7- Genetics, heritability and group differences Flashcards
Heritability (h²)
Heritability = the proportion of this total variation between individuals in a given population due to genetic variation.
* The question of whether a trait is heritable is a
question about the role that differences in genes play in the phenotypic differences between
individuals or groups within a population
* Heritability estimates for IQ are typically around 0.5
* Some estimates as high as 0.8 (e.g., see course manual page 227; Scarr, 1983)
3 major considerations:
1) A heritability coefficient is always specific to a particular population at a particular time
* Example: the heritability of weight is approximately 0.6 in North America. Would this coefficient be expected to change if we measured it 1000 years ago in Medieval
England?
2) Heritability is a measure of variation
* For some attributes there may be no person-to-person variation
* A study which tried to determine the heritability of
“having a heart” found that h2 was equal to 0. What
does that mean? – Does it mean that having a heart is not genetic? – No, heart development is strongly influenced by genes – It means that individuals who don’t have a heart are that way because of the environment they grew up in, not their genes?– No, it means there is no variation across the sample :
everyone has a heart so the coefficient is 0.
(Similar to a correlation coefficient : it’s a measure of variation !)
3) A high heritability coefficient does not mean that a trait “is genetic”
* If the heritability of weight was 1 would this mean that one’s weight is 100% determined by genes?
- No, it is not that simple.
– Genes cannot be understood outside the context of the
environment with which they interact.
– A heritability coefficient of 1 does mean that trait differences
between individuals within a particular population at a particular time are due to genetic factors. However, if a certain portion of the trait is invariant across the population, heritability does not account for genetic influences on the portion of the trait that does not vary.
– Also, if you change the population/time, you will change the
heritability coefficient
3 ‘LAWS’ OF THE GENETICS
OF COMPLEX TRAITS
(INCLUDING INTELLIGENCE)
- Intelligence has some genetic basis
- Intelligence is not 100% heritable
- Intelligence may be caused by many genes of small effect
- Intelligence has some genetic basis
* Rett syndrome
* William Syndrome
* Fragile X syndrome - Intelligence is not 100%
heritable
* All traits show some
environmental influence
* There is also an Interplay
between genes and
environment
* IQ = Genes + Environment
+ Interaction + error
Gene x Environment interaction:
– An interaction effect is not the same as a correlation
– Interaction effects quantify how genetic characteristics can change the influence of an environmental variable
* Example: in one study, researchers were able to identify a gene that
was associated with severe violent behaviour, but only for individuals who had been victims of abuse
during childhood.
* Individuals who had not been
victimized, but who carried the
gene, were less likely to commit
violence
- Intelligence appears to be caused by many genes of small effect
(polygenic model)
- Intelligence appears to be caused by many genes of small effect
- Genome-wide association (GWA) studies
- Benyamin et al 2014
- 18 000 children
- No genome-wide significant associations of IQ with 1 gene
- Largest effect sizes = 0.2% of the variance of intelligence scores
- Desrivieres et al 2014
- 1500 children
- Largest effect sizes = 0.5% of the variance of intelligence scores
- Genes can also interact (i.e., the expression of one gene might depend
on the presence of another gene)
The traditional way to study nature-nurture
- Twin studies – Compare identical (monozygotic) vs fraternal (dizygotic) twins
- Identical twins share 100% of their genes; fraternal share 50%
- Identical and fraternal twins share 100% of their environment (or as close as we can get to 100% in the real world)
- Fraternal twins share more of the same environment than regular siblings do
- Niche picking
- Adoption studies – Compare identical (monozygotic) twins raised apart or together
- Important notes:
– Identical twins raised together share more than just genes - Correlation between IQ scores of monozygotic twins reared together: 0.86 (explains approx. 70% of the variation in IQ) - Correlation between IQ scores of monozygotic twins reared apart: 0.72 (explains approx. 50% of the variation in IQ) – Twin studies have small sample sizes
New data : human genome project
1959: Trisomy 21 was described in Down syndrome by Lejeune
2005: Human Genome Project was officially completed
Human Genome Project:
*Allows estimates of genetic influence on
complex traits using genome-wide genotypes in large samples of unrelated individuals
*Limited to measuring additive genetic variance
Plomin and Deary (Nature, 2015)
* New method : Genome-wide
Complex Trait Analysis (GCTA)
that allows much larger analysis
than the typical twin studies
5 Main Findings:
1- Heritability of intelligence increases from
infancy to adulthood.
