery and MISIC - the standard Indian performance battery and the Indian adaptation of WISC.
• VSMS - Vineland Social Maturity Scale; gives social age and social quotient for adaptive functioning.
• Intellectual disability - low intellectual plus low adaptive functioning with onset in the developmental period; a specified disability under the RPwD Act, 2016 (benchmark 40%).
• Heritability - proportion of variance in a population attributable to genes; about 0.5-0.8 for adult IQ, not a personal fixed share.
• Flynn effect - secular rise of about 3 IQ points per decade, requiring re-standardisation of norms.

Formulas and examples

Example 1 - Ratio IQ and its breakdown. A child aged 10 years 0 months scores at the level of an average 12-year-old.

• MA = 144 months, CA = 120 months → IQ = (144/120) × 100 = 120.
• At age 20 the same person has MA 24 years → IQ = (288/240) × 100 = 120, but at 40 with MA still 24 years the ratio gives (288/480) × 100 = 60 - absurd. Hence mental growth is taken to plateau in late adolescence and adults are scored by deviation IQ.

Example 2 - Deviation IQ and percentile. On a subtest with norm mean M = 52 and SD = 8, a candidate scores X = 68.

• z = (68 − 52)/8 = +2.00.
• Deviation IQ = 100 + 15(2.00) = 130 (SD-15 scale); on an SD-16 scale it would be 100 + 16(2) = 132.
• Area below z = +2.00 is 0.9772 → 97.7th percentile, i.e. the top 2.3% - the conventional gifted cut-off.

Example 3 - Confidence band around an IQ. A WISC-type FSIQ of 112 from a test with SD = 15 and reliability r = 0.95.

• SEM = 15 × √(1 − 0.95) = 15 × √0.05 = 15 × 0.2236 = 3.35 points.
• 95% CI = 112 ± 1.96 × 3.35 = 112 ± 6.6 → 105.4 to 118.6, i.e. roughly 105-119. So 112 and 108 are not meaningfully different, and a borderline FSIQ of 71 cannot be treated as safely above the ID cut-off.

Example 4 - Heritability from twin data. In a sample, r_MZ(together) = 0.86 and r_DZ(together) = 0.59.

• h² = 2(0.86 − 0.59) = 2 × 0.27 = 0.54 - about 54% of the IQ variance in that population is genetic; the remaining 46% is shared plus non-shared environment and error. It does not mean 54% of any individual's intelligence is inherited.

Example 5 - Flynn-effect correction and social quotient. A child is tested in 2026 on a scale last standardised in 1996, and obtains IQ 104; the same child's Vineland social age is 7 years 6 months at CA 10 years.

• Norm age = 30 years → expected inflation ≈ 3 × 3 = 9 points → re-normed estimate ≈ 104 − 9 = 95 (an estimate; the drift is not uniform across tests or abilities).
• SQ = (90 months / 120 months) × 100 = 75 - adaptive functioning clearly below age level, so the near-average IQ must be interpreted together with this deficit.

Where it is used

• Clinical and rehabilitation psychology: IQ plus VSMS/adaptive assessment for intellectual disability certification, learning-disability assessment and school concessions - NIMHANS Bengaluru, NIEPID Secunderabad, AIIMS, state mental health institutes.
• Education and special education: identification of slow learners and gifted students, CWSN provisions under Samagra Shiksha and the RPwD Act, CBSE/state board exemptions and extra time, NEP 2020 gifted-nurture programmes.
• Recruitment and assessment in India: cognitive-ability sections of TCS NQT, Infosys, Wipro, Capgemini, AMCAT, CoCubes, eLitmus and bank/SSC/railway exams; vendor platforms Mettl, Wheebox, SHL and Aon; competency and assessment-centre work in HR and L&D roles.
• Defence and government selection: the SSB's officer intelligence rating uses verbal and non-verbal reasoning tests; DRDO's DIPR develops and standardises such tests for the armed forces.
• Public health and nutrition policy: cognitive outcomes as an argument for ICDS, PM POSHAN and salt iodisation; studies of lead and malnutrition on child development.
• Research and job roles: psychometrician, test developer, school counsellor, clinical psychologist (RCI-registered), I-O psychologist, UX researcher and data roles that build or validate ability tests.

