influences of intelligence- lecture 9

maltby chapter 13 & 14

francis galton 1822-1911

francis galton 1822-1911

Polymath
• Created the first weather maps
• Designed fingerprint classification systems
• Invented correlation
• Invented regression
• First to use the survey as a form of data collection
• Forefather of individual differences
• Measurement and quantification
• First to attempt to measure intelligence directly
• Reaction time, sight and hearing, ability to distinguish colour
• Coined ‘nature versus nurture’
• Coined ‘eugenics’

Galton. (1869). Hereditary Genuis.?

Theorised that intelligence ran in families

Galton and eugenics

1885: Outlined genetic principles of intelligence
• Proposed that intelligence was inherited through genes
• Thought that people with lower intelligence were having more
children, causing the human race to become weaker
• Proposed that white people had higher levels of intelligence
than Black people and other ethnic groups
• Proposed using artificial selection to increase intelligence
among humans
• Protecting the ‘weak’ was at odds with natural selection

eugenics

Theory that humans can be improved by selective breeding
• Based on incorrect understanding of genetics
• Believed complex human traits such as intelligence were inherited in a simple
fashion from parent to child (like Huntington’s disease)
• Believed complex human traits were solely the outcome of genetic inheritance (not influenced by environment)
• ‘Positive’ eugenics: encouraging increased reproduction in people with ‘superior’ traits
• Encourage white people from higher social classes to have more children
• ‘Negative’ eugenics: discouraging or stopping reproduction in people with ‘undesirable’ traits
• Encourage people from low social classes to not have children
• Used pseudoscientific theories to argue for the superiority of white Europeans

eugenics in the US

Lewis Terman, president of the American Psychology
Association (1923)
• Low intelligence often found in Spanish-Indian, Black, and Mexican
families
• This low intelligence was inherited
• These groups should be segregated
• Concern that these groups had more children than white Americans did.

sterilisation laws in the US

922: Model Eugenical Sterilisation Law: Mandatory sterilisation
for certain groups
• ‘Feeble-minded’ (low intelligence)
• The ‘insane’
• The criminal
• The epileptic
• The blind
• The deformed and dependent (orphans, homeless)
• Estimated that >42,000 people were sterilised between 1941 and
1943
• Not just US → many other countries had eugenic policies,
including compulsory sterilisation, marriage restrictions, birth
control and immigration control
• Canada, Sweden, Australia, Norway, Finland, Demark and Switzerland

eugenics after WW2

Eugenics fell into disrepute
• US: Most states based eugenic sterilisation programs ended after
WWII
But, sterilisations continued for:
• Intellectual disabilities and mental illness
• Native Americans
• Black women
• Mexican women
• Prisoners

Behavioural genetics of intelligence:

estimate the degree of genetic heritability of intelligence across a population

Genetic heritability:

the degree to which variation in a phenotype (observable trait) is due to genetic variation between individuals of a given population

Heritability estimate (h2):

Average estimate of the proportion of variance for intelligence thought to be accounted for by genetic factors across a population

additive assumption

The combined effects of genes and environment are equal to the
sum of their individual effects
• G + E = 100%
• Relative strength of both → if one is high, the other is low
• 100% - Genetic contribution = Environmental contribution
• Simplification: we know genes and environment interact (G * E)
• Considered a starting point → used as the upper limit of genetic
heritability

Heritability estimate = 50%

• Does not mean you inherited 50% of your intelligence from your genes
• It means that 50% of the variation between people in a given population on intelligence is due to genetic factors

environment- shared/ non-shared

Shared environment (c): parents, household, teachers
• Non-shared environments (e): peer group, illnesses

assumptions about non-/shared environments

Assumed that shared environment makes the same contribution to
both identical (MZ) and fraternal (DZ) twins
• Assumed that non-shared environment is not shared

Correlation between pairs of identical twins consists of:

100% shared genes
• Identical shared environment

Correlations between pairs of fraternal twins consists of:

50% shared genes
• Identical shared environment

gene contribution-

double the difference
between the size of the identical twin
correlation and fraternal twin correlation
• (MZr – DZr)x2

studies & their estimates of intelligence heritability

Eysenck (1979): 69%
• Herrnstein and Murray (1994): 74%
• Chipeur et al (1990): 50%
• Carroll (1993): 40%
• APA Taskforce on Intelligence (Neisser et al., 1996): 40% to 80%

haworth et al 2010: conclusions

The genetic influence of intelligence increases with age
• May account for some of the variation in heritability estimates of
intelligence