Plomin and Deary (Nature, 2015)
20% in infancy
40% in adolescence
60% in adulthood
Hypothesis : Small genetic differences are magnified as children select, modify and create environments correlated with their genetic propensities (active
model of selected environments)
Proximal vs distal effect of genes!
« Gene-environment interaction » (Krueger et al 2008)
2 - Genetic correlations among
cognitive tests are typically
greater than 0.6
-The same genes are
responsible for the
heritabilities of these tests.
-These tests seem to tap into different cognitive processes….
But still, these sub-tests would tap into
“similar” genes (Genetic correlations of 0.6)
3 - Assortative mating is greater for
intelligence than many other traits
(like personality, psychopathology or
physical traits)
* spouse correlations for
intelligence (~0.40)
* spouse correlations for height
and weight (~0.20)
4 - Unlike psychiatric disorders, intelligence is normally distributed
-There is a positive end of high performance as
well as a problematic end of intellectual
disability.
-Some suggested that genius emerges from unique combinations of genes: Discontinuity Hypothesis. …
Not confirmed by these data
5 - The same genes underlie variations
in intelligence, education and SES
We cant really know what is causing
what but it is in line with some of the
data from 2 weeks ago…
Nurture : Environment and IQ
Prenatal Effects
* Birth weight – Correlated with later IQ
– Very modest Whites .07 , Blacks .11 – Effects are driven by extremely low birth weights.
* Smoking / alcohol use… fetal alcohol syndrome
* Nutrition – Mixed results
- Dutch famine study showed little long term effect of famine.
- A large meta analysis suggested that all prenatal effects accounted for very little variation beyond maternal education.
– Correlation increased from .42 to .47 with correction for prenatal effects
– Sometimes post natal nutrition has effects, or not.
– When malnutrition is severe it most certainly does.
– Probably not a big factor on midrange IQs in the west.
* General health
– High IQ is related to having better health
- Health → IQ or IQ → Health??
* Pollutants
– Has a demonstrable effect on IQ at high levels
- If lead levels in the bloodstream reach 25 mg per 100 mL there will be deleterious effects on IQ
- This level of lead exposure is rare and associated with poverty – Low levels of pollution appear to have either very small or no effect
* Education
– Has meaningful effects on IQ (Brinch and Galloway 2012)
– Effect on other cognitive tests (Rithcie et al 2015)
-Another study reported that children’s IQs drop over the summer holidays (Ceci, 1991) Particularly in lower SES
-Effects of age and schooling in grades 5 and 6 on raw scores at IQ tests.
They compared the younger and older of each school year
(quasi-experimental paradigm)… see slide 24 because idgi
Geographic / Group differences in IQ
- It has been argued that there are large genetic differences between members of different geographic cultures.
- Sub-Saharan Africans and Australian aboriginals are claimed to have a mean IQ of around 65.
- This may be true of their test scores but what does this mean?
- Is it a reflection of broad overarching structures of oppression?
- Flaws in IQ testing or underlying assumptions around what IQ
represents? - Genetic differences between different “races”? Does the term “race”
have a valid genetic basis? - Consider Mackintosh (2011): “the genetic differences between the
populations of the rest of the world are smaller than those between
different inhabitants of Africa: the genetic diversity of modern
Africans is greater than that of the rest of the world put together
(Lewin and Foley, 2004; Tishkoff et al., 2009). This makes it all the
more implausible to talk of ‘the IQ of sub-Saharan Africans’.”
