Outline • • • • • • • David Wechsler – some history Wechsler’s Scales IQ scores Index scores Verbal subtests Performance subtests WAIS Psychometrics David Wechsler (1896 – 1981) • Studied at Columbia University (M.A., 1917; Ph.D., 1925) • Went to France with US Army in 1919, then to London • Studied with Pearson & Spearman • Also with Anna Freud in Vienna David Wechsler • Worked at NY’s Bellevue Hospital. • Unhappy with the Stanford-Binet – – – – – Content appropriate only for children Rapport problems if used with adults Produces only a single score Norms not appropriate for adults Binet’s emphasis on speed hurt older adults’ scores Wechsler Scales • 1939: the Wechsler-Bellevue, later called the WAIS. • 1945: the Wechsler Memory Scale • 1949: the children’s version, the WISC • 1955: Wechsler Adult Intelligence Scale • 1967: the WPPSI for children ages 2½ -7 • 1981: WAIS-R (revised form of 1955 test) • 1997: WAIS-III Wechsler Scales “Intelligence is the aggregate or global capacity of the individual to act purposefully, to think rationally and to deal effectively with his environment.” David Wechsler (1939) • global: one score ultimately describes a person’s intellectual ability • aggregate: that one score is composed of a number of sub-scores Wechsler’s original test • Wechsler believed that intellectual ability involves two major types of skills: verbal and performance abilities. • Each of these broad types includes a variety of specific skills that are assessed by the various subtests of the WAIS. • However, all these subtests measure g as well as specific skills Full Scale IQ Verbal IQ VCI Vocabulary Similarities Information Comprehension Performance IQ WMI Digit Span Arithmetic Letter-Number Sequencing POI Block Design Matrix reasoning Picture completion Picture arrangement PSI Digit-Symbol Coding Symbol Search Wechsler test produced 3 IQ scores • Full-scale IQ: • reflects both verbal and performance IQs • most reliable and valid score extracted from WAIS • why do you think this is the most reliable? Wechsler test produced 3 IQ scores • Verbal IQ: • Responses require person being tested to use language – to understand instructions or to make a response Wechsler test produced 3 IQ scores • Performance IQ: • Performance subtests involve doing something, not just answering questions Structure of the WAIS • WAIS-III has four index scores: – verbal comprehension – working memory – perceptual organization – processing speed. • Index scores added recently because, with new subtests, factor analysis suggests these four factors Index scores • Verbal comprehension • assesses general verbal skills, such as verbal fluency, ability to understand and use verbal reasoning, and verbal knowledge • based on both formal and informal educational opportunities, Index scores • Working memory • “The blackboard of the mind” (Goldman-Rakic, 1992) • Encode information into STM, store it there, retrieve it when needed • Manipulate information (e.g., addition) • Thinking, learning, planning Index scores • Perceptual organization • Use visual, spatial, and visually-guided motor skills • Organize thoughts • Assesses comfort with new, unfamiliar situations Index score • Processing speed • The speed at which cognitive processes can be carried out • focus, scanning speed, sequentially ordering visual information • sensitive to motivation, difficulty working under time pressure. • biological • cultural factors have little impact WAIS Verbal Tests • • • • • • • Vocabulary Similarities Arithmetic Digit Span Information Comprehension Letter-Number Sequencing Vocabulary • Subject is given one word at a time, asked to define it • Sensitive to word knowledge, linguistic development, acquired knowledge, verbal expression ability, crystallized intelligence • The best single measure of g • Correlation of vocabulary score with g is .83 Vocabulary • Very stable • Resistant to effects of poor concentration in schizophrenia • Not affected by mild concentration trouble • Of all WAIS subtests, the one most resistant to brain damage • Thus, allows a good estimate of pre-morbid functioning Similarities • Task is to say how two (superficially) dissimilar items might be similar • Evaluating details – which details are “diagnostic”? • Earlier items in series are known through experience (e.g., “In what way are a cake and a pudding alike?”) • Later items require abstract thinking (e.g., “How are affection and approval alike?” Similarities • Logical, abstract thinking • Concept formation • Crystallized and fluid intelligence • With Information subtest, the second best for measuring g • Correlation with g =.79 Similarities • Of VCI (Verbal