There are a couple of different groups we can examine:

• Men and Women differences are generally seen in 1) spatial and verbal differences, 2) gender treatment, 3) hormonal and brain development. More recent studies, however, are showing many of these differences are disappearing. The sex differences that continue to show up include:

  • abstract/mathematical and verbal differences (we should note the mathematical differences don't show up until we get to the level of Albert Einstein kinds of math though)

  • gender treatment (e.g., choice of careers counselors stress, and a talking Barbie who says "Math class is hard. Do you want to braid my hair?")

  • hormonal differences and brain development (e.g., differences in hemisphere sizes, testosterone and spatial relations), and some even include iron deficiency (malnutrition) which women are ten times more likely than men to experience

  • different patterns of decline, with fluid intelligence declining earlier for women, crystalized intelligence declining earlier for men

  • different patterns of recovery, with women recovering from brain injuries faster in terms of oral expression, but not comprehension

• Asian Americans often score higher on IQ tests due to schooling differences, social factors like motivation and high value placed on learning, as well as temperamental differences. Social values such as respect for others, hard work, and work with the family to support the family can effect achievement. Some studies show the same IQ as Caucasians, but higher academic achievement compared to their IQ scores. Other studies do show higher IQ scores for Asian Americans.

• Hispanic Americans often score lower due to language differences. Often their family's reasons for coming to the country effects their scores, especially when emotional trauma's underlie the family's move. Large splits between their verbal and nonverbal scores are seen due to language differences. They generally fall between African Americans and whites, and younger children might seem delayed if you rigidly use some chart of developmental stages. Hispanic families often do not push children to develop as quickly as Caucasian families.

• Native Americans - first realize that this term covers over 200 different cultures and languages. However, some studies show that test setting can make a big difference, as well as undiagnosed ear infections. Native Americans may prefer to work in situations where collaboration is required, as opposed to individualized work, and so examiner pressure to find the "right" answer "on your own" and "as quickly as you can" may not mean much.

• African Americans often score lower than Caucasians on IQ tests; the differences range from 2-3 to 15 points lower. The difference in these estimates seems to stem from SES. Studies that control SES got smaller numbers, while studies that did not got larger differences. We can probably rule out decreased familiarity with the test content, as the WAIS and WISC development included a stage of giving all items to different ethnic groups, and throwing out the ones that distinguished between ethnic groups compared to Caucasians. A larger family size might be related too, as noted above, but there's some doubt as to whether African Americans actually have more children on average than Caucasians. IQ scores for African Americans are still accurate, however, as a prediction of years in school and school grades. Steele offers that "effort optimism" and effects of discrimination make a difference, making African Americans feel that testing will reflect poorly on them, which decreases their efforts.

You might find this amusing…. Another intelligence in the history of our field is the BITCH (The Black Intelligence Test of Cultural Homogeneity or sometimes the The Black Intelligence Test Crossreferenced for Honkies) by Williams, 1972. It is comprised of 100 items that lower SES black children could answer, but middle class white children could not. For example:

Boot is
• a) cotton farmer,
• b) black,
• c) Indian,
• d) Vietnamese

Clean is
• a) just out of the bathtub,
• b) well-dressed,
• c) very religious,
• d) someone who has a lot

I note it here because it shows that some black children could learn a variety of cultural terms and vocabulary, as well as their meaning in more than one context, but this likely was not clear from their SB or WISC scores. While you might be hoping to learn this test in this class…

• it was only normed on 200 children,

• it hasn't really been shown to correlate with anything stronger than familiarity and exposure to the black culture,

• MR children scored lower on the BITCH than on the WISC,

• the WISC is a better predictor of achievement test scores as well as grades

Another interesting test is the Leiter, a totally non-verbal test. It involves sorting objects, matching colors, copying designs… based on modeling by the examiner. The Leiter is nice and has been revised twice since 1948 (1979 and 1997 I think). Some have criticized it as not really being any more culture free than the SB4, and as a poor predictor of school performance.

Thus, a "culture free" test is very hard to make. And the question remains, "Would we want it if we could make one?" What do you think?

On these issues though… Suzuki and Valencia warn us that within group differences are often bigger than between group differences when we study ethnic differences. They site a study by Scarr and Weinberg from 1978 which studied 176 adopted black children who had been adopted into white homes. The white parents' IQs averaged 119, their biological children's IQs averaged 116.7, and their adopted African American children's IQs averaged 110. This was 15 points higher than the control sample of black biological children in black biological homes.