An example of the position that differences in intelligence among geographical groups is influenced by genes:
“The position of environmentalists that over the course of some 100,000
years peoples separated by geographical barriers in different parts of
the world evolved into ten different races with pronounced genetic
differences in morphology, blood groups, and the incidence of genetic
diseases, and yet have identical genotypes for intelligence, is so
improbable that those who advance it must either be totally ignorant of
the basic principles of evolutionary biology or else have a political
agenda to deny the importance of race. Or both.”
(Lynn, 2006, p. 244)
Other estimates say 50 000 rather than 100 000 years (Lewin and Foley,
2004)
A counter point (the cause of racial IQ disparities is the environment):
“Rushton and Jensen (2005) ride roughshod over the evidence concerning whether the Black-White IQ gap has a hereditary basis. The most directly relevant
research concerns degree of European ancestry in the Black population. There is not a shred of evidence in this
literature, which draws on studies having a total of five very different designs, that the gap has a genetic basis.”
(Nisbett, 2005)
A second counter point (the debate is unscientific):
“There is not a single example of a group difference in any complex
human behavioral trait that has been shown to be environmental or
genetic, in any proportion, on the basis of scientific evidence.
Ethically, in the absence of a valid scientific methodology, speculations about innate differences between the complex
behavior of groups remain just that, inseparable from the legacy of
unsupported views about race and behavior that are as old as
human history. The scientific futility and dubious ethical status of
the enterprise are two sides of the same coin. […The genetic
argument] is based on a weak brew of unexamined intuition and
sketchy empirical evidence. In a free country and a free academy,
scientists can speculate about whatever they want, but their speculations should not be mistaken for a scientific consensus or a legitimate basis for social policy.” (Turkheimer, Harden, & Nisbett,
2017)
Discussions about the
determinants of IQ invoke a moral
imperative
- Attributing racial differences in IQ to genetic causes is inherently dangerous,
because these claims have been used in the past to justify racist political
agendas - Consider the legacy of eugenics (e.g., killing of the institutionalized; the
segregation and genocide of races perceived as inferior) - Genetic arguments can be used as a justification to avoid changing social
conditions and to minimize the importance of oppression as a cause of racial
inequality - What is the point of improving social conditions if racial inequalities are
immutable?
Are Cultural Differences Immutable?Indian children’s scores on the Performance half of the WISC and on Raven’s Matrices
as a function of the length of their residence in Britain (data from Sharma, 1971)… longer they lived in Britain, better their scores got
* Apparently not.
* People from a non-western culture lose all of their disadvantage
with a fairly short exposure to our culture, and the abstract reasoning skills that are privileged within Western education
systems
* This is also true of blacks in Britain
* But there is still that troubling low level of performance that remains
- At least in these old data.
Black-White IQ differences in the U.S.
Black-White IQ differences in the U.S.
* Group IQ differences have been studied extensively in the U.S.
* Random samples show mean IQ differences between Black
and White Americans of approximately 10-15 IQ points, despite both being exposed to Western culture
* After controlling for numerous demographic variables
including socioeconomic status, the size of this difference is generally reduced by a third
* Additional environmental factors would explain more of this IQ difference
* This IQ difference appears to keep shrinking with each
successive generation
* A smaller gap is associated with the current youngest
generation
The Black-White IQ difference is in decline
The Heritability of IQ and Group
Differences
* There is reason to believe that racial IQ differences may not
be subject to the same degree of genetic influence as IQ
scores
* The heritability of intelligence is lower among poor children
raised in the United States (h2 = 0.26) than among children
from wealthy families (h2 = 0.61)
* Based on these findings, we would expect a smaller
contribution of genes on the racial IQ gap (Tucker-Drob &
Bates, 2015)
Race vs. Ancestry
* “The racial groups used in the US — white, black, Hispanic,
Asian — are such a poor proxy for underlying genetic ancestry
that no self-respecting statistical geneticist would undertake a
study based only on self-identified racial category as a proxy
for genetic ancestry measured from DNA.” (Turkheimer,
Harden, & Nisbett, 2017)
Is there evidence that “black environments” generate lower IQs in US?
