User talk:Maunus/RIdraft

Hunt 2011

 * 5 positions (Hunt & Carlson 2007) (p. 407)
 * All cognitive skills are ultimately determined by biology. However, it is important to distinguish between genetic contributions, which establish potentials, and life experieces, which determine the extent to which a potential is realized.(p. 408)
 * Racial and ethnic distinctions are both fuzzy in the sense just defined and inconsistent over time and place....Given that the answer to the question "who is what?" changes over place and time, it is no wonder that it is hard to find how a behavioral capacity, such as intelligence, correlates with race and ethnicity. Race and ethnicity are seldom causal variables in themselves. The only exceptions are in studies of the reactions of others to a person of a given race, and the studies of the effects of self-identity on people's conceptions of themselves.(p. 409)
 * Nyborg and Jensen's conclusions, and many similar conclusions about differences between racial/ethnic groups, rest on the validity of the method of correlated vectors (Dolan has shown this method to produce positive findings when measurement invariance does not hold, and that MGCFA should be used instead (p. 365)). Rushton and Jensen misstated Dolan's conclusions. Dolan using MGCFA concluded that Spearman's hypothesis lacks specificity and that it is very  difficult to distinguish between competing hypotheses concerning the latent sources of B-W differences. Rushton and Jensen say Spearman's hypothesis should be regarded as empirical fact. The technical arguments and the various demonstrations by Dolan adnd his colleagues show that this is not the case.(p. 415)
 * The drop in the educational gap over time has been mirrored by a similar gap between African Americans and Whites on cognitive tests less tied to the educational curriculum.(p. 417), Hedges and Nowell found that the reduction in the gap was due to a reduction in the portion of African American students in the lowest portion of the test scores. They did not find an increase in the proportion of African Americans in the highest ranks.(p. 418)
 * Test scores widen markedly over the school years. On average, Black students enter elementary school with fewer of the skills required to benefit from formal instruction than do White students.(p. 418) Phillips, Crouse & Ralph (1998) found that all students who are ill prepared fall behind regardless of race, although a small portion of the gap remained. Schooling increases the variation between those who were or were not initially well prepared for schooling.(p. 420)
 * It is posible that IQ tests are unfair in the sense that they test an ability that Whites have a more opportunity to develop than Blacks, but this is not an inequality of the test but of society. The very scarce studies of predictive accuracy show that tests accurately predict performance for all groups. (p. 424)
 * Social causes: Intelligence is defined by society - i.e. a post-industrial society values skills that allows succes in this society - this is a kind of cultural bias, but it is appropriate. Stereotype threat is demonstrated in experimental settings but not in field situations where success/failure has real consequences.(p. 426)
 * SES cannot be the sole explanation for the gap because SES measurements do not fully account for the differences, SES is itself confounded with intelligence, and SES is itself a statistical abstraction of several variables: parenting practices, Gap is greatly reduced and no longer statistically reliable when not just child's score is considered but also parent's scores - three best predictors was occupational status of household head, mother's verbal comprehension test score and nature of interaction with child. (yeung & Conley 2008)(p. 428)
 * Differences in attitudes towards education, Steinberg 1996, Ogbu, Cook & Ludwig showed black tenth graders expecting social costs to academic achievement.(p. 429)
 * In sum there are many social practices that have an impact on racial and ethnic differences in cognitive skills. They include parenting and cultural beliefs about educational processe. These in turn interact with issues of school financing and teacher behavior. ... I do not doubt, though, that the social environment of racial and ethnic groups has a major influence on the development of intelligence within those groups.(p. 430)
 * To sum up, there are powerful arguments and a good deal of evidence that racial/ethnic differences in intelligence are influenced by a variety of social variables.(p. 432)
 * Brain size is correlated to intelligence by .35 within the White population. Brain size is almost entirely genetically determined. Based on Rushton's (1992) calculations of racial differences in cranial capacity (proxy for brain size) there should be a .19 difference in achievement. One short paragraph. Lynn's nutrition hypothesis contradicts Rushton's brain size hypothesis since cranial capacity is highly dependent on nutrition. (p. 434)
 * "Admixture studies" by Rushton and Jensen - "ambiguous unless these variables are controlled" one short paragraph.
 * Rushton, Jensen, Lynn: their arguments not so much wrong as vastly overstated. (p 434)
 * R & J are correct that 100% environmental cannot be maintained.
 * 80% default hypothesis is extreme. Based on assumption that same factors that conribute to within group differences contribute to between group differences. (p. 435)
 * Gene environment interaction is complex. We do not know today which genes are associated with intelligence. Moving target: environment shifts quicker than gene frequencies.
 * Plausible cases can be made for both genetic and environmental cases. Evidence required to quantify relative contributions is lacking. (436)
 * Lynn and Vanhanen: Many of the samples are grossly unrepresentative (e.g. Equatorial Guinea datapoint was a school for the mentally disabled in Madrid), residual much greater for developng countries (Hunt & Wittmann 2008), Wicherts, Dolan & Van Maas found many more prolems of selectivity and scholarship, raised subsaharan IQ to 82 (the equivalent of Dutch after WWII), highly biased towards underestamiating the IQ of subsaharan Africa.(p. 439)
 * Rindermann: makes a compelling case that investment in education increases national cognitive skills, which in turn increase wealth, which allows for more investment in education.(pp. 440 - 443)
 * Lynn and Vanhanen's work has been widely publicized but, to my concern, too often accepted and too rarely carefully critiqued.(p 445)
 * Rindermann's analysis shows that they have badly undervalued the effects of investment in education (p. 445)
 * No genes related to the difference in cognitive skills across the various racial and ethnic groups have ever been discovered. The argument for genetic differences has been carried forward largely by circumstantial evidence.(p. 447)

