Lifespan A: Intelligence, WEEK 4 Flashcards
(30 cards)
1
Q
Correlation co-efficients
A
- correlation is a way of describing the relationship between two things > shows direction + strength of a relationship from +1 to -1
- are group level statistics
- positive = both increase
- negative- = one increases while the other decreases
- non-linear association = +ve association up to a certain point then is -ve
- no association = 0
2
Q
Psychometrics
A
- refers to the science of psychological assessment + is primarily concerned w/ individual differences
- often uses indirect measures (e.g: measuring temp is reliant on height of mercury which indicates temp > not direct measure)
- concerned with individual differences + is a domain general theory (broad intelligence)
- people studying psychometric intelligence care about quantitative intelligence > sees dev as gradual, gaining speed + knowledge
3
Q
Reliability
A
- refers to the extent of consistency of a result on repeated trials
4
Q
Validity
A
- the extent to which something measures what it claims to measure
5
Q
What is intelligence?
A
- Psychometric intelligence refers to human differences measured by mental ability tests
- these tests do NOT claim to measure all human capability
- psychologists are interested in how people differently understand complex ideas, learn from experience etc..
6
Q
How is intelligence measured?
A
- Different types of tests which must be standardised available to trained psychologists to offer
- E.G: Single item type tests (e.g., the Ravens Progressive Matrices Test, the British Picture Vocabulary Test) > measures one core ability
- Multiple item type tests designed to measure wide-ranging cognitive abilities: Wechsler Preschool and Primary Scale of Intelligence (ages 2.5 – 7.5)
Wechsler Intelligence Scale for Children (ages 6 – 16), Wechsler Adult Intelligence Scale (ages 16 – 90)
7
Q
IQ test:
verbal ability
A
- verbal ability measures come in 2 broad formats:
1. receptive measures: show images and ask the ppt to show me this (e.g: show me the pen) > (WPPSI)
2. expressive verbal ability measure: researcher asks questions like how are red and blue alike? and the child has to generate a verbal response, not just point at a image > (WISC)
8
Q
IQ test:
Non-verbal reasoning
A
- Assesses problem solving skills which cannot be studied for
- e.g: matrix reasoning task (WPPSI) > have to find a pattern from the info available to understand pattern
- block design task (WISC) > give a set of blocks + a picture of a pattern they have to recreate + time them
9
Q
IQ test:
Working memory
A
- different from memory as in WM you have to remember something and perform an operation with it > assesses capacity to retain info + apply it efficiently
- E.G: Digit span backwards (WISC) > a set of number are said and the child had to repeat them backwards
- Letter-number sequencing (WISC) > where the child has to organise a set of numbers and letters in order starting with numbers
10
Q
IQ test:
Processing speed
A
- Assesses how accurate + quick you work in unseen conditions > captures how quick your perception + reaction is
- e.g: Cancellation test (WPPSI): put a cross through certain images like of clothing
- Bug search (WPPSI): find a certain insect and circle like find the worm
11
Q
What do these tests have in common?
Spearman (1914)
A
- All of these tests measure intelligence
- Spearman argues there is a positive manifold > tend to get strong positive correlations between the tests + is highly replicated finding (if you do well on one test, you do well on all tests usually) > meta analysis on 460 datasets show this (Carroll,1993)
- Spearman uses the positive manifold to argue this is proof of general intelligence (g factor) + all our intellectual ability is underpinned by an underlying difference between us > tested this using factor analysis
12
Q
Factor analysis
A
- used to determine the amount of latent variables (factors accounting for variation)
- method of simplifying large amount of data so we can visually see what happens in it
13
Q
Is there more than one form of intelligence
Thorndlike (1921)
A
- based on data of 800 soldiers across 7 tests, the results do not support spearman’s idea that there is one basic form of intelligence that underpins all ability
- Catell (1943) later found there are two broad types of intelligence:
1. Fluid Intelligence (gF) - the ability to work things out with no previous knowledge.
2. Crystalized intelligence (gC) – the established store of knowledge.
14
Q
Hierarchical organisation of intelligence
Carroll 1993
A
- description of what you see in data (not a theory)
- Carroll used the meta analysis of 460 datasets + analysed it statistically w/ factor analysis + created hierarchical organisation of intelligence
- 3 levels: Level 1 > give ppt all sets of tasks to do. Level 2> analysis shows they all correlate but certain tests relate more due to similar variance (e.g: remembering digits backwards + ordering letters + numbers both need WM) > this is cognitive domain variance
level 3 > above all this, all the abilities for the tests are related to eachother due to G (underlying intelligence factor) - we all have different intelligence due to G but we are stronger in certain domains like WM or verbal ability > underpinned by g (common intelligence)
15
Q
What is IQ?
A
- statistic representing someones reasoning ability in comparison to the average score for their age
- IQ tests are standardised + a mean score is set > calculate people’s difference from mean using standard deviation
- can be used to classify learning disability (if IQ is 2 SD less than the mean)
16
Q
How reliable are these tests?
