Individual tests of intelligence

1
Individual tests of intelligence
2
The Wechsler scales

• have dominated the intelligence tests
• main advantages of using the WAIS
• certain referral questions will demand use of
  different tests.

3
Origins of the Wechsler tests

• Wechsler-Bellevue (1939)
• to address better testing with adults.
• from parts of the Binet-Simon Army Alpha and
  Beta
• the aggregate or global capacity of the
  individual to act purposefully, to think
  rationally, and to deal effectively with his
  environment.

4
Wechsler origins (cont)

• specifically designed for adults
• introduced performance subtests
• new formula for IQ
• actual score/ expected mean score for age group
• based on IQ constancy
• Use for psychiatric diagnosis

5
General features of the Wechsler tests

• Normative test
• Examinees score is compared to same-aged peers
• Several tests and revisions
• All have
• multiple subtests, some are common to multiple
  ages
• a verbal and a performance component
• a mean FSIQ of 100 (sd15) and a mean subtest of
  10 (sd3). Index means and SDs

6
Revisions

• Included re-norming
• Expanded age group on WAIS
• Each attempted to improve items and, thus,
  reliability and validity
• Update materials and friendly administrations
• Each brought more in depth interp (ex WAIS R -gt
  WAIS III)
• Latest version deemphasize V/P as factors
• Later versions include more instruction/samples
• Expand floor and ceiling

7
Wechsler test dates

• WB (1939)
• WAIS (1955)
• WAIS-R (1981)
• WAIS III (1997)
• WAIS IV (2009)

8
Flynn effect across revisions

• The WAIS-III yielded a 3-point lower FSIQ than
  the WAIS-R reflecting that the standardization
  sample performed slightly better than the
  standardization sample for the WAIS-R.
• Important to re-norm

9
WAIS-IV

• 15 subtests
• 10 core
• 5 supplemental
• Four index scores (VC, WM, PR, PS)
• General Ability Index (GAI)
• FSIQ

10
standardization

• 2005 US census and stratified
• 2200 16yrs - 90yrs, 11mos

11
Reliability - WAIS III

• Split-half test-retest - in the .90s for the
  IQs
• close to this for the Index scores.
• Reliabilities for individual subtests are weaker
  but all above .70. Info vocab are highest.
• The SEM for the FSIQ is 2 ½ points for most age
  groups, meaning that 95 of the time, the
  persons true IQ falls within /-5 points of
  the IQ you obtain.

12
Validity WAIS III

• Good criterion-related validity
• Verbal IQ predicts academic success as well as
  FSIQ
• Vocab - highest correlation with academic
  standing
• Construct validity evidence
• IQ correlates with educational attainment and
  occupational attainment.

13
WISC history

• first WISC - 1949
• flaws inappropriate items for kids,
  non-representation of minorities and girls in
  test content, norming sample was exclusively
  white.
• WISC-R (1974)
• WISC III (1991)
• Freedom From Distractibility Index
• WISC IV (2003)

14
Flynn Effect

• WISC-IV scores tend to be 2.5 points lower than
  WISC III
• (WISC III scores tended to be 5 points lower than
  WISC-R scores)

15
WISC-IV (2003)

• 10 core subtests and 5 supps
• supps arent used for FSIQ or composite scores
• supps can give other info or sub for a core
• Four Index scores VC, perceptual reasoning, WM,
  and PS
• Based on factor analysis
• VIQ and PIQ excluded

16
WISC IV standardization

• Standardization sample 200 kids at each year
  between 6 ½ and 16 ½.
• Based on the 2000 census, stratified
• broader range of ability than the norming group
  for the WAIS-III.

17
Reliability validity

• Reliability
• IQ and composite in .90s
• Subtests .79-.90
• Validity
• correlates highly with WISC III others wechsler
  tests
• IQ and factor scores correlate highly with
  similar tests and low with dissimilar tests
• Theory consistent group differences

18
Assessment of Preschool IQ

• At age 2 ½ , standardized measures that are
  similar to adult measures are typically used
• tapping more cognitive skills
• Also tapping the childs cooperation, attention,
  responsiveness, etc.
• Children tend to be more variable in abilities
  than adults.
• Need to be more cautious in interps

19
WPPSI history

• WPPSI (1967)
• WPPSI R (1989)
• WPPSI III (2002)

20
WPPSI-III

• Similar format to other Wechsler scales
• Shorter, more game-like
• Less emphasis on acquired knowledge
• Elimination of time bonuses
• Use of queries prompts - generally unrestricted
• Many new subtests with III

21
WPPSI-III

• Ages
• Two age bands, 26-311 years and 40-73 years
• Composite scores FSIQ, VIQ, PIQ, PSQ
• Core, supplemental, optional
• Improved reliability and validity

22
WPPSI-R

• Normed on 1700 kids (stratified)
• 2000 census
• Split half reliabilities are mid.8 to mid.9
• Good concurrent and predictive validity
• IQ score extends down to 3.9 sds below the
  population mean
• Useful in diagnosing mental retardation (with
  adaptive measure)
• IQ score extends past range needed to determine
  giftedness

23
WPPSI-III Blueprint
24
SB5(Stanford-Binet 5th edition)

• Binets definition of intelligence collection of
  faculties judgement, practical sense
  initiative, ability to adapt oneself to
  circumstances.

