Week 5: Intelligence Flashcards
1
Q
Alfred Binet
A
- Hired by French Govewrnment to devise test to identify children with special education needs
2
Q
- Definition
A
o To find and maintain a definite direction or purpose
o To make necessary adaptions to achieve the direction or purpose
o To engage in self-criticism so that adjustments in strategies can be made
- Hence, he aimed to test reasoning, judgement and attentions
3
Q
Criteria for selecting an item
A
- Item has to relate to ‘common sense’
- Item has to be part of daily life
- Item must separate dull from bright children
- Item must be practical and easy to administer
- In developing his tests he used trial and error and hypothesis testing
4
Q
1905-First Binet-Simon test of intelligence
A
- Looked for tasks that could be achieved by 66.67% - 75% of children of a particular age
- Contained 30 items presenting in ascending difficulty level
5
Q
3 levels of intellectual disability: Idiot
A
o Most severe form of intellectual disability
o Item 6 was upper limit in this range for adults – to follow simple directions and imitate gestures
6
Q
3 levels of intellectual disability: Imbecile
A
o Moderate form
o Item 16 upper limit – stating difference between 2 common objects
7
Q
3 levels of intellectual disability: Moron
A
o Mildest form
8
Q
1905-First Binet-Simon test of intelligence: Problems
A
- Little evidence of validity
- Normed with only 50 children based on average school performance
- Idiot, moron etc hardly adequate classification system
9
Q
1908 Binet-Simon
A
- 1908 Binet tested 203 school children
- dropped simplest items
- added more difficult so now had 58
- it was an age scale rather than increasing difficulty
10
Q
1908 Binet-Simon: Calculated Mental Age
A
- Pass 5 tests at 5 year old level so basal mental age = 5
- Pass 2 tests at 6 year old level so mental age= basal mental age (5) + 2/5 = 5.4 years
- Passes 0 tests at 5 year level
- Passes 5 tests at 3 year level
- Passes 1 test at year 4 level
- Mental age= 3.2 years
11
Q
Spearman’s 2 Factor Theory
A
- correlation and factor analysis of different intellectual tasks
- all tasks intercorrelated > underlying g or general intelligence
- some groups of tasks intercorrelate more strongly > underlying s or specific ability- binet’s test involved different tasks but…
o specific factors averaged out
o account for a small proportion of measured intelligence
o binet’s test predominantly a measure of g. - later test developers maximised the amount of g in their tests
12
Q
Spearman g factor
A
General ability for complex mental work
13
Q
Spearman s factor
A
Series of specific abilities such as math and verbal
14
Q
Louis Terman – 1916 iteration
A
- 1916 revised Binet and called it the Stanford-Binet
- Revised items that didn’t perform as expected
- Normed on many more people but all white children native to California
- Added adult items
- Introduced use of the Intelligence Quotient
15
Q
Stern’s Terman’s IQ
A
Case 1 (C.A. 5 years)
- Mental age 5 and 2/5 (5.4)
- Chronological age 5
- I.Q. = 5.4/5X 100
- I.Q. = 108
- Allows direct comparison between children of different ages
- I.Q. = Mental Age/Chronological Age x 100
16
Q
Problems with Stern’s I.Q
A
- Iq ideas works well for children, as abilities are developing
- Development slows in adolescence, stabilises in adulthood
- If MA was less than CA, IQ would be under 100
- 1916 test had a upper mental age of 19.5 if you passed every group of items so anyone older would have a lower IQ
(e. g. Mental age 16/Chronological Age 22. I.Q. = 16/22 X 100. I.Q. = 72.73 - So had to set an upper CA. As thought mental age didn’t improve after 16 years old, it was used as a norm, a maximum CA
17
Q
Revised Version Stern’s IQ 1937 used by Terman
A
- Adding tasks possible mental age raised to 22 yrs, 10 mths
- Better standardisation of administration
- Increased inter-scorer reliability
- More performance items but still only 25%
- Standardised with 3184 people from 11 states but still urban whites
- Equivalency forms (L&M) included so could look more closely at psychometrics
18
Q
Problems with 1937 scale Terman
A
- Correlation coefficients higher for older than younger people so not as stable for younger
- More reliable for low IQ than normal or higher
- Different age groups had different variability (SD)
- So IQ in one age range could not be compared to another age range
19
Q
1960 Stanford-Binet
A
- Deviation IQ concept solved variability issue
- New IQ tables were constructed with a mean of 100 representing the 50th percentile and the deviation IQ resulted from evaluating variability so MA at each CA
- So could compare IQs of different ages
- But no new normative samples were used
- 1972 edition used 2100 children including non-whites
20
Q
So 1960 Stanford-Binet
A
- Stanford-Binet 4th and 5th editions were hierarchical with g reflecting common variability of all tasks
- Next level had 3 groupings
o Crystalised abilities reflecting learned knowledge
• Verbal reasoning
• Non-verbal reasoning
o Fluid-analytic abilities to represent original potential/capabilities to acquire crystalised knowledge
o Short-term memory – how much info can you store briefly after only seeing something quickly
21
Q
Thurstone 1938
A
- Argued against Spearman’s ideas and said that IQ comprised independent factors “primary mental abilities”
- 1986 version saw 15 categories so all verbal together rather than using the age format
- 2003: there are 5 factors which comprise equally weighted verbal and non-verbal items
- uses a hybrid of the age and multi-construct approaches
