Intelligence Testing in the Schools

A Brief History and Critique

Thomos J. Kampwirth Ph. D.

Professor of Educational Psychology

California State University, Long Beach

History

Intelligence tests have been used in the public schools of America since just after the turn of the present century. The first major test, the Stanford-Binet, wa brought to this country from France, having been invented there by Alfred Binet, a French scientist charged by his government to develop an instrument for the measurement of abilities in children so that school officials could separate true slow-learners from malingerers.

Binet’s test has been modified and updated a number of times since is beginning, most recently as a fourth edition in 1985. During the period of time between 1920 and 1950 numerous other tests were developed and marketed, but none quite reached the popularity of the Stanford-Binet until David Wechsler developed his tests, starting in the 1930’s with the Wechsler-Bellevue Intelligence Scale, since modified to the Wechsler Adult Intelligence Scale (WAIS, 1955) and most recently updated as the WAIS-R(1981).

In 1949, Wechsler developed a downward extension of his original test, designed for children between 6 and 16, and called it the Wechsler Intelligence Scale for Children. This test has most recently been updated in 1974 as the WISC-R.

Wechsler’s most recent entry in the intelligence measurement arena was another downward extension of the WAIS and WISC, one suitable for children between four and six and one-half, called the Wechsler Preschool and Primary Scale of Intelligence (WPPSI), which was published in 1967.

Since Wechsler’s efforts from the 40’s through the 70′ s there have been only a few other tests that have reached the prominence of these scales and the Stanford-Binet. These are the Woodcock-Johnson Cognitive Battery (1977) the Kaufman ABC (Assessment Battery for Children; 1983) and most recently, the Detroit Tests of Learning Ability, Revised (1985).

Individual intelligence testing of school-aged children is dominated be these five instruments, roughly in this order of popularity: WISC-R, K-ABC, Woodcock-Johnson, Stanford-Binet IV, and DTLA-R. Other tests that comprise a second level of popularity include the McCorthy Scales of Children’s Abilities, the Peabody Picture Vocabulary Test, the Slosson Intelligence Test, Columbia Mental Maturity Test, the Leiter International Performance Scale, and the Pictorial Test of Intelligence.

These second level tests are used as back-up instruments to gather additional information, or with children who are unable to take one of the top five tests listed above. NATURE OF INTELLIGENCE TESTING The most common format for intelligence testing with children is for a trained examiner to sit at a table across from a child and to ask questions designed to determine the child’s fund of information, short-term memory, reasoning ability (usually verbal and mathematical), and general “common sense”.

Most tests also include puzzles and other tasks that require visual-spatial skills, such as the copying of patterns or designs, or the detection of missing parts of objects, or the ability to do analogies presented visually.

Different tests yield different types of scores, though generally one con get a final score which is essentially on “I. Q” *these I. Q’s are standard scores that have a mean of 100 and standard deviation of 15 points. Use of this metric system allows us to communicate about results in a convenient, readily understood way.

If a child gets an I. Q of 100 we know that no matter how old the child is, he or she has earned a score that is exactly at the average for his or her age. We also know that about 50# of all children will earn scores between 90 and 110, a band of scores generally regarded as being average.

Finally, we know from this system that children who score 2 or more standard deviations below the mean (i. e,I.Q. 70 or below) are generally “retarded” for school learning purposes, while those 2 or more standard deviations above the mean (i.e. I.Q. 130 or above) are usually “gifted” learners.

Individual intelligence tests are not routinely given to school children. In fact, only about 10 to 12 percent of children ever take them, and these are almost exclusively children who have been referred to school psychologists (or private-practice psychologists) because of a learning problem or because they might be eligible for a program for gifted children.

Other “normal” children are not given individual I. Q. tests for two reasons: it probably wouldn’t make any difference in their schooling, especially given the grade-dominated, large-group type of educational service-delivery method prevalent in America: and, it is costly to administer them.

DIRECT AND INDIRECT COSTS OF INTELLIGENCE TESTING

In the public schools of California, the average salary for a school psychologist is approximately $30, 000 per 10 month school year. This is about $170. 00 per day, or $24. 00 per hour for a seven-hour school day. The average individual intelligence test takes about one and a half hours to administer, and another one and a half hours to score and write the report that accompanies it. If a second-level intelligence test is added to the battery, as it is in about half the cases, then the original three hour investment becomes more like five to six hours depending on what test were used, the speed at which the child responded, etc. Therefore the direct costs related to individual intelligence tests will range from about $75. 00 to $150. 00 per child. If a child is placed in a special education program, that child might be reassessed with an intelligence test three or four times in the course of their K-12 education. Costs in this case would run up to about $600. 00.

In addition to these direct costs for psychologists’ salaries, there is the cost of the tests themselves (ranging from about $100. 00 to $600. 00), test forms, secretarial time, computer scoring and interpretation software, and non-salary compensation (e. g, insurance and retirement packages) paid by the district.

Indirect costs would include the expense incurred in training programs (most programs require a least one full course devoted to intelligence tests), which are largely at tax-supported public universities, and the loss of other possible services that school psychologists could be providing if they are not giving individual intelligence tests. In regard to this latter point, it is a very common complaint among school psychologists that they are required to spend an inordinate amount of time in individual assessment of children, time which would be more productively used in counseling and consultation areas.

THE USES OF INTELLIGENCE TESTS

Any instrument(s) that cost the taxpayers hundreds of dollars to give (based on the direct and indirect costs mentioned above) should have very specific uses or results which are essential, valid, and cannot be obtained in other ways. To the extent that individual intelligence tests meet these criteria, they probably should be retained; to the extent that they fail to meet these criteria, their use should be questioned. Essential: the only reason for which an intelligence test is essential is to meet a legal mandate for its use.

