Scholars Develop Intelligence Test, First Since 30's
Washington--Using a definition of intelligence markedly different from that underlying most intelligence tests, two California psychologists last week introduced a new test--the first in nearly a generation--that they say counters many of the problems identified in existing intelligence tests.
Called the Kaufman Assessment Battery for Children, the test was developed over a five-year period by Alan S. Kaufman and Nadeen L. Kaufman, both of whom worked previously on the revised Wechsler Intelligence Scale for Children (wisc-r), the most commonly used I.Q. test. The researchers, who are regarded as among the foremost experts on intelligence testing, described the test at the annual convention of the American Psychological Association (apa), held here last week.
Like the wisc-r and the Stanford-Binet I.Q. test, the Kaufmans' test is intended for use as a diagnostic tool. Approximately 20-to-30 percent of all schoolchildren are given such individual intelligence tests. The results are generally used to plan special-education programs or identify gifted children, said Mr. Kaufman.
It was those goals, in fact, rather than the aim of ranking people by intelligence that led to the development of the first I.Q. test in 1904. The test was the work of the French psychologist Alfred Binet who was requested by the minister of public education to find ways to identify children who needed special help with their schoolwork.
Like the Kaufmans' test, Mr. Binet's was not intended as a measure of inborn intelligence.
The Kaufmans' test differs from the wisc-r and the Stanford-Binet I.Q. test, which is used less frequently, in other important respects as well, the researchers said.
By defining "intelligence" as the ability to process information and to solve problems, the test allows psychologists to distinguish between a child's "present intellectual functioning" and his or her acquired knowledge. Other intelligence tests do not make this distinction and have been criticized by some educators, who believe they are largely measures of achievement.
The test's design and content are based on recent neuropsychological research, making it the only intelligence test that has a "sound theoretical foundation." The other tests, developed in the late 1930's, are not based on any scientific theory, according to Mr. Kaufman.
By separating the "mental processing" score from the achievement score, the test is less likely to discriminate against culturally disadvantaged students and minorities.
The test scores are intended not only to help educators place children, but also to help them determine the best way to teach children with varying styles of learning.
The researchers used 1980 census data to establish a "norm--" the yardstick against which all scores will be measured. The other tests use 1970 data, they say, and hence do not take into account the many demographic changes that have occurred in the past 10 years.
Working with other psychologists around the country, the Kaufmans have conducted nearly 40 studies of the test's validity, in which they have addressed questions of racial, sexual, and other forms of bias. The results of all of these studies will be available when the test is published in March 1983, the researchers said.
'Novel, Game-Like Tasks'
The test consists of 16 "novel, game-like tasks," and is intended for use with children between the ages of 2 and 12. The tasks include activities such as picture identification. Some of the tasks were developed by the Kaufmans; others were adapted from existing intelligence and achievement tests.
The test contains separate scales for measuring two styles of information processing--"simultaneous" and "sequential"--and another scale that measures achievement. This distinction, they said, represents a major step forward in intelligence testing.
In sequential processing, "the child must solve problems by rearranging the stimuli presented so they are in sequential order," the test's developers explained. In simultaneous processing, "the child must simultaneously integrate and synthesize the information in order to solve each problem."
Children--and adults--may favor one style over the other, they noted. For example, a person who draws a map when asked for directions is using simultaneous processing, Mr. Kaufman said. A person who gives directions--"Go two blocks, then turn right ..."--is using sequential processing.
The scores on the information-processing scales should show educators not only how a child handles new tasks, but also what is the best method with which to teach him. For example, Mr. Kaufman said, some methods of teaching reading require better sequential abilities, while others require the child to process information simultaneously.
'Styles of Problem Solving'
"The bottom line is to focus on the two different styles of problem solving," Mr. Kaufman said. "Children differ in their abilities.
"From that general comparison, you can develop a program that emphasizes the child's strengths. What you do is capitalize on that child's better process. This is the goal. We wanted our test to be an active test."
Mr. Kaufman said the test scores reveal how well a child solves "new problems that do not depend on a lot of school-acquired skills?"
By using this information-processing model, Mr. Kaufman said, "We're able to use [test] tasks that are more fair to individuals from different backgrounds."
The test is designed to counter cultural or linguistic bias, Mr. Kaufman said. In the first several items on the intelligence test, he noted, the person administering the test may explain the question to the child--in a language other than English or using gestures, if necessary. "Nonverbal" subtests also permit the testing of hearing-impaired children, Mr. Kaufman said. Although other intelligence tests can be translated into sign language, he added, the sign-language versions are not "normed."
Other individual intelligence tests do not permit such "teaching items," Mr. Kaufman said. As a result, some children do poorly because they do not understand what is expected of them.
The test includes a section that measures the child's achievement in school--a perfectly valid thing to measure, but one that should not be confused with problem-solving ability, Mr. Kaufman said. "If you want to predict school achievement, give an achievement test," he said. Acquired information is important, he said, "but let's not call a child dumb because he hasn't acquired it."
Scores on the mental-processing and achievement scales may also be compared to one another. Such a comparison provides a basis for comparing a child's intellectual ability with his or her academically acquired skills, according to Gary Robertson, director of the test division of the American Guidance Service, a Minnesota-based publisher that will publish the test and an accompanying testing manual in March 1983.