- starts with identifying basal level and moves to find ceiling
22
Q
2003
A
- SD changed from 16 to 15
- New subsets added
- Retained new network, toys and better items
- 2-85 year age range
- normed from 4800 people stratified by gender, ethnicity, region and education + 3000 people from sub-populations (e.g. speech, language and hearing problems)
- possible scores from 40-160
- reliability for full-scale is .97-.98 for all 23 age ranges (see chapter for other reliability coefficients)
23
Q
Cattell – Fluid and Crystalised intelligence
A
- Cattell (1965), Horn (1985) and others argued for a quite different structure in intelligence
- Fluid intelligence (Gf) – logical thinking and problem solving in novel situations – relatively independent of cultural experience such as schooling
- Crystalised intelligence (Gc) skills and knowledge – heavily depending on schooling and life experience
- WAIS measures fluid intelligence on the performance scale and crystalised intelligence on the verbal scale. The overall IQ score is based on a combination of these two scales
24
Q
According to Wechsler, 1939, Intelligence is
A
- “the aggregate or global capacity of the individual to act purposefully, to think rationally and to deal effectively with his environment”
- he believed that factors beyond intellectual ability impacted on intelligent behaviour
25
Currently
- There are 3 Wechsler intelligence tests:
o The Wechsler Adult Inteligence Scale (WAIS-IV) 16 years and over
o Based on the WAIS, he developed the Wechsler Intelligence scale for children (WISC-V) 7-16 years
o Weschler preschool and primary school Scale of Intelligence (WPPSI-IV) 3-6 yrs
26
David Wechsler (WAIS)
- He objected the since score offered by the 1937 revision of the Stanford-Binet
- As Binet scales developed for children he also challenged that they were valid for adults
- Said Binet test’s emphasis on speed unfairly disadvantaged older test takers
- Mental age norms clearly didn’t apply to adults
- So, 1939 he developed a test specifically at adults
27
WAIS
- Tests deviated in 2 main ways from Binet’s
o Used a point scale concept rather than an age scale
o Included a performance scale
28
Point scale concept
- In age scale format items are not arranged by content but rather, in group of mixed tasks capable of being passed by 2/3 to ¾ of people of a particular age
- In binet’s scales it took ¾ of tasks to pass, you go no credit for passing ½
- In point scale, you get credit for everything passed
- Items of a particular content were grouped together
- Today this concept is the standard
29
The performance scale concept
- Binet’s earlier versions criticised for emphasis on verbal skills and language
- So Wechsler included entire non-verbal scale (performance scale)
- Tasks required people to actually do something 8
30
The performance scale concept
- Early Binet scale tasks aimed at younger children
- Also difficult to separate tasks as grouped together
- So WAIS etc had 2 major scales from the start
- Performance scales aim to overcome bias (cultural, language, education)
- Performance tasks also take more sustained effort, concentration and attention
31
First Efforts
- First try was poorly standardised; used 1081 whites primarily from NY
- By 1955 rectified then standardised again in 1981, 1997 (WAIS-III)
- Currently WAIS-IV 2008
32
WAIS-IV 1
- Until this version the WAIS tests had 14 subsets measuring verbal and performance IQ
- Test now has 10 core subsets that provide a full scale IQ score
- And 5 optional subsets
- Now provides index scores
- The General Ability Index (GAI) is based on the 6 subsets of the VCI and PRI)
33
WAIS-IV 2
- Standardised of 2200 adults in 13 age groups from 16-90
- Stratified by age, gender, education level, ethnicity, and region
- Revised, based on 4 years of national and clinical research
- Strengthened theoretical foundations
- Reduced administration time
- Enhance psychometric properties
- Improved developmental appropriateness
- Longitudinal research suggests this decline is not nearly as rapid as these norms indicate from the previous version
- Decline as a function of reaction times
34
Uses in diagnosis WAIS-IV
- Can provide clues for diagnosis and differentiate between groups
o E.g., Theiling and Peterman (2014) tested 116 people with ADHD and 116 matched controls; adults with ADHD showed decrements in subtests with working memory and processing speed demands and a higher GAI in comparison with the FSIQ
o Holdnack et al. (2011) found people with autism had deficits in social perception, verbal comprehension, and processing speed compared to controls and inconsistent performance on auditory working memory and perceptual reasoning tasks
35
WISC-V – 2014
- 1949, 1974, 1991, 2003, now 2014
- 6-16 years and 11 months (there is an overlap in ages with all 3 scales)
- Australian version
- Measures cognitive abilities, processing speed and working memory
- Contains 15 subsets
- Standardised on 2200 children
- Scoring etc parallels the WAIS
- Reliable using split halves
36
WPPSI-IV
- 2 years and 6 months-6months and 11 months
- also aimed at identifying children with learning difficulties
- there are practice items (as with the WISC-V)
- examiners can repeat items
37
So why use the WAIS as opposed to the Stanford-Binet
- Stanford-Binet still regarded as best for younger children and children of below average intelligence as is more user friendly
- WAIS etc better in adult population and older children/adolescents as
o More neuro
o Able to more precisely pinpoint deficits
o Training
