Intelligence Flashcards

Question

How is Divergent thinking measured?

Answer 1

* Torrance Test of Creative Thinking (TTCT): * most widely used test, intended for children, * Includes both verbal and visual components * Scoring Divergent thinking * *ideation fluency*: number of different responses * *flexibility*: number of different categories of responses * *originality*: statistical frequency of answers (reciprocal of count number ) * Also considered for visual: elaboration (degree of detail), titles (complexity)

Answer 2

* Originality confounded with flexibility * Statistical rarity is ambiguous * Uniqueness scoring penalises large samples * Tests are timed * Originality and fluency change with instructions (write as many/weirdest) * May not actually result in creative achievement * mixed results for divergent thinking correlation

Answer 3

* Corelations with OCEAN: * Scientists: small/med association with openness, extroversion * Artists: high openness, high neg correlation with C+, high psychoticsm * Compounding factors: gender, causation * Divergent thinking and Psychoticism * Different for different types of DT scores (fluency .12, flexibility .00, originality .16-.40)

Answer 4

* The Threshold hypothesis * A theshold of intelligence is required for creativity (above which unrelated) (120) * Meta-analysis: no sig dif at 120, but found diffs between diff DT scoring * Segmented Regression: different thresholds for different DT scoring (achievement none vs fluency 86 vs originality 120)

Answer 5

* EI is a set of abilities involved in processing and manipulating emotional information * Example: *The 4 branch Hierachical model*: * 4 hierachical abilities (higher level require lower level) * 1. Emotional Regulation, * 2. Emotional Understanding, * 3. Emotional Facilitation (dubious - using emotions) * 4. Emotional Perception

Answer 6

* EI is a mix of constructs: traits, emotion-related abilities, motivation, beliefs etc * Examples * Trait EI (tested using the TEIQue) * 15 facets including the 4 ability model, personality traits (eg low impulsiveness * Problem: Extremely broad definition makes testing/conculsions difficult * Social and Emotional Competence (Test = EQ-i) * 5 broad sets of skills, each with subsets of skills * Intra-personal Abilities, Inter-personal Abilities, Adaptation, Stress Management, General Mood * Not supported by factor analysis (particularly the subscales but still widely used

Answer 7

* Ability scales (maximum performance): * eg rate effectiveness of responses to scenario, rate emotions present in face * Benefits: relate to intelligence, cannot be faked * Issues: dont tend to corelate to personality, more difficult to measure, time consuming * (Self) Rating Scales (typical performance): * eg Schutte self report scale, 'i know why my emotions change' * Benefits: high corelations with personality, easy, fast, high internal consistency, * Issues: very easily faked * The MSCEIT : most common but commercialised

Answer 8

* **Stream 1:** Ability model + Ability measurement * Example question: Given situation X what is most likely to have caused this response? * **Stream 2**: Ability model + Rating Measurement * I have problems dealing with my feelings of anger . [R] * **Stream 3**: Mixed mode + Rating Measurement * On the whole, I`m pleased with my life * Large overlap with general personality tests, particularly with neuroticism and extroversion scales - poor discriminent validity

Answer 9

* 2010 meta-analysis of EI with personality, intelligence, job performance separately by stream of measurement * OCEAN corelations * Stream 3 \> Stream2 \> Stream 1 measures * Emotional stability highest corelate (for streams 2 and 3) * Agreeableness highest corelate for stream 1 * Emotional management highest corelate with agreeableness * Intelligence (IQ) * Stream 1 \> stream 2, 3 * 4 branches: highest corelation with understanding

Answer 10

* Stream 1 EI * 4 branches: Highest for management branch (7% for high EL jobs) * Emotional Labour: Higher for high emotional labour jobs than low emotional labour jobs * Stream 2,3 EI * Stream 3 best predictor (higher corelates), 2 still better than 1 * Higher for high emotion labour than low emotion labour jobs * Valuable? * 7% is non trivial: EI differences are more influential when IQ is relatively equal across job environment (all lawyers high IQ) * Job performance is not everything: EI has high predictive power in coping with stress (10-15%) and wellbeing (29%)

