week 7 intelligence Flashcards
Intelligence Tests: Construct Validity
Need to know how test developer defined intelligence, for example:
Spearman’s g:
Factor analysis should yield a single large common factor that indicates the different questions or tasks largely reflected the underlying characteristic (g).
Thurston’s PMA theory
No single factor should be expected to dominate, but instead many different factors reflecting a diverse set of abilities.
CHC model
Expect hierarchical structure- basis for many modern IQ tests
Revision: CHC Hierachical Model
Stratum 111 Pervasive “g” general intelligence
Stratum 11 Broad Carroll’s list (others have added) Fluid IQ/Reasoning (Gf) Crystallised IQ/Knowledge (Gc) Domain Specific Knowledge (Gkn) Visual/Spatial (Gv) Broad Retrieval/Memory (Gr) Cognitive Processing Speed (Gs) Decision/Reaction Time (Gt) Auditory Processing (Ga)
Stratum 1 Specific Carroll’s list (others adding to it)
- approx. 70 specific abilities, e.g., general verbal knowledge (Gc narrow ability)
Summary of Stratum II (Conceptual Groupings)
Motor > Gp (psychomotor abilities
Perception > Gv, Gk, Ga, Go, Gh
Fluid Reasoning > Gf, Gsm(short-term mem)
Knowledge > Gc, Gq, Grw
Speed > Gps (psychomotor speed), Gt(speed of percept), Gs(attentional Fluency), (Glr) Learning Efficiency & retrieval fluency
Tests Informed by CHC Model
Kaufman & Kaufman (2004)
- revised KABC-II
- CHC primary organisational structure
Elliott (2007)
- revised Differential Abilities Scales (DAS-II)
- heavy CHC influence
WISC-IV (2003), WISC-V (2016), and WAIS-IV (2008)
- heavy influence by CHC theory
Woodcock-Johnson III (2001, 2007)
- heavy influence by CHC
Stanford-Binet V5 (2003)
- heavy influence by CHC
Measuring Intelligence: The Beginning
Binet’s revolution
Binet-Simon Intelligence Scale- first practical intelligence scale
Scored by identifying what 75% of an age group could answer.
30 individual tasks
Easy- name various body parts
More difficult- how are a rabbit and a bird different?
Construct a sentence from words
Remember digits
Draw a design from memory (below)
Evolution of the Binet Scale
The Stanford-Binet Intelligence Scales
The first published intelligence test with clear instructions on use
The first test to introduce the concept of an alternate item
Ratio IQ: the ratio of the test-taker’s mental age divided by his or her chronological age, multiplied by 100 to eliminate decimals:
RATIO IQ = Mental age/chronological age x 100
What is each child’s ratio IQ?
Fred has a mental age of 6 years, and a chronological age of 6 years.
MA/CA X 100 = 6/6 X 100 = 100
Charlie has a mental age of 10 years and a chronological age of 8 years.
MA/CA X 100 = 10/8 X 100 = 125
Charlie stays two years ahead, what is his ratio IQ when he is 10 years of age?
MA/CA X 100 = 12/10 X 100 = 120
Standford-Binet-III
Third edition changes:
Deviation IQ replaced the ratio IQ
Deviation IQ: a comparison of the performance of the individual with the performance of others in the same age in the standardisation sample
Fourth Edition:
Point scale implemented
Subtests organised by category of item rather than the age
Stanford-Binet-5
The fifth edition was designed for administration from ages 2 to 85
Upward extension to test adults over time
The test yields a number of composite scores, including a Full Scale IQ, an Abbreviated Battery IQ score, a Verbal IQ score, and a Nonverbal IQ score
Draws from CHC model
CHC and Corresponding SB5 Factors
see slide 12 for picture
Weschler Tests
Wechsler dissatisfied with Stanford-Binet (designed for children originally)
devised Wechsler Adult Intelligence Scale (WAIS)
age range 16+ yrs
Downward extensions of WAIS due to popularity to allow assessment of children
Wechsler Intelligence Scale for Children (WISC)~ age range ~ 7-16 yrs
Wechsler Pre-school/Primary Scale Intelligence (WPPSI) ~ age range 3-7 yrs
Earlier versions very similar
Wechsler Tests
are what?
and work under what premise?
