What kinds of questions do IQ tests involve? The test is divided into several components, each asking questions that tap into a specific kind of thinking. In one part, children are asked to say what is missing from a picture (a door knob from a picture of a room that includes a door, for instance). Another component requires children to recall a string of numbers. Another involves moving around a collection of colored blocks so that they form a pattern presented on a card. Another asks children to answer questions about famous books, presidents, the weather, and different parts of the country. Although the test has been criticized for favoring children who live certain kinds of lives (if a child looks at a picture of someone riding a horse and doesn't know that the missing piece is the stirrup, it might well be because the child has lived her whole life in the inner city, with little access to books about riding). On the other hand, many of the questions test more content-free abilities, such as memory span. However, even these supposedly content-free questions might well favor children with certain kinds of experiences.
In the early 1960s, Sylvia Scribner and her colleagues set out to show the invisible bias in intelligence tests. Scribner was sure that something about school experience was helping some children do better even on parts of IQ tests that had been considered relatively culture-free. Her previous work in nonliterate communities had shown her that children who go to school regularly seem to acquire, without even realizing it, specific techniques that might help them do well on the kinds of memory tasks used in IQ tests. Sure enough, when she and her colleagues asked students to recall a fairly long list of words, such as apple, banana, desk, hat, plum, shoe, chair, and sweater, the children who had missed a lot of school days because of poverty, migrant work schedules, and segregation seemed to struggle. Scribner knew, from her literacy work, that learning to read leads people to conceptualize in a different way. Thus, the school children were using categories to chunk the items, making it easier to remember them: first the fruits, then all the furniture, then all the tools. Unschooled children didn't have this strategy at their disposal. However, when Scribner provided the category names ("Tell me all the kinds of fruit on the list, now all the pieces of furniture, now the clothing"), the children with little schooling performed similarly to the others. It seemed, then, that memory span per se did not differ between the two groups. Instead, what differed was the savvy to use category labels as a mnemonic, a skill found in school.
Scribner's research, which was really so simple, dealt a serious blow to the notion that any aspect of the IQ test, even the most seemingly culture-free part, was the same for all children. However, for all of its weaknesses and built-in biases, it taps into something pretty steady and real. But it's been hard for psychologists and lay people to put their finger on just what the test measured.