g and group differences

The size of the average Black–White (B–W) difference (expressed in standardized units) on various psychometric tests varies as a direct function of the tests’ differing g loadings. In other words, the B–W difference on psychometric tests (and their many external correlates as well) is mainly a function of g, so that the more a test measures g, the more it discriminates between unselected groups of Blacks and Whites. No other feature of tests is as highly correlated with the variable size of the mean B–W difference on various tests. Nor are socioeconomic status or other social background factors as sharply predictive of B–W differences on a given test as is the g factor. Because g is the primary effective factor both in the practical validity of tests and in the magnitude of the B–W difference in unselected groups, the conjunction of these two effects is the unavoidable cause of adverse impact when g-loaded tests are used in selection. This is a more serious matter than if the tests were merely biased in cultural content: first, because it is not possible to rid cognitive tests of g and still have them remain valid for any practical purpose; and second, because individual differences in the level of g resist intentional change.

–Jensen, 2000. The Dilemma of group differences: Testing: The Dilemma of Group Differences

Average differences in g as of the 90’s, d = .8 to 1.1 (age >16)

Academic

Industrial and Military

Differences in Job performance (2), d = .27 as of the 2000’s. See also (12).

Notice, objective tests are significantly more “biased”:

Differences on work sample test (i.e. job simulation exercises) as of the 2000’s (3), d = varying. See also (12).

The difference is larger in cognitively loaded tests:

Differences in trainability as of the 2000’s (4), d = varying

The difference is larger for cognitively loaded skills:

Differences on Academic Achievement tests (grades k-12) as of the 2000’s (5), d (averaged) = .7 to .8

Differences in IQ across all ages (6) as of the 2000’s, d (averaged) = 0.9; d (cumulative across age) = .95

Differences on various non-IQ psychological tests as of the 2000’s, d = varying

Differences in pyschomotor ability (7), d = – .7

Different rates in mild mental retardation (9), d = .66 SD (see: 10)

Different rates in Giftedness (11)

Implications of differences, were the US to embrace meritocratic equal opportunity, abolish ethnoracial quotas, and use standard forms of assessment (8):

References

(1) Roth et al., 2001. Ethnic group differences in cognitive ability and educational setting: A meta-analysis

(2) Roth, Bobko, and Huffcut, 2003. Ethnic Group Differences in Measures of Job Performance: A New Meta-Analysis

(3) Roth, et al., 2008. Work Samples tests in personal selection: A meta-analysis of Black-White differences in overall and exercise scores

(4) Roth, 2010. Updating the trainability tests literature on black-white subgroup differences and reconsidering criterion-related validity.

(5) Sackett and Shen, 2008. Subgroup differences on Cognitive tests in contests other than personal selection

(6) Dickens and Flynn, 2006. Black Americans reduce the racial IQ gap; Gottfredson, 2005. Implications of cognitive differences for schooling within diverse societies

(7) Ployhart, 2008. THE DIVERSITY–VALIDITY DILEMMA: STRATEGIES FOR REDUCING RACIOETHNIC AND SEX SUBGROUP DIFFERENCES AND ADVERSE IMPACT IN SELECTION

(8) Gottfredson, 2006. Social consequences of group differences in cognitive ability

(9) Chapman, et al., 2008. Public Health Approach to the Study of Mental Retardation

(10) Plomin and Spinath, 2004. Intelligence: Genetics, Genes, and Genomics

Although no twin studies of severe mental retardation have been reported, an interesting sibling study shows no familial resemblance. In a study of over 17,000 children, 0.5% were moderately to severely retarded (Nichols, 1984). As shown in Figure 4 (dotted line), siblings of these retarded children were not retarded. The siblings’ average IQ was 103, with a range of 85 to 125. In other words, moderate to severe mental retardation showed no familial resemblance, a finding implying that mental retardation is not heritable. In contrast, siblings of mildly retarded children (1.2% of the sample) tend to have lower than average IQ scores (see Figure 4, solid line). The average IQ for these siblings of mildly retarded children was only 85. Similar findings—that MMR is familial but moderate and severe retardation are not familial—also emerged from the largest family study of MMR, which considered 80,000 relatives of 289 mentally retarded individuals (Reed & Reed, 1965).

(11) Yoon and Gentry, 2009. Racial and Ethnic Representation in Gifted Programs: Current Status of and Implications for Gifted Asian American Students.

(12) Roth et al., 2011. TOWARD BETTER META-ANALYTIC MATRICES: HOW INPUT VALUES CAN AFFECT RESEARCH CONCLUSIONS IN HUMAN RESOURCE MANAGEMENT SIMULATIONS