Why do genetic influences increase with age?
• In childhood, children have little say over their activities and learning
• As children grow up, they are increasingly able to select their own
experiences and may be more able to express their genetic differences
• Although it decreases with age, shared and non-shared
environment still contribute substantially to intelligence

limitations- behavioural genetic studies

Additive assumption is a simplification → Genes and
environment can interact
• Families used in family/twin studies may not be representative
of the normal population
• Kamin and Goldberger (2002): assumption that identical and
non-identical twins have similar environments may be wrong
• Carey (2002) estimating heritability and environmental
influences from behavioural genetics
• Abstract, theoretical concepts: tell us nothing about the specific
genes or environmental influences of intelligence
• Population concepts: estimates refer to a group of people - tell us
little about individuals

davies et al 2018

Many genetic variants (SNPs) associated with intelligence
• Many genes associated with intelligence
• Each SNP/gene individually had only a very small effect
• SNP-based heritability for intelligence ranged from 12% to 25%
• Common genetic variants accounted for between 12% and 25% of the differences between people on intelligence test scores
• There was genetic overlap between intelligence and many health- related traits
• The same genetic variants associated with intelligence were also
associated with a range of physical and mental health-related traits

a complex trait

is influenced by a large number of genes

the difference in heritability estimates using behavioural techniques vs molecular

Heritability estimates are lower using molecular compared to
behavioural genetic techniques
• Behavioural genetic studies likely overestimate heritability of intelligence
• SNP-based heritability likely underestimates heritability of intelligence

is intelligence heritable?

Despite the differences in estimates, behavioural genetic and
molecular genetic studies show that intelligence is heritable

genetic influences- reduce/ increase or stay the same?

increase with age

the bell curve

Herrnstein and Murray (1994). The Bell Curve: Intelligence and Class
Structure in American Life.
• Analysis of IQ test scores in the US
• Extremely controversial book
• Huge debate

• The extent to which intelligence differences are the cause of social issues in the US
• The extent to which intelligence is genetically inherited
• Racial differences in intelligence test scores
• Whether genetic differences accounted for the racial differences in intelligence test scores

bell curve- racial differences in IQ

Wechsler Intelligence test scores:
• White Americans scored 15 IQ points higher than Black Americans
• White Americans averaged IQ of 102
• Black Americans averaged IQ of 87
• Gap between the two races is around 1 SD
• Argued that intelligence is largely genetically heritable and
therefore the racial differences in intelligence test scores are
genetic in origin

criticisms & the APA taskforce on intelligence ( Neisser et al 1996)

Large differences in the average IQ scores between Black and
white Americans
Differences in intelligence test scores may not reflect
differences in intelligence
• Tests could be culturally biased

There is no evidence to suggest these differences on
intelligence test scores between white and Black Americans are
caused by differences in genetics
• Social and economic differences

bell curve assumptions

Herrnstein and Murray’s argument:
• Intelligence is heritable
• There are ethnic differences in intelligence test scores
• Therefore, ethnic differences in intelligence are due to genetic
differences
• Taskforce agreed that intelligence is substantially heritable
• Taskforce disagreed that heritability can be used to explain
differences in IQ test scores between ethnic groups
• Heritability can only be used to explain average differences within a population group
• It cannot be used to explain average differences between groups (white Americans versus Black Americans)

issues with the bell curve

Heritability is not deterministic
• Bell curve ignored environmental influences of intelligence
• Genes and environment interact

Possible environmental explanations include:
• Bias
• Discrimination
• Poverty/deprivation

brain size and IQ

Brain size is correlated with higher intelligence
• McDaniel (2005): Meta-analysis of 37 studies (n = 1,530)
• r = 0.33
• Correlations larger in female than male participants
• Correlation larger in adults than in children
• For all ages and sex groups brain volume positively correlated with intelligence

• Race classifications do not reflect genetic similarity

Genetic variance within a race can be nearly 10x greater than the
genetic variance between races
• Between-group genetic differences do not map onto socially
recognised categories of race
• Consensus in biological and social sciences that race is a social
construct and not a genetic construct

cox et al 2019 results

Total brain volume correlated with intelligence
• r = 0.28, p < 0.001
• Examined sex differences in the relationship between total brain
volume and intelligence: No sex differences
• Grey matter:
• r = 0.28, p < 0.001
• Normal appearing white matter volume (healthy white matter):
• r = 0.25, p < 0.001
• White matter hyperintensities (damaged white matter):
• r = -0.11, p < 0.001

cox et al 2019- conclusions

Higher intelligence associated with: bigger brains, more grey
matter, more healthy white matter, less damaged white matter
• Differences in these brain indices accounted for 6.16% of the
variance in general intelligence

