Usefulness of IQ as measure of intelligence

Binet - Intelligence as judgement skills

• Created it to see who would succeed (or not) in traditional schooling
  - disagreed with Galton's prior method of measuring IQ (using eminence)

Binet and Simon (1916)

• Used verbal and social comprehension measures
  - knowing steps to complete a task
  - adapting strategy during tasks
  - monitoring performance 
• Test administered individually, untimed and increased in difficulty 
• Each test designed with particular age group in mind (3-13)

Mental Age and Ratio IQ (Stern) scored from results

• Mental Age = the highest age at which a child performed well 
• IQ = childs age/mental age x 100 (e.g age 6/mental age of 7 = 1.17 x 100 = IQ of 117)

The test was better than predecessors + revised many times
- first major revision by Terman (1916) who renamed it Stanford-Binet Intelligence Scale

Wechsler (1939) = developed IQ test based on his experience testing in the military

WAIS

• Administered individually
• Tests range of different cognitive abilities
  - e.g. speed of processing, working memory, verbal comprehension, perceptual organisation

Deary et al (2007) = General cog ability, as measured by psychometric tests ('g' and IQ), is the single most important predictor of academic success

Niesser et al (1996) = IQ scores negatively correlated with juvenile crime 

Furnham and Chamoro-Premuzic (2005) = Rarely accounts for more than 50% of variance in academic performance 

Plomin et al (2009) = Even after IQ is accounted for, meaure of self-perceived ability (SPA) predicts academic performance. SPA is genetically determined. 

Schmidt and Hunter (1998)

• Assessed predictive power of 19 diff ways of selecting people for a job
  - including: interviews, IQ tests, work sample tests, examining handwriting 
• Found the work sample tests had the greatest correlation with job success
  - however not used because expensive 
• Psychometrics were useful + practical
• Graphology was useless 

A lot of variance in work success explained by other factors 

IQ tests have excelled because of their diagnostic utility and economic/useability benefits

Research has shown in general, the higher you score on IQ tests...

• The more education you'll get
• The more money you'll earn
• The longer you'll live (Deary, 2001; Ciancolo & Sternberg, 2004)

However the relationship between IQ and success only works up to a point

• After 115 on IQ scale, the relationship breaks down (Jensen, 1980)
• Real differences between 115, 150 and 180 but the differences have far fewer personal implications
  - analogy of height in basketball
• Some people with very high IQ's don't achieve anything
• Some people excel at the type of thinking required for IQ tests and intelligent with respect to those tests
• Hudson (1966) = asked 1 convergent + 1 divergent thinker the possible uses of a brick and a blanket
  - convergent thinker (had highest IQ in their public school) had limited practical responses 
  - divergent (much lower IQ) had huge variety of responses. Also a better indicator of creativity

Stanovich (2009) = Most people recognize that IQ tests don't measure all important mental faculties. Yet, we behave as if they do. 

External Reliability (are results consistent over time?)

• Benson, (2003) = IQ scores can vary by 15 points (1SD) from one test session to the next

Internal Reliability (do subscales correlate with eachother?) + Validity (does it test what it claims to?)

• Uncertain definition of intelligence, thus uncertain subfunctions

Stanovich (2009)

3 minds 

• Algorithmic Mind = individual differences in fluid intelligence
  - what IQ tests measure
  - efficiency of goal pursuit
  - strategies and production systems
  - assessed in artificial situations with clear instructions
• Reflective Mind = indi diff's in rational thinking/ chrystallized intelligence
  - measures of critical/rational thinking
  - beliefs, goals and general knowledge 
  - assessed in typical, natural conditions where free thinking is required
  
• Autonomous Mind = few continuous indi diff's 
  - encapsulated knowledge base (modular)
  - tightly compiled learned information
  - not whats assessed in intelligence tests, more specialised 'mind-ware'

Rationality requires good reflective and algorithmic mind (for correct actions to be taken).

Fluid intelligence guides the computer-like power of the algorithmic mind, to sustain decoupling (separation). 

Override and simulation are important operations for rational thought. 

(1) Is IQ the best measure of intelligence?

(2) How useful is IQ as a predictor of success?