* Black and mixed race children raised by Black and white families.
* Observational studies show black parent interact less with the children and
tend to use a maintenance style of up bringing.
* Poor black children are spoken to less than white children
* White children hear 20,000,000 words by age 3; Black children hear
10,000,000 (Hart & Risley, 1995
Expectations and Threat
* Children given Raven’s told either it was an achievement test or just a survey
* Black did more poorly on the high threat condition.
* A similar finding arose when boys and girls were tested on math problems.
Black-White gap conclusions.
* There is a gap in IQ performance.
* It is reflected in and perpetuated through economic,
educational and social inequality.
* The proportion of this gap that is due to genetic
differences is likely negligible or very small
* The gap is in decline
Sex Differences
- Generally there is little evidence that boys and girls, or men and women differ in overall IQ.
- This is not because the tests were standardized to generate
this result.
– That is, the constructors did not choose and re-choose questions until there was no difference - However, there are still sex differences.
– Girls excel on some verbal abilities and boys on some spatial abilities.
Sex Differences in IQ Variability
* IQ among males is more variable than it is among females, hence there are more males at the positive and negative
tail ends of the IQ distribution
Verbal differences
* The general consensus is that girls do better than boys.
– The difference is not in vocabulary except at a young age
– It occurs in a wide range of tests including verbal fluency.
– It continues into adulthood.
Math
* Early research suggested boys were better than girls.
– But this difference has got smaller and has even perhaps
reversed. (see figures).
– There has been a serious attempt to change societal
expectations for girls.
– Females are negatively affected by stereotype threat on
mathematics exams
– Early 12-1 ratio in masters degrees in math for boys but
similar numbers of bachelors (1988).
– This ratio is declining, 3:1 recently (Spelke, 2005)
Recent Female advantage on Math reasoning
Differences between 18-year-old male and female scores on the three reasoning tests
from the DAT, in the 1980 and 1995 standardisation samples. (Data from Feingold, 1988,
Table 1, and Hyde and Trickey, 1995)
Worldwide Study Finds Few Gender Differences in
Math Abilities: Gender Gaps Linked to Status of Women,
According to New Analysis
WASHINGTON – Girls around the world are not worse at math than boys, even though
boys are more confident in their math abilities, and girls from countries where gender
equity is more prevalent are more likely to perform better on mathematics assessment
tests, according to a new analysis of international research.
“Stereotypes about female inferiority in mathematics are a distinct contrast to the
actual scientific data,” said Nicole Else-Quest, PhD, a psychology professor at Villanova
University, and lead author of the meta-analysis. “These results show that girls will
perform at the same level as the boys when they are given the right educational tools
and have visible female role models excelling in mathematics.”
Spatial abilities
* Men perform better on some subtests,
– Most obviously 3D MRT (three dimension mental rotation).
* Spatial reasoning differences are “real” in the
sense that they are reflected in differences in map reading, route finding, puzzle solving, and skills associated with a diverse array of disciplines such as engineering and
filmmaking
Men and women may use different
route finding strategies.
* Men are good at vector reasoning (sense of direction).
* Women tend to memorize landmarks better.
– A verbal strategy.
* This difference is reflected in the behavior of some animals.
* The evolutionary argument is that this has
evolved because humans are hunter gatherers and the men use this ability more.
– Some evidence supporting this in species differences in sex differences in rodents.
The Flynn Effect
Flynn effect – What accounts for it?
