Board Critiques Use of IQ Tests

K-12 Testing

The Board of Assessment and Testing (BOTA) of the prestigious National Research Council (NRC) has issued a report concluding that "the usefulness of IQ tests in making special education decisions needs reevaluation." This conclusion, summarized in The Use of IQ Tests in Special Education Decision Making and Planning (1996), was based on research evidence and discussion at two workshops. BOTA was created by the NRC to further research and evaluate the use of testing in education and employment.


An IQ test is one of the assessment instruments generally used in diagnosis for possible special education placement. A major problem, BOTA found, is the lack of connection between assessment practices and effective treatments. The report explains that assessment information should be useful for instruction and improving student learning, but IQ test-based information typically is not; and that while the educational consequences of using IQ tests in special education decision-making must be considered, often they are not.


The BOTA report also analyzes the continuing problem of overrepresentation of African Americans in special education. In 1990, African Americans were two-and-a-half times as likely as Whites or Hispanics to be classified as having "mild mental retardation," continuing a pattern reported in 1978, though the overall use of this classification has declined.


However, one research paper for BOTA reported that poor children were much more likely to be placed in special ed., and African American and White children in poverty had equal rates of special education placement. In the Larry P. v. Riles federal court decisions of 1979 and 1986, California was barred from using IQ tests for placing Black students in classes for the "educable mentally retarded" or equivalent categories, on the grounds that the tests were biased against and not valid for African American children (see Examiner, Winter 1988, Fall 1992).


At times the BOTA report defends IQ tests with arguments that critics such as Stephen Jay Gould have thoroughly exposed, such as reliance on factor analysis to claim that there is one general form of intelligence (g-factor). However, the report clearly recommends greater caution in the use of IQ tests, notes many flawed assumptions made about the tests and their uses, criticizes practices that have been discredited by research, and recognizes the tests as limited instruments.


Several alternative testing approaches are considered in the report. Other good methods and instruments are not, for example, the Primary Language Record/California Learning Record (see Examiner, Summer 1992). Preliminary evidence suggests that teachers who use the PLR/CLR are far less likely to recommend that students be tested for learning disabilities. This implies that the disabilities may be artifacts of the instructional practices used in many classrooms or that they can be overcome in regular classrooms with improved instruction. Approaches of this sort need further research, but tend to be studied less than test-based approaches.


-- The report, edited by P. Morison, S.H. White, and M.J. Feuer, is available from BOTA, NRC, 2101 Constitution Ave., NW, Washington, DC 20418.


--Gould's Mismeasure of Man is available from FairTest; use the order form on p. 15.