286 CHAPTER 7
That’s right—think about the definition of intelligence for a moment: the ability to
learn from one’s experiences, acquire knowledge, and use resources effectively in adapt-
ing to new situations or solving problems. How can anyone define what “effective use
of resources” might be? Does everyone have access to the same resources? Is everyone’s
“world” necessarily perceived as being the same? Intelligence tests are useful measuring
devices but should not necessarily be assumed to be measures of all types of intelligent
behavior, or even good measures for all groups of people, as the next section discusses.
IQ TESTS AND CULTURAL BIAS The problem with trying to measure intelligence
with a test that is based on an understanding of the world and its resources is that
not everyone comes from the same “world.” People raised in a different culture, or
even a different economic situation, from the one in which the designer of an IQ test
is raised are not likely to perform well on such a test—not to mention the difficul-
ties of taking a test that is written in an unfamiliar language or dialect. In the early
days of immigration, people from non–English-speaking countries would score very
poorly on intelligence tests, in some cases being denied entry to the United States on
the basis of such tests (Allen, 2006).
It is very difficult to design an intelligence test that is completely free of cultural
bias, a term referring to the tendency of IQ tests to reflect, in language, dialect, and con-
tent, the culture of the person or persons who designed the test. A person who comes
from the same culture (or even socioeconomic background) as the test designer may
have an unfair advantage over a person who is from a different cultural or socioeconomic
background (Helms, 1992). If people raised in an Asian culture are given a test designed
within a traditional Western culture, many items on the test might make no sense to them.
For example, one kind of question might be: Which one of the five is least like the other four?
DOG—CAR—CAT—BIRD—FISH
The answer is supposed to be “car,” which is the only one of the five that is not
alive. But a Japanese child, living in a culture that relies on the sea for so much of its food
and culture, might choose “fish,” because none of the others are found in the ocean. That
child’s test score would be lower but not because the child is not intelligent.
In 1971, Adrian Dove designed an intelligence test to highlight the problem of cul-
tural bias. Dove, an African-American sociologist, created the Dove Counterbalance Gen-
eral Intelligence Test in an attempt to demonstrate that a significant language/dialect
barrier exists among children of different backgrounds. Questions on this test were derived
from African-American culture in the southeastern United States during the 1960s and
1970s. Anyone not knowledgeable of this culture will probably score very poorly on this
test, including African-American people from other geographical regions. The point is sim-
ply this: Tests are created by people from a particular culture and background. Questions
and answers that test creators might think are common knowledge may relate to their own
experiences and not to people of other cultures, backgrounds, or socioeconomic levels.
Attempts have been made to create intelligence tests that are as free of cultural influ-
ences as is humanly possible. Many test designers have come to the conclusion that it may
be impossible to create a test that is completely free of cultural bias (Carpenter et al., 1990).
Instead, they are striving to create tests that are at least culturally fair. These tests use ques-
tions that do not create a disadvantage for people whose culture differs from that of the
majority. Many items on a “culture-fair” test require the use of nonverbal abilities, such as
rotating objects, rather than items about verbal knowledge that might be culturally spe-
cific. One example is Raven’s Progressive Matrices, a test of abstract reasoning. The test
consists of a series of items containing abstract patterns, either in a 2 * 2 or 2 * 3 matrix,
from which test takers have to identify a missing portion that best completes a pattern
(see Figure 7. 6 ). However, although once believed to be largely culture free, or at least fair,
even this test is not immune to the influence of culture, as age, generational cohort, and
education appear to impact performance (Brouwers et al., 2009; Fox & Mitchum, 2013).How might these two women, apparently from
different cultures, come to an agreement on what
best defines intelligence?