Nisbett, Aronson, Blair, Dickens, Flynn, Halpern, Turkheimer (2013)

 * (a) Heritability of IQ varies significantly by social class. In low SES populations the variation is mostly explained by environment, in high SES families more of the variation is explained by genes. This is also borne out in studied cited by Rushton and Jensen as contradicting this, and also in intervention based studies. Ergo: High SES means that an individual is more likely to develop the full biological potential, low SES means that one is unlikely to reach one's full cognitive potential. Results of Asbury, Wachs & Plomin (2005) purported to show opposite result using telephone surveys. A study using the same data by Docherty, Kovas & Plomin (2011) was reanalyzed by (Hanscombe et al. 2012) and showed the expected result. This effect also means that adoption studies systematically overestimate heritability (Stoolmiller 1999 and others). A number of studies are thus biased by sampling only moderately high and high SES groups and hence excluding the populations where environmental effects would be strongest.(pp. 132-134)
 * (b) Almost no genetic polymorphisms have been discovered that are consistently associated with variation in IQ in the normal range.
 * (c) Much has been learned about the biological underpinnings of intelligence. A small number of GWAS studies have found only a handful of genes that are associated with intelligence variation, together they may account for less than 1% of variance in general cognitive ability. Likely intelligence is too polygenic to be feasibly studied in the foreseeable future. (p. 135) [ comment: none of the genes have been associated with specific continental populations]
 * (d) “Crystallized” and “fluid” IQ are quite different aspects of intelligence at both the behavioral and biological levels.
 * (e) The importance of the environment for IQ is established by the 12-point to 18-point increase in IQ when children are adopted from working-class to middle-class homes. We don't know that the home environment effect is due to environment or due to gene, but adoption studies of lower SES children to higher SES homes show that it does account for a substantial part of the IQ advantage of childen from privileged homes.
 * (f) Even when improvements in IQ produced by the most effective early childhood interventions fail to persist, there can be very marked effects on academic achievement and life outcomes. Adoption studies overestimate the genetic effects because of the SES self-selection effect.
 * (g) In most developed countries studied, gains on IQ tests have continued, and they are beginning in the developing world.
 * (h) Sex differences in aspects of intelligence are due partly to identifiable biological factors and partly to socialization factors.
 * (i) The IQ gap between Blacks and Whites has been reduced by 0.33 SD in recent years.


 * (a) the relationship between working memory and intelligence,
 * (b) the apparent contradiction between strong heritability effects on IQ and strong secular effects on IQ,
 * (c) whether a general intelligence factor could arise from initially largely independent cognitive skills,
 * (d) the relation between self-regulation and cognitive skills, and
 * (e) the effects of stress on intelligence.

Differences from Neisser et al. 1996:
 * (a) Due in part to imaging techniques, a great deal is now known about the biology of intelligence.
 * (b) Much more is known about the effects of environment on intelligence, and a great deal of that knowledge points toward assigning a larger role to the environment than did Neisser and colleagues and toward a more optimistic attitude about intervention possibilities.
 * (c) More is now known about the effects of genes on intelligence and on the interaction of genes and the environment. Our article also presents a wide range of new theoretical questions and reviews some attempted solutions to those questions.

Mackintosh 2011

 * Flynn effect provides some of the strongest evidence for a sizeable environmental effect on IQ scores. Its explanation is less certain: the best bet is that it relfects the operation of a large number of factors each of limited impact. That is probably just as true of environmental causes of differences in IQ within any one generation. NUmerous factors are correlated wth IQ differences: parental social class, behaviour and attitudes, prenatal and perinatal complications, nutritiion and health, family size and birth rder, amount of schooling. Some genetic effects on environmental factors (e.g. child's innate IQ affects parental behavior, parental SES status may be genetically based) cannot be ruled out. Behaviour geneticists think that family background has virtually no importance on children's IQ scores after age 18. The evidence for this is "less than overwhelming".