Interview questions

• Q: Why was ratio IQ replaced by deviation IQ? A: Mental age stops growing in late adolescence while chronological age keeps rising, so the ratio falls artificially for adults; deviation IQ instead places a person against the mean and SD of their own age norm group.
• Q: What exactly does Spearman's g mean? A: A general factor loading on all cognitive tasks, inferred from the universal positive correlations between ability tests; each test also has a specific factor s unique to it.
• Q: Difference between fluid and crystallised intelligence? A: Fluid intelligence is novel reasoning, largely culture-free and declining from early adulthood; crystallised intelligence is accumulated knowledge and skill, which is maintained or grows into middle age.
• Q: Is Gardner's theory accepted by psychometricians? A: Only partly - it is influential in education, but the proposed intelligences correlate with each other and lack validated measures, so most psychometricians treat them as talents within a g-based hierarchy.
• Q: Which test would you use for a 13-year-old with no formal schooling, and why? A: A performance or culture-fair test such as Bhatia's Battery or Raven's Progressive Matrices, because verbal and reading-based items would measure schooling rather than ability.
• Q: Can a person with IQ 68 be denied the diagnosis of intellectual disability? A: Yes - DSM-5 and ICD-11 also require significant deficits in adaptive functioning with developmental onset; adequate adaptive functioning, or an invalid test administration, rules the diagnosis out.
• Q: What does heritability of 0.7 tell you about one child's intelligence? A: Nothing about that child - it describes the share of variance within a specific population at a specific time, and it can change if the environment changes.
• Q: What is the Flynn effect and why does it matter in practice? A: Mean test performance rose about 3 points per decade, so scores on outdated norms are inflated - which can wrongly lift someone above a disability cut-off; tests must be periodically re-standardised.
• Q: Why report a confidence interval with an IQ? A: Every score has measurement error (SEM = SD√(1 − r)); the interval shows the band in which the true score probably lies and stops over-interpretation of small differences.
• Q: How does an aptitude test differ from an intelligence test? A: An intelligence test samples general ability broadly; an aptitude test targets potential in a specific area (clerical, mechanical, numerical) and is validated against success in that area.
• Q: Is emotional intelligence measured on the IQ scale? A: No - EI is a separate set of abilities about perceiving and regulating emotion, measured by its own instruments; it correlates only weakly with IQ and its mixed models overlap with personality.
• Q: A candidate scores poorly on a timed reasoning test but performs well on the job. What do you conclude? A: The test has limited predictive validity for that role - speed, language, anxiety or practice effects may have depressed the score; selection should use multiple predictors, not one cognitive test.

Common mistakes

• Using ratio IQ (MA/CA × 100) for adults, or mixing years and months in the ratio.
• Quoting an IQ without the test, edition and SD - 132 on an SD-16 scale is the same rarity as 130 on an SD-15 scale.
• Treating IQ as a fixed, precise number instead of a band with measurement error, and diagnosing intellectual disability from IQ alone without adaptive-behaviour data.
• Reading heritability as "the percentage of my intelligence that is genetic", or using it to explain differences between groups.
• Confusing Thurstone's seven primary mental abilities with Gardner's multiple intelligences, or Guilford's divergent production with Thurstone's reasoning factor.
• Calling Raven's Progressive Matrices "culture-free" - it is culture-fair (reduced, not eliminated, cultural loading).
• Quoting the ICD-10 IQ severity bands as current DSM-5 practice, or stating RPwD disability percentages from an old notification without checking the current gazette.

Related topics

Learning · Memory · Personality theories · Motivation and emotion · Schools of psychology · Psychological testing, statistics and research methods