part 2, https://www.youtube.com/watch?NR=1&feature=endscreen&v=U8rhIZJAdd0
if intelligence is innate and unchangeable, an educational system looks very different from if it is socially determined
different societies have come up with different answers at times
policies are often inconsistent
intelligence is an objectively measurable quantity
it correlates highly with with social and economic success, in particular for the most important jobs
intelligence is highly heritable
intelligence cannot be strongly influenced by environmental factors
"race" (according to US statistics)
average group IQs for four major races: 85, 89, 103 106
Well known result, but the book still caused quite a stir because it talked about implications.
Given high correlation with salaries, educational success, health, etc. IQ may be a significant factor explaining group differences in income etc.
APA = American Psychological Association, representing mainstream US psychology
about IQ, IQ scores have high predictive validity for individual differences in school achievement., IQ scores have predictive validity for adult occupational status, even when variables such as education and family background have been statistically controlled., There is little evidence to show that childhood diet influences intelligence except in cases of severe malnutrition.
about race, There is certainly no such support for a genetic interpretation... . It is sometimes suggested that the Black/ White differential in psychometric intelligence is partly due to genetic differences (Jensen, 1972). There is not much direct evidence on this point, but what little there is fails to support the genetic hypothesis., The differential between the mean intelligence test scores of Blacks and Whites (about one standard deviation, although it may be diminishing) does not result from any obvious biases in test construction and administration, nor does it simply reflect differences in socio-economic status. Explanations based on factors of caste and culture may be appropriate, but so far have little direct empirical support. There is certainly no such support for a genetic interpretation. At present, no one knows what causes this differential.
The Mismeasure of Man
A strong argument against biological determinism, i.e., that biological differences account for differences in class, social mobility, etc.
Updated in its second edition to respond to the Bell Curve.
Hypothesis 1: differences in intelligence scores between racial groups in the US are caused by non-hereditary factors, such as economic and educational opportunities etc., The degree to which this hypothesis can be true is limited by many studies on the influence of training on intelligence scores., This is likely to be more important when looking at countries where access to nutrition and health care may be a limiting factor., Effectively, the argument made by Gould.
Hypothesis 2: race is causally related to differences in intelligence scores, This idea has currency in many racist ideologies, but is untenable from a biological point of view., "Race" is not a biologically well-defined category, so it can't cause anything., This hypothesis is patently false because its premises are ill-defined.
Hypotheis 3: race is a marker that happens to be associated with differences in intelligence scores in US populations, but is otherwise unrelated, This would be consistent with existing data. There is no specific evidence for or against it.
Hypothesis 4: intelligence scores are not meaningful; intelligence is not reliably measurable or quantifiable and/or there is some other methodological error, This is always a possibility in science, but there are no clear proposals for what these errors might be., Another argument made by Gould.
Summary, There is a real and strong correlation that needs to be explained., The source of the correlation remains unknown and controversial., What do you think? Alternative hypotheses?
correlations at the group level are much weaker than individual variation, therefore you cannot infer individual intelligence from group membership
to the degree that it is true, heritability might affect the interpretation of statistical data on integenerational social mobility, but authors like Gould dispute that as well
How did assumptions about intelligence and race influence Nazi ideologies?
How do assumptions about intelligence influence contemporary German politics?
Has "German culture" or "cultural heritage" ("Erbe", inheritance) become a proxy for "German race" in German political discussions?
heredity of intelligence is a controversial but politically and socially important topic
political and social context determines how we think about the nature of intelligence: "tabula rasa" general purpose machinery, or heritable and predetermined machinery
scientific data supports egalitarian and race-blind policies, but does not necessarily support the "tabula rasa" assumptions that they are commonly justified with
term coined by Francis Galton
Galton, 1822-1911 (UK) King's College, Cambridge, invented correlation, inventor of fingerprint classification, psychometrics, no statistical effect of prayer, inventor of meteorology
take individuals with different genetic relatedness
measure their IQ
determing the correlation coefficient
.95 same person (tested twice)
.86 identical twins reared together
.76 identical twins reared apart
.55 fraternal twins reared together
.47 biological siblings reared together
.42 parent-child living together
.35 fraternal twins reared apart
.30 unrelated children reared together
.24 biological siblings reared apart
.22 parent-child living apart
.19 adoptive parent-child living together
Source: Kaufman: IQ Testing 101
depends on definition
IQ is defined as innate general intelligence, training can't improve "IQ", but it may improve general performance; to the degree that it does, the IQ tests are measuring the wrong thing
IQ is defined as general intelligence without requiring that it is innate, training can improve "IQ"; care must be taken that improvements on only a few specific tests are not mistaken as an improvement in general IQ, in particular... IQ tests are constructed for average participants; improving scores on only one or two skills doesn't improve "general intelligence", but it will improve IQ scores without recalibration, there is some evidence that training may be able to improve "general intelligence" somewhat
IQ is defined as that which IQ test measure, with this definition, any kind of intervention that raises IQ test scores "raises IQ"
innate, general IQ is correlated with prestige, high pay, and health
BUT ... those correlations have been established only for innate, general IQ without training
there is no evidence for (or against) the idea that increasing IQ scores will give you the same benefits as innate general IQ
height, height and growth are heritable, being taller amounts to higher earnings (about $3200/year for every 10cm)
difference, "Asians are shorter on average" is just considered a factual statement., "_______ are less intelligent on average" is considered offensive
intervention, why don't we give everybody lifters?, (Turns out the real factor is height at age 16, so lifters don't help. It may be the same with IQ.)
(talk about this more below)
Eugenics movement, Eugenics tried to improve human societies by preventing people considered from reproducing, usually by sterilization, but in some cases by extermination., problems with this view, There is no obvious reason that more intelligence is automatically better, either for the individual or for society., Humans traits are tradeoffs, just like any organism or even any device.
colonialism, Colonialism was justified by various ideas related to intelligence and souls: either native peoples were said not to have souls, or they were considered of inferior intelligence and in need of European (or other empire) guidance.
religious views, In the 16th century, both the Catholic church and Luther considered mentally retarded children "masses of flesh without souls" and recommended they be drowned.
The Age of Enlightenment and philosophers like John Locke.
"All men are created equal"
protect humans regardless of abilities / disabilities
assume that intelligence is independent of race
"intelligence" and "reason" used to be the defining attribute of being human
"sapience", "morality", and "compassion" are the defining attributes now
the confusion about this is not unique to AI or Turing tests, and many of the discussions surrounding AI are related to these political and historical discussions