A
- the tests are repeatable measures
- IQ tests show strong levels of test re-test reliability > results will be strongly correlated if tests are done at two points in time (+0.8 shows strong corr) > this is the case even if it is a different IQ test
- Deary et al (2000) found strong stability across life span when tested at age 11 and at age 77 > same rank order but IQ mean can change (discontinuity)
17
Q
Intelligence across lifespan
A
- sharp increase in crystallised and fluid intelligence early in childhood
- around age 25, crystallised intelligence stops > doesn’t increase or decrease but remains stable at a plateau
- fluid intelligence however gradually declines
(discontinuity)
18
Q
Twin studies
Genetically sensitive research design
A
- Twin studies can be used to calculate to what extent intelligence is affected by additive genetic factors, common environment and non-shared environmental factors
19
Q
Adoption studies
Genetically sensitive research design
A
- comparing adoptive parents who aren’t genetically related to biologically related parents and kids allows us to see the effect of environment and genes
20
Q
Is intelligence heritable?
Plomin & De Fries (1980)
A
- Created a graph which shows genetic relatedness and the correlation of IQ between those related
- e.g: test re-test = 0.87 corr, MZ twins = 0.86 (as if they re-took the test themselves) DZ twins = 0.64, siblings 0.34 etc..
- results from MZ twins show genetics have an impact as they share 100% of genes + had a very strong corr
- however, DZ twins + normal siblings share 50% of DNA, yet the correlation was significantly different which shows the ev has an affect
21
Q
Heritability
A
- degree to which individual differences in a trait is accounted for by genetic factors
- Plomin + Vonstumm (2018) find inherited differences in DNA sequence account for 50% of the variance in measures of intelligence found from twin studies
- twin studies allow us to see the contribution of genes without having to know exactly what genes are involved
22
Q
What heritability doesn’t tell us
A
- doesn’t claim that 50% of someones intelligence is due to their DNA > it is a population stat saying people’s intelligence differ partly due to genetics > is not simple transmission across generations
- Does not tell us anything about average IQ. Heritability estimates can be 50% even when the mean IQ is 85.
23
Q
Heritability changes with age
Davis et al, 2009
Haworth et al, 2010
A
- influence of heritability fluctuates throughout lifespan
- heritability is low in early childhood but gradually increases its influence (genes don’t change)
- this may be seen as environmental factors being most effective in early childhood as research shows genes explain 26% of variance and common EV explain 68%
24
Q
Genes & intelligence
A
- No study as of yet has indicated a particular gene to play a significant role in intelligence
- Individual genes tend to have very small effects (e.g., .05% variance in IQ)
- Genome wide association studies indicate therefore that intelligence is highly polygenic
25
Environmental influence on intelligence
- Environmental influences can be biological (e.g., nutrition, chemical exposure) or social (e.g., schooling, family circumstances) (Neisser et al., 1996).
- IQ is correlated with a range of family factors (e.g., resources in the home, parental language input, warmth) (Neisser et al., 1996
26
Environmental associations
- there are 2 core ways which can explain ev associations
1. social causation: this is where parents are of high socio-economic status and can provide the child with an enriching ev with all resources = high IQ of child
2. social selection: where parents intelligence influenced by their genes leads to them getting better ed and better jobs which lead to high socio-economic status thus the child has high IQ (kids are intelligent bc their parents are)
27
Adoption studies & intelligence
| Ijzendoorn et Al (2005)
- Adoptive parents strictly have a environmental relationship with adoptive kids so IQ correlation can be compared to biological families to see effect of ev
- Ijzendoorn (2005) did a meta analysis + found adopted children had higher IQ than the children who stayed in care or w/ biological parents (explained by ev change) > the adopted children then show the same kind of IQ as their adoptive siblings
- No longitudinal studies available to track changes in IQ
- Specific factors helping adopted children remain unclear
28
Flynn effect
| Flynn (1987)
- used Raven Matrices data from 5 countries + found on average each generation seems to go up slightly in IQ > 2 SD increase in IQ per 100 years
- Flynn argues this cannot be due to genetic change as this time is too short for that so it must be due to EV perhaps due to nutrition, technology or ed (but we don't know what ev influence exactly affect IQ)
- alternatively, it could be due to problems with IQ measurement > IQ tests might be able to rank people within a particular generation but not across generations.
- We must be cautious of comparing different ethnicities, races and other groups.
29
Education and IQ
| Ritchie & Tucker-Drob, 2018
- IQ is associated w/ longer time spent in ed
- meta analysis of 142 studies where there was a change in time spent at ed due to ed reform > 1 extra year spent in ed = 3 point increase in IQ > supports social causation theory, can't be social selection as kids didn't choose to stay in ed, they had to
(social causation = more ed > higher IQ
social selection = higher IQ = more time in ed)
30
Gene x Environment interaction
| Tucker-Drob & Bates (2016)
- research on origins of individual differences pits genes against environment
- genetic influences are expressed more in nourishing environments
- meta-analysis of socioeconomic status effect on heritability of IQ
- genes exerted weaker influence on IQ in low SES groups (in the USA but not in Europe) > this is because places in Europe like the UK have a social safety net which supports people financially unlike USA
- suggests genes have more effect if you live in a privileged ev whilst people from low ses experience less of an effect from genes as their genetic potential is restricted by their environment.
