The originators of intelligence tests believed that they were assessing general intellectual competency that predicted success in life. Research has revealed, however, there are other factors not related to ability that may influence adolescents’ performance on intelligence tests.

Socioeconomic status (SES)

Children who are raised in poverty are severely limited in their intellectual potential by their environment (Turkheimer, Haley, Waldron, D’Onofrio, & Gottesman, 2003). Adolescents who come from low-income homes score as much as 15 to 20 points below their middle-class peers (Neisser et al., 1996). The reasons suggested are that poor children and adolescents may lack the material (e.g., books, music lessons, computers) and familial support (e.g., help from parents on homework, engagement in school-related activities) that promote intellectual development. Low-income adolescents are also more likely to live in neighborhoods that lack extracurricular opportunities (e.g., parks, recreational facilities) and contain poor-quality schools. Furthermore, adoption studies using a behavior-genetic framework have shown that when low-income children are adopted into middle-class families, their IQ scores rise.


African American youth score, on average, 15 points below Caucasian youth. Hispanic youth fall midway between these two groups, and Asian Americans score slightly higher than Caucasians (Ceci, Rosenblum, & Kumpf, 1998). Part of this pattern may be explained by the large percentage of African American youth who live in poverty—42%, compared to 22% of all American children. Studies that equalize SES show differences in IQ scores, with Caucasians scoring higher than African Americans (Jensen & Reynolds, 1982).

These findings are not controversial in themselves; it is how they have been interpreted that has raised the concern of many social scientists. Controversies have stemmed from the claims some researchers have made about the origins of intelligence. They have suggested that intelligence is largely biologically determined and therefore not amenable to improvement (Jensen, 2001). Opponents of this view argue that scores on intelligence tests can be influenced by many nonintellectual factors, such as years of schooling, SES, and familiarity with the culture for whom the test was written. Researchers have in fact shown that when any one of these factors is manipulated, IQ changes (Grissmer, Williamson, Kirby, & Berends, 1998).