• Observational study: cannot disentangle directionality
• Smaller brain → lower intelligence
• Lower intelligence → smaller brain
• Something else cause both? (confounding)

sex differences in intelligence- johnson et al results

Two of the most population-representative samples of intelligence
ever complied
• SMS (Scottish mental survey) 1932
• No sex differences between males and females
• 0.018 standard deviation difference
• 0.28 raw points higher for female participants
• SMS1947
• Very small significant difference between males and females
• 0.109 standard deviation difference in favour of females
• 1.72 raw points higher for female participants

sex differences specific abilities

Some differences in specific abilities

Female participants tend to be better at verbal tasks
• Reading, comprehension, writing

Male participants tend to be better at some spatial tasks
• Especially mental rotation of 3D objects
These differences are small
These differences may cancel each other out when examining
general intelligence measured using a wide range of different
types of tests

sex differences- context

If/when sex differences exist, they represent the average differences between men and women, not differences between
individuals
• Knowing whether someone is male or female does not reveal
anything about their intelligence
• Many female participants will score higher than male participants
• Great variability

iodine and intelligence- results (qian et al 2005)

children receiving supplements scored 8.7 IQ points higher
than children not receiving supplements
• 3.5 years after the supplementation program was introduced,
increased to:
• 12 IQ points on non-verbal (Raven’s)
• 17.25 IQ points for verbal (Stanford-Binet)
• This study: Children who were iodine deficient
• No evidence iodine supplements improve intelligence in
children who are not deficient

education and intelligence- ritchie and tucker-drob

Meta-analysis of educational effects on intelligence
• 42 studies, >600,000 participants
• 1 additional year of education was associated with an increase in IQ of 1-5 IQ points
• Larger for ‘composite’ intelligence tests (mixture of fluid and
crystallised) than for fluid tests alone.

lead exposure and intelligence

Lead is a heavy metal that was extensively used in paints,
petrol, plumbing construction
• Lead is toxic, especially to children
• Linked to damage in brain and nervous system, slowed
development
• Evidence that lead exposure in childhood has a negative impact
on intelligence.

port pirie cohort study

Largest lead smelting plant in Australia
• Amount of lead in the body and pregnancy outcomes in 749
pregnant women
• Higher levels of lead in the blood of the mothers was associated
with premature babies
• Followed-up children up throughout childhood and adolescence
• Measured levels of lead in the blood of children antenatally, at
birth, and during follow-up
• Ages 11-13: Full scale IQ scores assessed at age 11-13 with
the Wechsler Intelligence Scale for Children

port pirie results

Children with more lead exposure had lower intelligence test scores
• Lead was significantly associated with IQ at all time points, except lead levels at birth
• After adjusting for covariations (including socioeconomic status & mother’s IQ) lower
IQ was associated with lead levels at:
• Age 5, Age 7 and Age 11-13

summary of influences of intelligence

Genetics influence intelligence → Intelligence is heritable
• Although intelligence is heritable, the assumptions about the
heritability of intelligence made by eugenicists and in The Bell Curve
were flawed
• Higher intelligence is associated with brain differences (larger brains,
more grey matter, more healthy white matter)
• There is mixed evidence whether there are sex differences in
general intelligence, although there are sex differences in specific
abilities
• Environment influences intelligence
• Environmental conditions can lower intelligence (lead exposure)
• Environmental conditions can increase intelligence (iodine supplements in
those who are deficient, education)

education and intelligence in Norway

Brinch and Galloway 2012

Genetics influence intelligence → Intelligence is heritable
• Although intelligence is heritable, the assumptions about the
heritability of intelligence made by eugenicists and in The Bell Curve were flawed
• Higher intelligence is associated with brain differences (larger brains, more grey matter, more healthy white matter)

• There is mixed evidence whether there are sex differences in
general intelligence, although there are sex differences in specific
abilities
• Environment influences intelligence
• Environmental conditions can lower intelligence (lead exposure)
• Environmental conditions can increase intelligence (iodine supplements in
those who are deficient, education)