Individual Tests Outline 1 Aptitude ability and achievement
Individual Tests - Outline 1. Aptitude, ability, and achievement 2. Alternatives to major tests of abilities 3. Specific individual tests – infants – – Brazelton Gesell Bayley Cattell
Individual Tests - Outline 4. Specific individual tests – young children – Mc. Carthy – Kaufman 5. Specific individual tests – special populations – – Columbia Peabody Leiter Porteus
Individual Tests - Outline 6. Specific individual tests – learning disabilities – Illinois – Woodcock-Johnson
Ability, Aptitude, & Achievement • Let’s begin by reviewing the distinctions amongst these three related concepts: – Ability – refers to general capacity to do things. – Aptitude – capacity to carry out a specific function. – Achievement – what has been learned.
Alternatives to the major ability tests – Major ability tests (e. g. , Wechsler scales, Stanford -Binet) are excellent tests with known psychometric properties. – However, these tests are not appropriate for use with some of the people we might want to give an ability test. – In particular, the major tests require certain kinds of input & output (e. g. , vision, speech) that may not be available to some subjects.
Alternative tests of ability • Alternative tests tend to be created for special purposes or populations. This results in: – fewer people using any given test – a smaller resource base for investigating these tests’ psychometric properties. – less knowledge about the reliability and validity of these tests.
Alternative tests of ability • Alternative tests are often useful as a supplement rather than a replacement for major tests.
Specific individual ability tests 1. Infant Scales A. B. C. D. Brazelton Neonatal Assessment Scale Gesell Developmental Schedules Bayley Scales of Infant Development II Cattell Infant Intelligence Scale 2. Tests for Young Children A. Mc. Carthy Scales of Children’s Abilities B. Kaufman Assessment Battery for Children
Brazelton Neonatal Assessment Scale • Infants 3 days to 4 weeks old • Well-constructed (makes sense conceptually). • Widely used in research, especially with at-risk infants. • Assesses: – Attention and social responsiveness – Muscle tone and physical movement – Control of alertness – Physiological responses to stress • Psychometric issues – e. g. , no norms • good inter-rater reliability; poor test-retest reliability • validity: test does not predict later intelligence – in fact, we don’t know what it is measuring • but Majnemer & Mazer (1998) say 1995 version of test is better • Lundquist & Saber (2000) – detected sex differences among healthy neonates
Gesell Developmental Schedules • Children 21 months to 6 years • First published in 1925 • Original observations provided data on developmental norms • Norms for milestones let tester report comparative rate of development for subject • Test score based on presence/absence of agerelated behaviors • Developmental Quotient based on test score • Unrepresentative standardization sample • Poorly documented reliability & validity • Vague instructions. • Questionable scoring system • Only predicts IQ in lowest range • Shepard (1992): GDS should not be used for placement. Issues with inter-judge reliability and predictive validity
Bayley Scales of Infant Development II • Infants 1 – 42 months old • Psychometric properties: – excellent standardization • 40 years in the making – good split-half reliability • Assessments based on – weakest in youngest age normative developmental ranges data. – test score does not predict • 2 primary scores: mental later IQ and motor. – widely used in research, especially with at-risk • Assumes that later mental infants functions depend upon motor development
Cattell Infant Intelligence Scale • Based on normative • Standardized on nondevelopmental data representative sample of 274 children • Items taken form Gesell • Published in 1939, & Stanford-Binet never updated • Also some new items • Arranged in an age scale • Does not predict IQ in normal range • Ricciuti (1994) says it is widely used in research
Mc. Carthy Scales of Children’s Abilities Children 2. 5 – 8. 5 years old. Developed in early 1970 s. Carefully constructed General Cognitive Index: composed of scores on 15 of the scales (µ = 100, �= 16) • More research support for verbal and perceptualperformance factors • Less research support for quantitative, memory, and motor factors. • • • Good reliability for GCI • Concurrent validity: good correlations with WPPSI Verbal, Performance, and Full Scale IQs (Karr et al. , 1993) • Widely used in research, especially with at-risk children.
Kaufman Assessment Battery for Children • Children 3 -18 years old • 18 subtests combined into 5 global scales: – sequential processing – simultaneous processing – learning – planning – knowledge • Grounded in research – but not the most recent research. • Consists mostly of nonverbal items that require child to perform various informationprocessing tasks.