Individually administered
(a) assess different aspects of ability (b) give opportunity to view personality, motivation, and other variables
Work under the premise that:
Intelligence is both global and specific
Aspects of intelligence cluster into higher-order
cognitive ability domains
Subtests are different measures of intelligence, not
measures of different kinds of intelligence
Until recently, all Wechsler scales yielded a Full Scale IQ (a measure of general intelligence), a Verbal IQ, and a Performance IQ
-Recent revisions greater informed by CHC and factor analysis
Sample Similar Tasks (Sattler, 2016)
Block design: reproduce designs using 2, 4 or 9 blocks
Similarities: “In what way are a pencil and a piece of chalk alike?”
Matrix Reasoning: complete the series
Digit Span: repeat digits forwards, backwards, and in order
Coding: Quickly copy symbols from a code in a set time
Sample Similar Tasks (Sattler, 2016) part 2
Figure weights: Select choice that balances scale
Visual Puzzles: Select three pieces that make the design
Picture Span: Look at one+ nameable objects then find in a set with other pictures
Symbol Search: Quickly find and indicate if a symbol is present in an array
Vocabulary: e.g., “What is a ball?”
WAIS Evolution
WAIS stands for what?
1939 – original Wechsler-Bellevue Intelligence Scale
1955 – Wechsler Adult Intelligence Scale (WAIS)
1981 – WAIS-R
1997 – WAIS-III (3rd Ed)
. to this point, almost 80% original
. improvements in presentation and focus on cultural fairness
2008 – WAIS-IV
substantial changes (reflect CHC developments)
~ new subtests, some subtests dropped
~ dropped PIQ and VIQ composites
~ normed with WMS (What is expected memory functioning given IQ?)
WAIS-III (1997): Structure
organisation identified via factor analyses (a) “g” factor (b) verbal and performance factors (c) four indices (factors) . verbal comprehension . working memory . perceptual organisation . processing speed (d) 14 subtests (2-3 per factor/index)
WAIS-IV (2008)
see slide 21
CHC Theory & WAIS
Stratum 111 Pervasive “g” general intelligence
Stratum 11 Broad Fluid IQ/Reasoning (Gf) Crystallised IQ/Knowledge (Gc) Domain Specific Knowledge (Gkn) Visual/Spatial (Gv) Short-term Memory (Gsm) Cognitive Processing Speed (Gs) Decision/Reaction Time (Gt) Auditory Processing (Ga)
Stratum 1 Specific - some specific abilities
Weschler Intelligence Scale for Children (WISC)
Many revisions over the years WISC 1949 WISC-R 1974 WISC-III 1991 WISC-IV 2003 WISC-V 2016
WISC-V designed for assessment of children aged 6:0 -16:11
WISC-V Overview
10 primary subtests
Full Scale IQ (FSIQ) and the five primary index scores
Perceptual reasoning now split into visual spatial and fluid reasoning
6 secondary subtests
Ancillary and expanded index scores
Substitute for a primary subtest in calculation of FSIQ if admin is spoiled
(5 complementary subtests- US/Q-interactive only)
Only on Q-interactive (iPad) administration
US norms
WISC V Subtests and CHC
see slide 25
WPPSI
The Wechsler Preschool and Primary Scale of Intelligence was developed to assess young children
WPPSI 1967 WPPSI-R 1989 WPPSI-III 2003 WPPSI-IV 2012 Two batteries: - Ages 2:6-3:11 - Ages 4:0-7:3
WPPSI-IV & CHC
see slide 27
Infancy and Early Childhood
Conceptualisation of intelligence in infants & young children is different
Includes motor and sensory abilities
Assessment of infants/young children requires unique skills
Shorter attention
Limited response options
Less stable than later assessments, useful in differential diagnosis, early identification of needs, informing intervention/goals
Mullen Scales of Early Learning (Mullen, 1995)
Individual assessment suitable from birth to 68 months of age
Theoretical basis: information processing and neurodevelopmental theories
Five subscales
- Subscale scores (T-scores, Mean = 50, SD =10)
- Cognitive: Visual reception, fine motor, receptive language, and expressive language
- Combine for an Early Learning Composite (“g”)
- Gross Motor
Administration
Individual vs. Group
Common features of individual adaptive tests:
-Routing Test & teaching items
-Basals & ceilings (https://www.pluralpublishing.com/media/media_acdc_SamplePages.pdf)
-Queries/prompts
Behavioural observations: more to testing that the numbers!