Environments have changed
* Nutrition
* Education
Tests have changed
Test sophistication
People have gotten used to taking tests
Therefore, there could be a practice effect
Evidence suggests practice effects on IQ tests may account for an
increase of up to 5 IQ points
For some authors, the Flynn effect is the strongest demonstration
that IQ tests are open to environmental change.
Applies to all regions of the world
* Is the Flynn effect different in developed countries as compared to
developing countries?
* Could the Flynn effect be slowing down?
Results:
Have gone up everywhere in the world
IQ Scores ↑, why?
* Tests have changed
* Environment has changed
- Demographics (e.g., decreasing family size)
- Nutrition
- Education
- People moving from working class occupations to “white
collar” occupations (unlikely given this would predict high gc
gains, when in reality we see higher gf
gains)
* Education
- Years of schooling has increased
- Recent changes in the type of schooling
- From rote learning to problem-oriented.
- Teaching more abstract, conceptual skills.
- Could explain improvements on Raven Matrices
* Multiplier effect?
- Many small causes combine to make a change that is greater
than their sum.
- E.g radio, TV, the Internet
Has moral progress contributed to IQ gains?
* Flynn suggests moral progress over the past hundred years
may have contributed to IQ gains.
* As he writes in your reading: “By 1900, the new scientific ethos had blind faith on the defensive. However, it did little
to banish the secular demons of racism and nationalism that culminated in the horrors of World War II and the holocaust.
We were still like domesticated animals. We had selected ourselves to resist violence within groups but not between
groups: we were happy to coerce “inferior” races or kill traditional enemies. Yet, over the last 70 years, these
demons have been on the defensive.”
Subtests most affected by the Flynn effect
* WAIS Subtests such as: Similarities, Block Design, Pictoral subtests, and Vocabulary most affected. Raven Matrices
highly effected.
* These subtests tend to be less (more you mean?) g loaded, indicating they reflect improvements in complex cognitive skills (e.g., thinking abstractly, categorizing, and language skills) more than “true intelligence”
(which is expected to be more strongly related to g)
- The idea that g and intelligence are related is supported by the
fact that tasks that are increasingly cognitively complex have
higher g loadings
Conclusions about Flynn Effect
* The Flynn effect is the strongest demonstration that IQ tests
are open to environmental change.
* Were people from the 1900s intellectually disabled?
- Their IQ scores were around 70 by modern standards, which is one criteria
used for diagnosing an intellectual disability
- No, they were not intellectually disabled – however, they did lack skills
related to abstract thinking, categorization, language skills, etc., and they
were not as good at taking tests
- Some of the Flynn effect may reflect an actual increase in intelligence,
however. This may be a response to demands within society for increasingly
complex cognitive skills (and therefore increased demands for education).
Compensatory education
- American Head start program
- Its “failure” used to argue that IQ is inherited and little affected by environment.
- Although expensive, the intervention was small in terms of hours in the program
- It does have a short term effects on IQ
- but this effect fades away
- Effective compensatory education programs can have a long term effect
on achievement and school performance - Garces et al 2002 – well controlled study (family background)
- Effective compensatory education programs also demonstrate how
education may have been a factor behind the Flynn effect.
An example of two effective
compensatory education programs
Abecedarian and Milwaukee
Effect of interventions on IQ compared to control groups
These were very intensive and long term interventions.
(read text cuz idgi)
Environment
* There have been studies showing that adopted parents’ SES has an
effect on the adopted children’s IQ
- But many suffer from selective placement etc
- A stronger French study (Duyme et al, 1999) looked at 4 year old children
who were removed from their homes and placed with adoptive parents of
different classes. So parents and children’s genes were not correlated. See
Table
Results:
- There was an effect of SES.
Pre adoption IQ (at 4.5 years old) vs. Post-adoption IQ (13.5 years old)
Low (N=24) Before: 78 After: 86
Middle(N=22) Before: 77 After: 92
High(N=19) Before: 79 After: 98
*see other chart cuz idgi