Kaufman Assessment Battery for Children • mental processing – measures fluid intelligence – Sequential processing (problems solved in a step-bystep fashion – e. g. digit span, typical math problems) – Simultaneous processing (bits of information organized and integrated to solve a problem – e. g. , understanding a paragraph; Ravens Progressive Matrices) • achievement – measures crystallized intelligence – – vocabulary reading comprehension general knowledge arithmetic knowledge
Kaufman Assessment Battery for Children • Strengths: first 3 items on mental processing subtests can be used to “teach the task. ” • internal consistency ~. 80 for subtests, ~. 90 for 5 global test scores. • Mental Processing composite positively correlated with school achievement; ~. 70 with WISC-R full-scale IQ. • some factor analysis studies support distinction between sequential and simultaneous tasks (but Strommer, 1988 doesn’t). • produces smaller average differences between African -Americans and whites • colorful and interesting items; good norms
Kaufman Assessment Battery for Children • Weaknesses: • Theoretical foundation may be crumbling – as new research makes ideas of Luria, Sperry, Neisser, out of date. • Match between tests and theoretical foundation not as clear as we would like. • Kahan & Noyman (2001) – K -ABC does not distinguish between ability and achievement (which questions a major claimed virtue)
Specific individual ability tests 3. Tests for Handicapped & Special Populations A. B. C. D. Columbia Mental Maturity Scale III Peabody Picture Vocabulary Test III Leiter International Performance Scale R Porteus Maze Test 4. Tests for Learning Disabilities A. Illinois Test of Psycholinguistic Abilities B. Woodcock-Johnson III
Columbia Mental Maturity Scale III • Evaluates ability in • No verbal response normal & variously needed handicapped children 3 - • No fine motor control 12 years old. needed • Tests general reasoning • Child indicates which ability with multipledrawing does not choice items. belong among 3 -5 drawings on a card. • 92 cards in 8 age scales
Columbia Mental Maturity Scale III • Strengths: • Scores relatively independent of reading skill. • Easy to administer and score. • Untimed (reduces pressure on subject). • Good test manual. • Good reliability • Good standardization sample (n= 2600, stratified by sex, race, region, parents occupn. )
Columbia Mental Maturity Scale III • Weaknesses: • Carvajal et al. (1993): weak correlations with • Scale is very vulnerable WPPSI-R for children 3 to error – that is, scores years 6 months to 7 can be seriously inflated years 3 months. Do not by guessing. use CMMS as substitute for WPPSI-R.
Peabody Picture Vocabulary Test III • Originally developed by Dunn & Dunn (1981). • Updated in 1997. • Intended to measure receptive vocabulary, producing a non-verbal estimate of intelligence. • Multiple choice items require no reading ability. • Subject indicates “Yes” or “No” only • Used with ages 2. 5 to 90 years.
Peabody Picture Vocabulary Test III • Two forms: IIIA and IIIB. – Each has 204 plates – Each plate presents 4 numbered pictures. – Task: specify which picture best relates to a spoken word • Items arranged in increasing order of difficulty
Peabody Picture Vocabulary Test III • Good reliabilities: . 86 to. 97 • Validity: – reasonable correlations with WISC-III VIQ: . 91 – Ukrainetz & Blomquist (2002): weak, but significant, correlation with NDW, a vocabulary measure • Tends to underestimate Wechsler or Binet IQs for children at higher and lower ends of IQ range • Only evaluates receptive vocabulary.
Leiter International Performance Scale R • Non-verbal alternative to Stanford-Binet for children 2 to 18 years old. • Strictly a performance scale. • Updated in 1997. • Can be used with deaf and language-disabled subjects. • Reasonably good validity • Useful as an aid to diagnosis in disabled children
Porteus Maze Test • Popular, non-verbal performance test. • First published around 1920. • Can be administered without verbal instructions. • Not much recent research on this test. • Issues: • Standardization sample is very old • No manual • Still used with certain groups (e. g. , BD).
Tests of Learning Disabilities • School problems may result from a variety of causes: – – low IQ emotional difficulty SES Parents’ characteristics (IQ, education) • Some children have difficulty learning in the absence of such factors.
Tests of Learning Disabilities • Learning disability is usually defined as: – A significant difference between IQ and achievement (often a 2 difference, with IQ higher). – Only in the presence of such a difference can learning disability be diagnosed. – So, thrust of most testing is to compare ability (potential) with achievement (actual)
Illinois Test of Psycholinguistic Abilities • Used with children 2 – 10 years old • Based on an information-processing model: – Failure to respond could result from any of: • defective input system • defective processing between input and output • defective output system
Illinois Test of Psycholinguistic Abilities • Separate subtests for these different “stages” – goal is to isolate the stages where problem lie
Illinois Test of Psycholinguistic Abilities • Issues: – manual gives no information about reliability or validity – manual gives little information about norming sample • normed only on middle-class children • criticized for culturally loaded content – Ottem (2002) says test is useful for languageimpaired children
Woodcock-Johnson III • Designed as a broad range test for school settings • Based on Cattell – Horn – Carroll stratified model of intelligence • Assesses – – – g specific cognitive abilities scholastic aptitude oral language achievement.
Woodcock-Johnson III • Contains both ability and achievement tests that were normed together – allowing evaluation of presence of discrepancies. • Intended to offer precise localization of impaired function, facilitating intervention. – – Cognitive abilities – 10 tests Extended CAT – 10 more tests Achievement test – 12 tests Extended Ach test – 10 more tests
Woodcock-Johnson III • Relatively good psychometric properties – standardization sample n > 8800, & sample was representative – split-half reliabilities for 38 of 42 tests – medians ~. 80 s to. 90 s – Construct validity supported by factor analysis • CAT – each test loads on a single factor • Ach Test – factorially complex: more than just g – Good correlations with other major tests
- Slides: 34