Standardized achievement tests are often misused as indicators of a school's quality or effectiveness relative to other schools. This is an incorrect use because it ignores variation among schools in student abilities, family support of education, student mobility, and other factors. People also misuse tests because they impute to them more statistical reliability and validity than they actually have. An example using the Iowa Tests of Basic Skills illustrates the possibilities for reliability errors with both single tests and retesting over time. The use of norm-referenced tests to evaluate Title I programs, under the Elementary and Secondary Education Act, is incorrect because it makes the doubtful assumptions that national norms are valid and that one can compare either different groups of students or students with and without Title I treatment. However, criterion-referenced tests have similar problems--with reliability, neglect of some learning areas and processes, and arbitrary choice of satisfactory performance levels. Hence it is unsound to use criterion-referenced tests alone in minimum competency testing, because the tests do not account for the range of student abilities, the student's employability, or educators' possible failure to solve a student's learning disability. (Author/RW)