Individual vs. Group Testing
Emerged in response to need for assessments for US military Army Alpha Test Army Beta Test More recent tests, e.g., Raven’s matrices Economic/efficiency benefits More widely used than individual tests
pros & cons of traditional group testing
see slide 32
Adaptive Testing:Routing Test & Teaching Items
Adaptive testing: testing tailored to test-taker
Routing test/item: task to direct examinee to particular questions
-Direct to items at an optimal level of difficulty
Often contain teaching or practice items
Illustrate response required to ensure understanding of task requirements
Qualifying items
Check sufficient skills for subtest
E.g., WISC sequencing component of digit span (counting to 3), and for letter-number sequencing (counting to 3 and alphabet to C)
If don’t qualify treat subtest as missing/invalid
Basal
Adaptive tests have rules for starting/ reversing/finishing subtests
-Scaled from easiest (floor) to hardest (ceiling)
Starting point: Floor (for individual)
-E.g., answers two consecutive items correctly
Reversing
-E.g., test-taker does not obtain perfect score on either first or second item, then reversal items administered until perfect scores obtained on two consecutive items (floor)
Basal: base level criteria in a subtest for test-taker after start/reverse rules applied
Ceiling
Ceiling: highest item completed for test-taker based on discontinue rules
Discontinue rules tell you when to stop administration for a particular subtest, may be:
- Number of items in a row incorrect (e.g., similarities, vocabulary, matrix reasoning tasks)
- Multiple trials of a task incorrect (e.g., number of digits recalled- both trials of X items incorrect)
- Time lapse (e.g., coding, symbol search, cancellation tasks)
Benefits of Adaptive Testing
Why are starting rules a good thing?
- Sets a reasonable difficulty level- no lull into “false sense of security” or frustration
- Facilitates rapport
Why are discontinue rules a good thing?
- Reduce test-taking time
- Reduce test-taker frustration
- Maximum information in minimum time
Queries/Prompts
Used to clarify task requirements or to clarify ambiguous answers
-(Q): say “What do you mean?” or “Tell me more about it
Behavioural Observations
Assessment starts the moment the client comes to the clinic
What does to client look like and act like
How do they present during the interview and/or assessment?
Is the client providing the best assessment of their current abilities?
- Test-taking effort e.g. cooperativeness, motivation, interest, response to failure
- Physiological factors – pain, sensory problems
- Psychological factors – depression, anxiety
Behavioural Observation Sample Record Form (Paynter & Fothergill, 2015)
see slide 39
Scoring: Key terms
Scaled scores (M = 10, SD = 3): age standardised score on subtests (e.g., WISC)
Composite scores (M = 100, SD = 15): age standardised score on composites and FSIQ
Confidence interval: margin of error for true score (conventionally use 95%)
Percentile ranks equal to or better than this percentage of children’s scores in the normative sample
Descriptor – verbal description of performance (e.g. Average, low average)
Why Transform Raw to Standard Scores?
Max Range of Raw Scores Vary
Scaled scores transform ranges for easier profiling
and comparison across subtests
(see slide 41)
Weschler Test Scores & Descriptors
see 42
Descriptive labels in SB-5
Measured IQ Range Category 145 – 160 Very gifted or highly advanced 130 – 144 Gifted or very advanced 120 – 129 Superior 110 – 119 High average 90 – 109 Average 80 – 89 Low average 70 – 79 Borderline impaired or delayed 55 – 69 Mildly impaired or delayed 40 – 54 Moderately impaired or delayed
Note similar to Weschler test categories
Successive-level approach(e.g., Sattler et al. 2016)
Interpretation of intelligence tests based on successive level approach
- Global composite (e.g., Full Scale IQ, GAI)
- Index scores (e.g., VCI, FR, VS, WM, PS)
- Subtest level (e.g., vocabulary)
- Item level (e.g., error patterns in subtests)
- Qualitative analysis (e.g., behavioural observations of problem-solving strategies)
WAIS-IV example: Interpretation of Index and FS Scores
VCI = Verbal Comprehension Index PRI = Perceptual Reasoning Index WMI = Working Memory Index PSI = Processing Speed Index FSIQ = Full Scale IQ GAI = Global Ability Index (FSIQ without effects of memory)
(X = 100, SD = 15)
Interpretation of Subtest Scores
(X = 10, SD = 3)
See slide 46
Issues in IQ Testing
No construct in psychological measurement generates as much debate and controversy as that of intelligence.