Comprehension Index) subtests, the one least affected by formal education or learning • Score on this subtest is impacted by psychopathology, and by brain damage – esp. left hemisphere lesions Arithmetic • Assesses working • Correlation with g is r memory, numerical = .75 reasoning, computation skill, concentration, retrieval from LTM • Influenced by emotional state Digit Span • Repeat a series of up to 7 digits in correct order • Digits presented 1 per second • Tested both Forward (DSF) and Backward (DSB) • Attention, auditory memory and sequencing, short term memory, mental alertness, sequential processing, cognitive flexibility Digit Span • Average 6.4 digits forward, 4.7 backwards • DSF – DSB ≥ 5 suggests brain damage • Correlation with g: r=.57 (the worst) Information • Acquired knowledge, crystallized intelligence, fund of information, range of general factual knowledge, long term memory • Affected by formal education opportunities • Correlation with g: r = .79 (2nd best) • Failure on easy items followed by success on harder items suggests retrieval difficulties • Resistant to psychopathology and brain damage – good estimate of pre-morbid functioning Comprehension • 3 different kinds of questions: – Appropriate responses to hypothetical situations – Logical explanations for everyday actions – Proverb interpretations • Assesses social and moral reasoning, judgment, verbal concepts, knowledge of ordinary standards of behavior, practical information Comprehension • Rewards conventional responses, not creative ones • R Hem patients may score high (their L Hem intact so they know answers) but still behave inappropriately • g: correlation r = .77 Letter-number sequencing • Optional – not needed to compute IQ • Task is to re-order intermixed, randomlysequenced numbers and letters • Taps STM, sequential reasoning ability, planning • g: correlation r = .65 • Impaired by anxiety, weak attention focusing ability • Dropping a letter: attention problem • Mixing up letters: sequencing problem WAIS Performance Tests • • • • • • • Picture completion Digit symbol-coding Block design Matrix reasoning Picture arrangement Symbol search (optional) Object assembly (optional) Picture Completion • Task: say what detail is missing from a picture. • Timed response • Pointing response is allowed • Raises issue of pointing precision (trained examiner required) • Alertness, attention to detail • LTM (Visual) • Crystallized intelligence • R Hem skill • g: correlation r = .64 Digit Symbol Coding 1 2 3 X ╫ ÷ 9 1 6 … 9 … 3 8 Digit Symbol Coding • Visual STM • Psychomotor speed • Visual-motor coordination • Visual sequencing • Accuracy & speed • g: correlation r = .59 • Very sensitive to any kind of brain damage Block Design • Assesses visual-motor coordination, visual analysis & synthesis, spatial visualization • Visual input, motor output • Fluid intelligence • Trial & error learning • g: correlation = .72 • Affected by R Hem damage • Bizarre solutions may indicate dementia Matrix Reasoning • 4 types of stimuli: – – – – pattern completion classification analogy serial reasoning • Subject given a series of matrices and must say what goes in the empty cell Ō ō Ǒ ? Matrix reasoning • Measures abstract thinking, nonverbal reasoning, analogy skill • Fluid intelligence • Visual organization • g: correlation r = .72 Picture Arrangement • Task is to put a series of (randomly-ordered) pictures into narrative order (so they tell a coherent story) • As if you found a comic strip jumbled up and had to order the frames sensibly • Non-verbal reasoning • Possibly social skill • Fluid & crystallized intelligence • Planning • Time concepts Picture Arrangement • Failure may be due to visual problems (compare with Picture Completion) • g: correlation r = .66 Symbol Search • Optional – not needed to compute IQ • Subject shown two target abstract symbols and asked whether either target appears in a set of ‘probe’ symbols • 120 seconds allowed • How many can subject do in that time? Symbol Search • Visual-motor • g: correlation r = .70 coordination and speed, planning, STM, perceptual speed • Large practice effects in young adults Object Assembly • Subject given pieces of a puzzle • Task is to assemble pieces into a whole • Taps knowledge of partwhole relationships, visual-motor coordination, anticipation of relationships among parts • Fluid intelligence Object Assembly • Sensitive to cerebral damage, especially in R Hem • Sensitive to disorder called neglect • g: correlation r = .62 Scales and Norms for the WAIS • Determine raw score for each subtest. • Convert raw scores to standard scores, called scaled scores (M=10, SD=3) • Convery