25

• 1st to include
• detailed admin and scoring instructions
  (standardization)
• IQ
• Initially used a ratio IQ
• Moved to a deviation IQ

26
Revisions

• Binet-Simon 1905, 1908, 1911
• Stanford-Binet 1916, 1937, 1960, 1973, 1986,
  2003

27
The SB-FE Model of intelligence (1986)

• Moved from age scale to point scale
• included multiple scores to reflect its
  hierarchical model of intelligence.
• 1st level -- g, or general ability
• 2nd level -- crystallized intelligence, fluid
  intelligence, and short-term memory
• 3rd level
• crystallized intelligence into 2 factors verbal
  reasoning and quantitative reasoning.
• fluid intelligence encompassed abstract/visual
  reasoning.

28
SB5 Model (2003)

• Based on CHC model
• Five factors of intelligence
• fluid reasoning, knowledge, quantitative
  reasoning, visual-spatial processing, working
  memory
• These load onto VIQ and PIQ
• All load onto FSIQ
• Other WAIS-like characteristics
• Subtests grouped by content
• Verbal and nonverbal component
• Means and SDs
• Subtest 10(3)
• IQs and factors 100(15)

29
(No Transcript)
30
Subtests

• 10 subtests (15 activities)
• 5 minutes per subtest
• Starting with the fourth edition, subtests are
  grouped by content (like the WAIS)
• Starting with the fourth edition, new subtests
  attempted to balance the previously verbal loaded
  tests

31
Standardization

• Nearly 5000 people, ages 2-85yrs (fourth edition
  was only up to age 24 yrs)
• Based on 2000 census and stratified on sex, race/
  ethnicity, geographic region, education.

32
Psychometric properties

• For the FSIQ, NVIQ, and VIQ, reliabilities range
  from .95 to .98 (average internal consistency
  composite reliability, across all age groups). 
  Reliabilities for the Factor Indexes range from
  .90 to .92.  For the 10 subtests, reliabilities
  range from .84 to .89. http//www.riverpu
  b.com/products/sb5/details.html
• Correlations with SB-FE, Wechsler scales,
  achievement scales, factor analytic support
• Types of validity? Problems?

33
Administration and scoring

• Routing Chronological age and performance on
  object series/matrices and vocab is used to
  determine the remainder of the testwhich
  subtests will be administered and at what level.
• Abbreviated IQ may be obtained from the routing
  subtests
• Floor, ceiling, basal levels
• new to 5th edition

34
Pros

• wide age range on one test
• routing makes administration more efficient
• less time (than WAIS) to administer
• based on item response theory
• does a good job of assessing low cognitive
  functioning
• good job of assessing giftedness
• nonverbal subtests do not require expressive
  language (as with WAIS)
• expert examination of cultural fairness issues,
  including religion
• WM factor may be useful for ADHD assessment
• can be used as a retest. 

35
Pros to routing

• Get max info in min amt of time
• Min examinee fatigue, frustration and/or
  complacency

36
General testing issues

• Teaching items
• Observation of extra-test behavior
• testing the limits
• Descriptive categories

37
Short forms

• Reducing of items will reduce reliability
• Reducing reliability will reduce validity
• Rec for screening only

38
Abbreviated SB

• Routing tests

39
WASI

• IQ screen for 6-89
• 2 subtest version (15 mins)
• 4 subtest version (30 mins)
• Adequate psychometrics
• Better than other brief measures

40
Kaufman Brief Intelligence Test, Second Edition
(K-BIT-2)

• Used mostly as a screening toolsomewhat for
  research.
• Crystallized (Verbal) Scale
• Verbal Knowledge and Riddles
• Fluid (Nonverbal) Scale
• Matrices
• Ages 4-90
• Takes less than ½ hour.

41
Kaufman Brief Intelligence Test, Second Edition
(K-BIT-2)

• Mean of 100 (sd15).
• KBIT scores were 5points higher than WAIS-R
  WISC III scores
• Good reliability and validity.

42
Group tests

• Army Alpha
• Army Beta
• Primarily used for screening purposes
• Shipley
• Compare group vs individual tests

43
Measures of specific intellectual abilities
Creativity

• Originality
• Fluency
• Flexibility
• Elaboration
• Convergent vs divergent thinking
• Guilford and examples
• Dont have great psychometric properties

44
Considerations in choosing a test

• Theory
• Purpose
• Ease of admin, scoring, and interp
• Appropriateness of norms
• Psychometrics reliability and validity
• Utility cost/benefit