 * "Lynn and others who think like him have* a wholly unrealistic and empoverished conception of culture"(p. 329) [i.e. not only the test is un fair if one population is not able to acquire the knowledge needed, but cultures also value and inculcate different kinds of intelligence differently]


 * "The genetic hypothesis needs nothing more than an average difference in the distribution of the no doubt vast array of genes affecting IQ. ONe should not attempt to rule it out on a priori grounds."(p. 334)


 * Adoption studies: [about Weinberg et al - genetic and environmental factors were confounded (although black children appear not to have benefited form adoption in to white families at all, which is embarrassing to any environmental hypothesis)] "Two of the three studies provide no support for a genetic hypothesis at all [Eyferth 1961, Tizard 1974]. The third contains data that are certainly consistent with a partly genetic hypothesis [Minnesota trans=racial adoption study - Scarr 1976, Weinberg et al. 1992]. If it would be rash to argue that they refute the genetic hypothesis it would surely be absurd to argue, that taken as a whole, they support it. If one recalls the cautionary note sounded at the onset of this discussion, that in a racist society it may never be possible to bring up black and white children in truly comparable environments, the results of these studies are surely consistent with the possibility that if environmental differences between black and whites could be miraculously eliminated, the two groups might well obtain approximately equivalent IQ scores."(337)


 * Admixture studies: Quotes Nisbett 2005 p. 309 "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"(p. 338)


 * "Scores on Ravens Matrices are highly heritable, while the Wechsler similarities test is the second most heritable test of the WISC (Rushton 1995). But these are also the tests that have shown the largest gains over the past 50 years, in other words two tests that are most influenced by environmental differences between generations (Flynn 2007). As Flynn (2000) has pointed out, therefore if black-white differences are more pronounced on these tests than on others, this implies, according to Rushton and Jensen's argument, that black-white differences musy be environmental in origin."(pp. 338-39)


 * Reaction times: "It is surprising that Rushton and Jensen never comment on another finding, that there appears to be no difference between black and white children's scores on a variant of the infant habituation/dishabituation test (Fagan 2000), a test that correlates rather well with later IQ, and seems as likely as simply RT to be culture fair" "It is also hard to resist citing a study by Washburn and RUmbaugh (1997) who reported that rhesus monkeys had shorter reaction times than American college students. A different between two groups in average RT may not really prove that they differ in intelligence."(p. 239)


 * Environmental causes: infant morality is about twice as high for American Blacks than whites, low birthweight is twice as prevalent (correlated with IQ), white mothers are twice as likely to breast feed, brestfeeding being particularly correlated with IQ for low birthweight infants, single parent raise children are also disadvantaged with as much as 5-10 IQ pts (Flynn 2008) 63% of black children are raised in single parent families as opposed to 23% of whites (pp. 343-344)


 * "Most (perhaps all) of the genetic evidence reviewed above was consistent with a negligible genetic contribution, and there is rather good evidence that a variety of environmental factors have contributed to the difference in average test scores, which is, at last, now smaller than it was 50 years ago. ONe could reasonably defend Nisbett's argument that the gap was entirely environmental in origin. But it would probably be even more reasonable to acknowledge that the evidence is simply not sufficient is simply not sufficient to provide a definitive answer - and possibly never will be.(p. 344)


 * "There can be little doubt that stereotype threat contributes to blacks' poor performance on tests of cognitive ability;"(p. 348)


 * "IQ scores do not systematically underestimate the educational achievements of blacks or most other ethnic minorities. The only exception to this generalization is in the case of Americans of Chinese or Japanese ancestry."(p. 353)


 * "But it is, after all, entirely possible that educational performance is just as serious an underestimate of blacks' true level of intellectual performance as are the IQ tests themselves. This is hardly a fanciful suggestion: it is not difficult to think of reasons why black schoolchildren might not do as well at school as their abilities dictate. ... It follows therefore that actual educational attainment might well be an even more biased measure of black children's intellectual competence than an IQ test."(pp. 353-54)


 * "While it is certainly true that AFican peasants and Australian aboriginals, for example, typically obtain substantially lower test scores than white Europeans or North Americans, it requires an astonishing insensitivity to the possibility of cultural differences to conclude that this establishs that they are less intelligent."(p. 357)


 * "In spite of claims to the contrary, there is remarkably little evidence that the difference is genetic in origin: when brought up in relatively comparable environments black and white children usually obtain relatively similar test scores; and studies of the degree of white ancestry in American blacks suggest that this has little or no impact on test scores. By contrast, there is quite good evidence that several environmental factors, prevalence of low birth weight, breast-feeding, and especially style of parental interaction, do contribute to the difference in test scores.(p. 358)


 * "When one turns to comparisons between different nationalities, North American whites against Australian aboriginals or illiterate peasants in sub-Saharan Africa, these problems surely, become insuperable, and render such comparisons completely worthless, and probably pernicious"(p. 359)