Answer 11

* Lynn and Irwing Meta analysis: 'g' as measured by Ravens Progressive Matrices * Findings: Small effect size in favour of men * \<15 years = no difference * 15 - 19 years = Males 2 IQ higher * Undergrads = Males 3-5 IQ higher * Adults = Males 5 IQ higher * Compounding issues * Mostly students measured (female bias, lower ability females incl) * Ravens PM may include spatial ability compound: when controlled for spatial abilites no diff found

Answer 12

* **Moderate female advantage**: Spelling, Language * **Slight female advantage**: Perception, Speech * **Slight male advantage**: Mathematics, Science * **Moderate male advantage**: Spatial perception * ***Large male advantage:*** Mental rotation, Mechanical reasoning

Answer 13

* No difference in test score averages * Argument from test scores * tests are intentionally constructed to reduce bias - differences masked by tests * Argument against averages * Mean will not capture variability * males tend to be more variable (overrepresented in high achieving and low achieving groups)

Answer 14

* Evolutionary * Male ‘foraging’ hypothesis: navigating land, throwing weapons * "range" hypothesis: males cover more area, seek polyamorous * Warfare hypothesis: males travel futher to fight other males * Brain Function * Tasks solved more efficiently using one side of brain: Men tend to use only one half solving spatial problems * Testosterone * Route learning: testosterone associated with male route learning strategies (compass points, not landmarks) * Mental rotation: testosterone predicts spatial rotation task performance

Answer 15

* Stereotypes/confidence: * Self reporting on abilities corelates with observed diffs in specific intelligence. Men higher overall * Toys * Boys play with trucks (moving parts) girls play with dolls (roll-play) * Activities * Boys given more freedom to roam, more encouraged to play videogames * Education/subject choices * Women given less motivation/encouragement to engage in "masculine" subjects like math/science * SES * No differences are found in low SES environments - no access to technology, more freedom to roam?

Answer 16

* Basic Premises * 1. Genetic component to intelligence * 2. There are mean differences in racial intelligence * 3. These differences are genetic not environmental in origin * Race Differences based on bell curve research * Black white: estimated ~1 SD, d=0.7-1, 10-15 IQ points * Asian \> White (~d=.33, 5 points) * White \> Hispanic (~d=.5, 7-8 points) * Conclusions * proposed that enrichment doesnt affect IQ gap * Classrooms should be segregated by race for optimal learning

Answer 17

* **TIME**: The IQ gap has narrowed (Dickens & Flynn, 2006): * 1972: 15-point IQ gap, 2006: 10-point IQ gap * **SES**: The associated achievement gap is partly due to SES * About half a standard deviation (7.5 points) is accounted for by SES variables * **AGE**: The gap increases with age – both IQ and achievement gaps begin small and get larger

Answer 18

* The logic: * IQ predicts success in life, IQ is genetic, therefore group differences in IQ are genetic and are therefore unchangeable * Lack of success among Black people is due to genetic causes, therefore, affirmative action/head start programs will not work * 4 part book 1. *Social Darwinism in the U.S.* 2. *IQ relates to social problems* (poverty, crime, teenage mothers etc) 3. I*Q and race*: IQ differs between races and is lower for immigrants * Low-IQ races and immigrants have more children, create more social problems and are less successful 4. *IQ and Social Policy* * Best way to “adopt out” children of low-IQ mothers * Put more money into “gifted education” less into remedial * Scale back “quota” programs in favour of a “colour-blind” system