Concerns are rooted in both misconceptions and genuine concerns about how intelligence scores have been used historically, and are used today.
One such misconception is that the intelligence quotient, or IQ is a measure of a person’s worth.
Intelligence testing is still widely used despite concerns that it discriminates against minority ethnic groups, women, disabled people and those with reduced access to education and civic benefits.
Culture
Items on an intelligence test tend to reflect the culture of the society where the test is employed and thus many theorists have expressed a desire to develop a culture-free intelligence test.
Culture-free intelligence tests are difficult if not impossible, and thus “culture-fair” intelligence tests began to be developed.
Culture loading: the extent to which a test incorporates the vocabulary, concepts, traditions, knowledge, and feelings associated with a particular culture
Australian Adaptation of WAIS-R: Information Subtest
USA Questions Australian Questions
What are the colors What are the colours in the American flag? in the Australian flag? Name four men who have Name four men who have been president of the US? Prime Minister of Australia? Who was Louis Armstrong? Who was Donald Bradman? In what direction would you In what direction would you travel if you went from travel if you went from Chicago to Panama? Adelaide to Darwin? What is the population of the What is the population of United States? Australia?
Stability of Intelligence
Intelligence in adulthood appears to be relatively stable.
Gold et al. (1995) compared archival intelligence test data from World War II to intelligence test data acquired 40 years later from a sample of 326 veterans and found stability in measured intelligence over time.
Young adult intelligence was found to be the most important determinant of cognitive performance as an older adult.
Ivnik et al. (1995) found that verbal intellectual skills tended to be highly stable over time, but newly learned information was much less stable.
In later adulthood (beyond age 75 especially), a decline in cognitive ability has been noted.
Stable…BUT…Flynn Effect
Flynn effect: the progressive rise in intelligence test scores in the population that is expected to occur on a normed intelligence test from the date when the test was first normed
James Flynn: https://www.youtube.com/watch?v=9vpqilhW9uI&feature=youtu.be
Example of FE: Change in Scores: WAIS-R (1981) to WAIS-III (1997)
Study (Wechsler, 1997)
- 192 adults; age 16-71 years completed both tests between 2-12 weeks apart - Higher scores on older norms- shows Flynn effect
___
WAIS-R (1981) WAIS-III (1987) Diff
Scale M (SD) M (SD)
VIQ 103.4 (14.5) 102.2 (15.1) -1.2
PIQ 108.3 (14.4) 103.5 (15.4) -4.8
FSIQ 105.8 (14.3) 102.9 (15.2) -2.9
- Norms date by 1/3 to 1/2 IQ points per year
- Are people getting smarter?
The Flynn Effect
In the past 60 years, intelligence scores have risen steadily by an average of 27 points.
Intelligence: What we do know?
Heredity sets limits on a child’s potential, environment permits potential to be actualised
Environments can restrict or promote IQ
Genetic influences increases with age
-20% infancy, 40% childhood, 60% adult, 80% older
Cannot measure innate potential – only actual behavior
Fairly stable beyond childhood within an individual, but changes over time at population level (Flynn effect)
Intelligence: What we do know?
GENDER,CULTURE,IMPORTANCE
Gender differences almost non-existent
- Females language acquisition, vocabulary
- Males spatial perception, mental rotation, more variability and overrepresented at extremes
Culture needs to be acknowledged in designing tests and interpreting scores
Big business and concerning history of use and abuse (e.g., eugenics)
–Important area for psychology and ethics
- Which of the following is not a WAIS-IV index?
a) Verbal comprehension
b) Quantitative reasoning
c) Perceptual reasoning
d) Processing speed
answer = : (b)Quantitative reasoning, is NOT a wais-iv index
THE ACTUAL 4 are:
Verbal Comprehension Index (VCI)
Perceptual Reasoning Index (PRI)
Working Memory Index (WMI)
Processing Speed Index (PSI)
Jane is 8 years of age, but has a mental age of 6 years, what is her ratio intelligence quotient?
a)80
B)86
c)133.33
d)75
answer = d) 75
ratio iq = mental age/chronological age *100 = 6/8 = .75 *100 =75