Federal law (P. L 94-142) requires that chil-dren being considered for special education (especially those who may be retarded or learning disabled) be evaluated to determine their level of mental functioning. The law does not, however, require the use of formal, norm-based, intelligence tests. Their use remains common, however, because of the general belief that they serve well the requirement to find a child’s level of mental functioning.

In spite of their acknowledged problems (see next section: Validity) And the costs discussed above, they remain the method of choice is almost 100% of the school districts around the country. Interestingly, however, this almost exclusive use of intelligence tests to meet the mandates of the law is being reconsidered in some districts. In California, the San Francisco Unified School District stopped using these tests about five years ago, and about one year ago, their use was terminated in both Los Angeles and Long Beach Unified School Districts.

The decision to stop using “I. Q. ” tests in favor of “alternative means of assessment” (which are still being formulated) was largely the result of court cases (specifically, Larry P. Vs. Wilson Riles), and validity concerns mentioned below, especially those related to minority and bias issues.

In summary, the use of individual intelligence tests in not essential for legal purposes; all that is necessary is a valid way of measuring mental ability. Validity: the concept of test validity has to do, essentially, with whether or not a test measures what it purports to measure. In order to determine this, one needs an accepted definition of a concept (e. g, intelligence) and some evidence that any given test satisfies the definition, and does so reliably.

Definitions of intelligence include the following: “goal-directed adaptive behavior” (Sternberg, 1982), “total information-processing capacity of the organism” (Jerison, 1982), “the overall capacity of an individual to understand and cope with the world around him” (Wechsler, 1974), “to judge well, to comprehend well, to reason well, those are the essential activities of intelligence” (Binet, 1916), and “the resultant of the processes of acquiring, storing in memory, retrieving, combining, comparing, and using in new contexts information and conceptual skills: it is an abstraction” (Humphreys, 1979).

Indeed, it is an abstraction. But not so elusive as to be unmeasurable. Like electricity, it cannot be seen, but it can be measured, even though not always in ways that everyone agrees on. Some of the current tests emphasize verbal skills (vocabulary knowledge, abstract problem solving) other rely on visual-spatial skills; others have a heavy memory component.

A child might take two or three different I. Q. tests and get scores that range from 90 to 110, or even 85 to 115, depending on the different content of the test used. Also, due to some degree of measurement error, it is expected that on re-testing, a child is not likely to get exactly the same score. For example, on the WISC-R, if a child earns an I. Q. of 100, retesting on the WISC-R at a later date will probably (i. e, with 95% assurance) be between 94 and 106. Can intelligence measures be obtained in some way other than through the use of I. Q. tests?

Yes, they can. As mentioned above, at least three of the largest school districts in California are measuring mental ability and satisfying the requirements of the federal law without using I. Q. tests.

They are obtaining their estimates of intellectual ability through the use of adaptive behavior scales, interviews with the child and his or her teacher and parents, observation in the classroom, achievement test. results, and other memory, reasoning, and general adaptability measures that do not constitute a standard intelligence battery.

Whether this trend away from formal I. Q. tests will continue is unknown for two reasons: 1. these alternative assessment provisions may not prove to be any more useful than standard I. Q. tests, and 2.

Courts may rule that their use violates the wording of P. L 94-142 which reads, in part, that assessment instruments shall have been normed for the purpose for which they are being used. (Note: this has already occurred.

After the San Francisco Unified School District stopped using I. Q tests, they started using the Learning Potential Assessment Device, a non-normed, subjectively scored instrument designed to allow a trained observer to infer a child’s level of mental ability.

The Federal Office of Civil Rights, responding to a parent’s complaint that the LPAD did not meet the federal guidelines regarding norming, ordered the district to stop using it as an instrument for determining special education eligibility).

Criticisms of the intelligence Test

Intelligence tests have been criticized on various grounds almost since their

inception. Some of these criticisms have been given above: difficulty in defining 4. Standard I. Q. tests tend to focus on verbal problem-solving ability. As Gardner (1983) points out in his book Frames of Mind: the Theory of Multiple Intelligences, there are numerous other ways of being intelligent, such as mechanical, artistic, or in the area of leadership.

5. The use of I. Q. tests as the basis for determining the ability part of the ability-achievement discrepancy formula is questionable to the extent that if the child has a true learning disability, the disability itself would lower the child’s ability to be successful on some or all parts of the I. Q. test.

6. In order to simplify scoring procedures, most intelligence tests emphasize items that require one, or a very small range of, correct answers. Divergent thinking, which is most closely related to creativity, and more sophisticated kinds of reasoning are rarely called for on I. Q. tests.

7. Contrary to the beliefs of many lay-persons, I. Q. tests cannot measure innate learning capacity. Theorists agree that both nature and nurture combine to yield present functioning.

Therefore, if a child comes from the non-dominant. culture or has been raised speaking another language, or has not been given the standard advantages readily available to other children, he or she will appear to be dull on I. Q. tests. It is easy to forget that the I. Q. is likely to be depressed because of the contribution of the child’s background.

8. I. Q. tests are not designed to measure the processes underlying a child’s responses. They are better suited to measuring the products (achievement) than the processes (intelligence that accounts for achievement).

9. I. Q. tests are very limited in their ability to predict out-of-school success.

10. Current I.Q. tests are not appropriate for curriculum building. Knowing a child’s score tells us virtually nothing about the way a child should be taught.

Summary

This paper has attempted to summarize the history, purpose, present status, and criticisms of individual intelligence tests. It should be clear that these tests leave much to be desired. It is possible that we are asking too much of them, and that no matter how they are designed or normed, they will always have some of the flaws mentioned above. We can hope that future generations will have entirely new ways of assessing mental abilities, perhaps based on direct measurement of brain function, which will go a long way toward solving our present problems in this are.