Answer 19

* **Premise: IQ predicts sucess**: about 25% of success at work and school is predicted by IQ but also predicted by other factors * Conscientiousness: Meta-analysis r=.22 (equal to Intelligence in review) * Termans Termites: High IQ predicted higher income, education but no differences in social success (alcoholism, divorce) * **Premise: IQ is genetic:** * Not completely: current estimates .4 - .7 genetic factor of intelligence * Definition of intelligence: the Bell Curve is based on measures of Crystalline intelligence only * **Premise: Group differences are genetic and therefore unchangeable** * Flawed logic: genetic component doesnt equal genetic cause of differences * counter example: identical plants one denied sunlight

Answer 20

* Black population IQ is increasing over time (~3pts per decade) * Currently at the level of white population in the 1950s * The current black environment is arguably equivalent to white 1950s environment in terms of * income * access to health care

Answer 21

* Mean IQ is different for Nations * Corresponds to GDP and other wealth measures * Arguments * 1. Differences between countries have genetic basis * 2. Differences in IQ cause differences in wealth

Answer 22

1. **Estimation of IQ** * IQ “estimated” for 104 of 185 (56%) of countries reported * Measured different time-points, tests (not adjusted for Flynn effect) 2. **Unsophisticated statistics** * Simple correlations – not appropriate for financial data (skew), do not account for structural factors 3. **Invalid logic**: * Cause of within group differences (genetics) NOT necessarily the cause of between group differences 4. **Causal direction is not clear** * IQ -\> Wealth versus Wealth -\> IQ 5. **Neglect of other important variables** * Motivation, structural factors, possibility of educational attainment

Answer 23

* IQ test scores are increasing over time * Each generation gets higher scores than the previous * ~3 points per decade * Relevance to nature v nurture * gene changes cannot possible explain an effect size this large * there must be some environmental changes causing this

Answer 24

* Use natural experiments to determine whether increases due to age or schooling * compared youngest and oldest children in same grade (just age effects) * compared children of different grades of same age (just school effects) * Cahan and Cohen Results: * 1. Schooling was a major underlying factor for an increase in intelligence * 2. Schooling showed a different impact on different types of tests * more impact on verbal (Gc), less impact on non-verbal (Gf) * Other Issues: * Those attending regular schooling score higher on IQ tests * Those who delay schooling lose 5 IQ points per year * Summer “drop-off” in IQ scores

Answer 25

* 'Headstart' initiative in the 60s to combat poverty * enriched pre-school program for disadvantaged children * Intended to build vocabulary, numeracy, and writing skills * Failure? Initial strong increases (7-8pts) decreased within 2 years * However, meta-analysis of enrichment programs * less likely to repeat a grade or require special education * more likely to graduate * less incarceration rates * more likely to attend university * tend to have higher earnings * siblings of “head start” participants gain advantages

Answer 26

* **Iodine:** Very strong evidence * iodine deficient children v non-deficient or supplements during pregnancy: between 4 and 17 point increase * **Vitamin supplements** * increased IQ by 9 points * BUT inconsistent replication * **Breastfeeding** * 1. Meta analysis * controlled for 5 or more covariates * 3.16 IQ points higher, 2.66 if healthy weight, 5.18 for underweight * 2. Meta analysis * small effect .5 IQ point * 3. Belarus Study * randomly allocate mothers to breastfeeding * 5.9 point difference overall * BUT confounding factors like cuddle time, support groups

Answer 27

* Evidence for: * Galton: men of science 48% first born sons, Current: First borns overrepresented in nobel prizes * Netherlands military testing: RPMs test, controlled SES * clear relationship between birth order and intelligence, and between family size and intelligence * confounding factor of religion (catholics in the 40s) * BUT: effect is social order not birth order * Theories * Resource: parent resources finite, 1st born gets 100%, 2nd 50% etc * Confluence: extra adult attention, more time socialising with adults not siblings, added responsibility of taking care of siblings * Within family effects * No relationship found - suggests compounding SES factors

Answer 28

* The major providor of longitudinal age/intelligence research * 1932 batch: * ALL 87,498 children born 1921 take Moray house test * + 1000 followed up with Standford Binet; followed into adulthood * 1947 batch * ALL 70,805 children born in 1936 tested * about 1000 still alive for followups * Moray House test #12: measure of crystalline intelligence, original purpose for high school admittance

