Measuring intelligence

• how each person is unique, compared to other people
  
  
• approach is correlational, what "goes together" statistically
  
  
  
  

Validity

• Does the test measure what it's supposed to?
  
  
  
  
• Are inferences based on the test scores correct?
  
  
  
  

Types of validity

• content validity: content of the test represents the content of the domain "intelligent behavior"
  
  
  
  
  
  
• predictive validity: test scores predict other, later, behavior of interest
  
  
  
  
  
  
• concurrent validity: test scores are related to other current behavior of interest
  
  
  
  
  
  
• construct validity: the whole case supporting the validity of a test, a nomological network of different kinds of evidence supporting validity
  
  
  
  
  
  
• "face" validity: intuitive judgment that the measure seems reasonable

Reliability

• are the test scores stable, repeatable?
  
  
  
  
  
  
• test-retest reliability: are test scores stable over a reasonable time period?
  
  
  
  
• split-half and internal consistency reliability: are scores stable over different versions of test?
  
  
  
  
• are test scores the same irrespective of who the tester is?
  
  
  
  

Some notes on reliability and validity

• you can have reliable measures that aren't valid
  
  
  
  
• without good reliability you can't assess validity adequately
  
  
  
  
• face validity is inadequate as a basis for conclusions about the validity of a measure
  
  
  
  

Standardization

• procedures, equipment, instructions and conditions are standardized
  
  
  
  
• test development includes deriving group norms against which to compare individual scores
  
  
  
  
• individual scores are compared against the normal distribution of scores for their age range from the standardization sample
  
  
  
  
• a variety of standardized scores can be computed, for example, percentiles, stanines.
  
  
  
  

History

• early approaches based on measuring skull
  
  
  
  
• Galton's approach based on reaction time, other sensory functions
  
  
  
  

Binet (1904)

• test for French schoolchildren needing special education.
  
  
  
  
• concept of Mental Age (MA).
  
  
  
  
• Intelligence regarded as a measure of current mental functioning, fluid
  
  
  
  

Terman (1916)

• The Stanford-Binet test, developed from Binet
  
  
  
  
• concept of Intelligence Quotient: I.Q.= MA/CA x 100 (Stern's ratio IQ)
  
  
  
  
• perceived intelligence as a fixed and inherited entity

World War 1

• Group I.Q. tests developed, Army Alpha and Beta.
  
  
  
  
  
  
  
  

Wechsler (1939)

• WISC & WAIS, commonly used today
  
  
  
  
• concept of Deviation I.Q., compare people to others of same age
  
  
  
  
  
  
• developed performance subtests in addition to verbal subtests
  
  
  
  
• gave profiles of strengths and weaknesses
  
  
  
  

Defining intelligence

• intelligence is an inferred internal psychological construct
  
  
  
  
• there are many definitions, with two general themes
  
  
  
  
  • ability to learn from experience
    
    
    
    
  • ability to adapt to one's environment
    
    
    
    
• Psychometric approach: factors
  
  
  
  
  • Spearman's "g": there is a general factor in intelligence
    this is inferred from intercorrelations among subtests
    
    
    
    
  • Thurstone's primary mental abilities: there is a set of specific factors comprising intelligence
    this is inferred from the relative independence of different abilities (also refer to savants, and the performance of people with brain damage)
    
    
    
    
• Sternberg's Triarchic Theory:
  
  
  
  
  • analytic
    
    
    
    
  • creative
    
    
    
    
  • practical
    
    
    
    
• Gardner's multiple intelligences:
  • linguistic
    
    
    
    
  • logical-mathematical
    
    
  • spatial
    
    
    
    
  • musical
    
    
    
    
  • body-kinesthetic
    
    
    
    
  • interpersonal
    
    
    
    
  • intrapersonal
    
    
    
    
• process approaches to intelligence: focus on
  
  
  
  
  • speed of mental activities
    
    
    
    
  • strategies of information processing (metacognition)
    
    
    
    
• fluid and crystallized intelligence
  
  
  
  
  
  • fluid intelligence is based on problem solving
    
    
    
    
  • with age fluid intelligence declines
    
    
    
    
  • crystallized intelligence is based on acquired knowledge
    
    
    
    
  • crystallized intelligence can compensate for decline in fluid intelligence with age
    
    
    
    

Heredity or environment?

• higher correlations with genetic closeness, irrespective of environment
  
  
  
  
• changes in IQ with different, impoverished or enriched environments
  
  
  
  
• genetics sets potential, environment nurtures or stunts its development
  
  
  
  
• group differences in IQ, controversies
  
  
  
  
• changes over the lifespan
  
  
  
  

Uses and abuses of intelligence testing

• pervasive use, but not always appropriate, e.g. job specific skills
  
  
  
  
• labeling and the issue of self-fulfilling prophecies (ref. Rosenthal study)
  
  
  
  
• cultural bias in tests (which is often also mirrored in criteria, e.g.. job performance)
  
  
  
  
• pretty good correlation between IQ and academic success in high school but not at higher levels (restriction of range, other factors also very important)
  
  
  
  
• High or low IQ guarantees nothing - neither success nor failure, nor level of life satisfaction!