Intelligence Testing

intelligence overview

theories

  • singular model (Spearman)

    • intelligence is singular (reflects a single ability)

    • “g” for general (single) Intelligence

  • plural model (Thurstone)

    • intelligence is multiple abilities that may not relate to each other

    • there are many different kinds of intelligence that do not necessarily correlate strongly, such as:

      • verbal comprehension

      • logical reasoning

      • perceptual speed

      • numeric facility

      • word fluency

      • associative memory

      • spatial visualization

  • hierarchical model

    • blends singular and plural models

    • a global, overall intelligence factor (“g-factor”) that includes relatively ability-specific factors

  • most current intelligence tests follow a hierarchical model of intelligence

    • single IQ score

    • several ability-specific scores

    • sub-test scores represent more narrowly defined skills

uses

  • predict future achievement

  • characterize functional impairment

    • academic planning

    • cognitive changes over time

  • influence approach to assessment

    • many strategies assume average intelligence for question comprehension

  • influence approach to therapy

    • vocabulary

    • expectations for abstract reasoning

    • use of written material

Wechsler Adult Intelligence Scale (WAIS)

  • 1939

  • “the global capacity of a person to act purposefully, to think rationally, and to deal effectively with his/her environment”

test structure

  • 5 index factors

    • verbal comprehension index (VCI): verbal concept formation and verbal reasoning

    • visual-spatial processing index (VSI): spatial processing and visual-motor integration

    • fluid reasoning index (FRI): deductive reasoning and logical thinking

    • working memory index (WMI): the capacity to store, transform, and recall incoming info/data in short-term memory

    • processing speed index (PSI): ability to rapidly and accurately process simple rote info

  • two summary scores

    • full scale IQ (FSIQ)

      • based on total performance

      • the “intelligence quotient”

    • general ability index (GAI)

      • based only on VCS and PRS

      • allows for an estimate of intelligence without motor skills-related measures (using a pencil to complete tasks)

  • hierarchical model

    • FSIQ

    • five subscales

      • VCI

        • similarities: abstract verbal reasoning

        • vocabulary: degree to which one has learned, can comprehend, can express vocabulary

        • information: degree of general cultural info

        • comprehension: ability to deal with abstract social conventions, rules, expressions

      • VSI

        • block design: spatial perception, visual abstract processing, problem solving

        • visual puzzles

      • FRI

      • WMI

        • running digits: attention, concentration, mental control

          • “say the last number of each sequence 5-2-2”

        • digit sequencing: measures working memory capacity, mental manipulation, sequential processing

          • “say the numbers in order from smallest to largest "2-3-1”

        • digits forward, backward

      • PSI

        • coding: visual-motor coordination, motor and mental speed

        • symbol search: visual perception, speed

        • naming speed quantity: visual-perceptual speed

    • subtests

      • sim

      • vocab

scores

  • individual’s score is scaled relative to scores of same-age peers in normative sample

    • same IQ score has same meaning at different ages

    • “norm-referenced assessment”

  • normative sample: U.S. english-speaking population

    • national sample collected from February 2023- January 2024

    • 2,020 people in the U.S. and 688 in Canada between ages 16 and 90, stratified by education, ethnicity, region

    • stratification enhances external validity

  • reference group for WAIS-5 is same-age peers

  • intelligence assumed to be “normally distributed” throughout population

    • most individuals fall closely together towards “average”

  • average IQ for age: 100

    • average range of IQs for age: 90 to 109

  • ranges of IQ scores based on cutoffs

    • average range: 90 - 109

  • higher ranges:

    • very superior: > 130

    • superior: 121 – 129

    • high average: 110 – 120

  • lower ranges:

    • low average: 80 – 89

    • borderline impaired: 70 – 79

    • impaired: < 70

psychometrics

  • high standardization

    • same questions, administration strategy

    • coding manuals, extensive training

  • high reliability for FSIQ and subscales

    • internal consistency (alpha >.90)

    • test-retest (r from .67 to .94)

  • good validity for FSIQ and subscales

    • measures what it says it’s supposed to measure

  • high utility

challenges

  • time intensive (45 minutes for FSIQ)

  • scores can be influenced by factors that are unrelated to intelligence

    • educational background

    • cultural background

    • motivation

    • intimidation/anxiety

    • physical testing conditions

  • scale may not adequately cover the concept of intelligence

    • significant questions about control validity

  • in initial development, Wechler was aware that his scale measured only a portion of intelligence

    • other attributes contributing to intelligent behavior include:

      • planning and goal awareness

      • persistence

      • enthusiasm

      • resisting impulsivity

      • managing anxiety

benefits

  • purpose of revising WAIS-5:

    • less time intensive than previous edition

    • expanded theoretical foundations (fluid reasoning, visual-spatial processing, verbal comprehension)

    • increase in developmental appropriateness (succinct instructions, simplified language)

    • increased user friendliness (item security, portable stimuli, online administration)

    • updated psychometric properties (norms, item bias, floors and ceilings)

    • increased clinical utility (test structure, interpretive clarity)

      • broader construct coverage of working memory, fluid reasoning, and processing speed

      • lower rates and clinician and client fatigue

predict

  • what DOES predict

    • grades in school

    • number of school years completed

    • academic performance through first year of college

    • landing a job

  • what DOESN’T predict

    • emotional adjustment/happiness

    • creativity

    • interpersonal relationships/marital status

    • motivation

    • character

IQ: genetic & environmental influences

  • genetic influences (Nisbett et al., 2012)

    • heritability= .40 to .80

    • heritability lower in low-SES families

    • intelligence clearly polygenic

  • environmental influences

    • lack of micronutrients, presence of environmental toxins decrease IQ scores

    • variables associated with low SES associated with lower IQ scores

    • breast-feeding increases IQ scores

    • educational interventions with children in low-SES families increase IQ scores

overview

  1. IQ scores don’t predict everything

  2. be careful not to over interpret IQ scores; to assume that they mean more than they actually do

  3. the WAIS is not assessing all of what we mean by intelligence

  4. current best practice recommendations indicate that the practitioner should view everyone as unique and consider attributes other than intelligence when interpreting test results