Answer 29

* Stability = rank order of people over time * tested using correlations (cutoff for correlations ~.6) * Theoretical meaning: intelligence is stable over tme * Change = absolute level of trait over time * tested by changes to mean over time * theoretical meaning: intelligence develops, changes, declines over time * Results from SMS * Intelligence is stable (less so in children under 6-8) * but changing (overall means are rising)

Answer 30

* Longitudinal research involves following a single cohort over time * Benefits * Controls for cohort differences (education, maternal age, flynn effect) * Doesnt show ridiculous drops that cross-sectional studies show over age * Problems * expensive, time consuming * can underestimate cognitive decline * retest inflation: repeating test (get better at it), older people more time * selective attrition: low IQ cohort members likely to drop out, die

Answer 31

* Looks at age differences in the population at a single time point * Cross-sectional research results * The Weshler-Belleview (39): Huge declines on performance (25pt by 47) * WAIS-R (81): large drops in performance (10pt age 45, 30 age 72) drops in verbal (none at 45, 7 by 72) * Gf decreases, Gc increases then plateaus * Benefits: Easy to do, cheap * Issues: overestimates cognitive decline * compounding cohort effects (education, wars, women in workforce etc) * IQ testing is restandardised regularly - higher drop due to flynn

Answer 32

* *The Standards for Educational and Psychological Testing* define test bias as * _"construct underrepresentation_ or _construct-irrelevant_ components of test scores that _differentially affect the performance of different groups_ of test takers" * **Construct underrepresentation** is when a test that covers only certain aspects of a construct being understood as representative of the whole (eg Ravens = IQ) * **Construct irrelevance** is when a test requires/examines factors that aren't relevent to the target construct eg knowledge of currency for arithmetic * **Note**: must differentially affect groups of the same ability for the construct being measured (ie not depressed people for depression scale)

Answer 33

* Construct irrelevant bias: * reading comprehension: familiarity with writing style ie bible, womens mag (but must be systematic) * very complex verbal instructions for a math test * Content bias: * culture biases: the army alpha/beta tests familiarity with tennis, different names for things across culture (truck vs lorry) * Test bias: * rating scales: depression scale in hospital (sick people will feel weary) * language bias: feeling blue (idiom)

Answer 34

1. Differences between groups * NOTE: Group differences alone are not an indication of test bias, since it could be a genuine difference 2. Differences between groups when ability is the same * Differential item functioning (DIF) is a way of testing for item bias * DIF occurs when individuals with the same level of the attribute obtain different scores on an item or subset of a test 3. Differences in prediction * systematic over/underprediction of a group's performance * Most important when test is used for selection for some valued opportunity (usually educational or employment) * Check if the regression line is different for different groups

Answer 35

1. Selection/placement: educational, job selection, MENSA * Issue: people cheat on obvious aptitude tests 2. Diagnosis: learning disabilities, mental health screening * inconsistent scoring (eg dyslexia) * Generally needs to be combined with achievement, history 3. Counselling: career transition advice 4. Policy making: national curriculum, workplace policies 5. Forensics: diminished capacity in court, custody issues 6. Management: best methods of handling different employees

Answer 36

* Early estimates placed intelligence test and work performance correlations at .35-.45 * But predictive biases in tests have adverse effects * Search for 'fairer' alternatives actually had worse bias * eg peer review(.41), biodata(.35) * Schmidt Meta analysis * best predictor of work performance is past work performance (.54)

Answer 37

* Best estimate is that intelligence predicts school achievement at r = .50 * But these are review papers rather than meta-analyses * Predictors of academic achievement at different levels of education * Intelligence is more important in the early grades * Conscientiousness is always important * Other personality variables are only important in the early grades

Intelligence Flashcards

(62 cards)