to standard scores using tables for age (13 age groups). • Subtest scaled scores are added, then converted to WAIS-III composite scores. • Three composite scores: Verbal, Performance, Full Scale, each with M=100, SD=15 Standardization of the WAIS • Standardized on a stratified sample of 2,450 adults representative of the US population aged 16-89. • There were 200 cases per age group, except for the smaller numbers in the two oldest groups. • Still difficult to know the effects of selfselection since participants had to be invited and accept to be included. Reliability of the WAIS • Internal consistency and • About .90 for test-retest reliabilities performance and these .95 or higher for full index scores: scale and verbal scores. perceptual organization, working memory, and processing speed. Reliability of the WAIS • Internal consistency reliability for the subtests range from upper .70s to low .90s. Test-retest is about .83. • Generally, performance reliabilities are lower than verbal reliabilities on the subtests. • Why might that be? Validity of the WAIS • We have a great deal of information on criterion-related and construct validity. • Factors analyses support use of 4 index scores. • Most widely used ability test today Johnson et al. (2004) • Measured correlation between WAIS scores and two other test batteries: – Comprehensive Ability Battery (CAB) – Hawaii Battery + Ravens • These batteries varied in their subtests, but correlations were essentially perfect Comprehensive Abilities Battery 1. Numerical Ability 2. Spatial Ability 3. Memory Span 4. Flexibility of Closure 5. Mechanical Ability 6. Speed of Closure 7. Perceptual Speed 8. Word Fluency 9. Inductive Reasoning 10. Associative Memory 11. Meaningful Memory 12. Verbal—Vocabulary 13. Verbal—Proverbs 14. Spelling Computations including fractions, decimal divisions, square roots, etc. Interpretation of two-dimensional figural rotation or reversal. Recall of digits presented aurally. Identification of embedded figures. Identification of mechanical principles and tools. Completion of gestalt. Evaluation of symbol pairs. Production of anagrams. Identification of pattern in sequences of letter sets. Rote memorization of meaningless pairings. Rote memorization of meaningful pairings. Multiple choice among possible synonyms. Interpretation of proverbs. Multiple-choice identification of misspellings. Hawaii Battery + Ravens 15. Card Rotations Matching of rotated alternatives to probe. 16. Mental Rotation Identification of rotated versions of 2-D pre-presentation of 3-D objects. 17. Paper Form Board Outline of cutting instructions to form the target figure. 18. Hidden Patterns Identification of probe figures in more complex patterns. 19. Cubes Identification of matched figures after rotation. 20. Paper Folding Identification of unfolded version of a folded probe. 21. Raven Identification of analogous figure to follow a sequence of figures. 22. Vocabulary Multiple choice among possible meanings. 23. Subtractn/Multn Completion of two-digit subtractions and two-digit by one-digit multiplications. 24. Word Beg/Endings Generation of words beginning and ending with specified letters. 25. Pedigrees Identification of familial relationships within a family tree. 26. Things Categories Generation of things that share assigned characteristics. 27. Different Uses Generation of novel uses for specified objects. 28. Immed Vis Mem Recall of illustrations of common objects immediately following presentation. 29. Delay Visl Mem Recall of illustrations of same common objects after delay. 30. Lines and Dots Trace a path through a grid of dots. 31. Identical Pictures Identification of alternative identical to probe. Johnson et al. (2004) • • • • Correlations: WAIS – Hawaii Battery + Ravens WAIS – Comprehensive Ability Battery Hawaii – CAB 1.00 0.99 0.99 • These results constitute strong evidence for g and show that the one-factor result does not depend upon particular tasks WISC-III • Most popular test for assessing intellectual ability of children ages 6 years, 0 months to 16 years, 11 months. • Similar to structure of the WAIS, with easier items • Both tests yield verbal, performance, and full scale IQ and 4 index scores • Most of the subtests are the same Psychometric Properties of the WISCIII • Standardization program involved 2,200 cases selected to represent the US population of children aged 6-16. • Composite scores generally have internal consistency reliabilities in the mid-.90s and testretest reliabilities around .90. • Subtest reliabilities are generally in the mid-.80s. • Object Assembly and Mazes are problematic, with reliabilities in the .60s.