W11L1 Lecture Notes on Theory Development of ICQ PDF

# 10/14 Wil Lecture I ## Theory Development of ICQ ### 1) Single factor "g" (Charles Spearman, 1927) - Positive manifold. - Good predictor of performance in real life ### 2) Primary mental abilities (Thurstone, 1938) - 7 types of mental ability - Verbal comprehension - Inductive reasonin...

# 10/14 Wil Lecture I ## Theory Development of ICQ ### 1) Single factor "g" (Charles Spearman, 1927) - Positive manifold. - Good predictor of performance in real life ### 2) Primary mental abilities (Thurstone, 1938) - 7 types of mental ability - Verbal comprehension - Inductive reasoning - Numerical fluency - Word fluency - Spatial ability - Memory - Perceptual speed - Levels differ among individuals - Can be impaired in isolation and brain damage ### Gf-Gc theory (Cattell 1941) - Current dominant theory - **Gf**: General _fluid_ intelligence - Ability to grasp things and deal with novelty - Non verbal abilities: inductive/ deductive reasoning - **Gc**: General _crystallized_ intelligence - Acquired knowledge and skills - Requires exposure to culture, formal/ informal education - Performance on a single task is likely to involve both ### Although different, Gf and Gc correlate to various events - Ability increases with age, then peaks and decreases - The strongest evidence is that Gf and Gc are different ### Hierarchy of theories - General intelligence "g" - Gf - Inductive reasoning, spatial ability, etc - Gc - Verbal comprehension, arithmetic, etc - Fluency ## How to Assess Tests 1. **Validity**: How well the test measures what it's supposed to measure/ used appropriately 2. **Reliability**: How accurate it is ### Reliability - The consistency of the measure - Should be able to differentiate between two different levels of ability ### Classical Test Theory - Two components in a test - True score (T) - Error component (variance) (E) - Sources of Error - Test construction or administration - Errors in scoring - Interpretation subjectivity - True score estimated off ideal long-term mean of true scores ### How we estimate reliability: 1. **Test-retest reliability** - Giving a test multiple times - Issues: carry-over effects (remembering from first test), change in true score from maturation 2. **Equivalent forms (like split-half testing)** 3. **Internal consistency - Cronbach's alpha (α)** - If every possible split-half correlation was computed, the average is Cronbach's alpha ### Systematic (Non-random) Error Variance - Doesn't decrease estimated reliability - Reduces validity ### Validity - 2 aspects - Is it measuring what we think it's measuring? - Is the test appropriately used for its intended purpose? ### Content Validity: Coverage of the "domain" - Define boundaries (what is and what is not in the domain) - And structure (structure of domain) ### Construct Validity: How well defined is the construct measured by this test? 1. **Convergent validity**: Is it related to other theoretically similar constructs or tests? 2. **Discriminant validity**: Is it independent of other unrelated psychological constructs? #### Diagram: - Reliable - Valid - Invalid - Unreliable - Valid - Invalid