In the United States, the use of nationally developed standardized tests has proliferated during the past decade (Council of Chief State School Officers, 2000). In a recent survey of U.S. public attitudes about education reform (Hart & Teeter, 2001), almost one-half of the respondents touted the benefits of these tests for improving education, but one-third expressed concerns about their possible misuse. The survey identified two recurrent themes: Testing is important in education reform, but we need to use tests carefully.
What Tests Can Measure
What purposes do tests like the National Assessment of Educational Progress (NAEP), Stanford 9, the Iowa Tests of Basic Skills, and the ACT and SAT serve?
The NAEP surveys the educational accomplishments of groups of U.S. students in a variety of subjects. It reports the results by student populations—grade levels, for example—and subgroups of those populations, such as male or Hispanic students, but does not provide individual scores of students or schools. The reporting metrics used by the NAEP allow performance comparisons within a subject from year to year and from one subgroup of students to another in the same grade (Brown, Dabbs, Kostad, & Horkay, 1999).
Unlike the NAEP, the Stanford 9 and the Iowa Tests of Basic Skills report individual student achievement. The questions on these tests align with standards that have been developed by such national organizations as the National Council of Teachers of Mathematics, the American Association for the Advancement of Science, and the National Council for the Social Studies, among others (Harcourt, 2001; Riverside, 2001). The Stanford 9 combines open-ended and multiple-choice questions to measure students' academic achievement, and the Iowa Tests of Basic Skills provides a comprehensive assessment of student progress in basic skills, including the ability to read maps and locate and evaluate different sources of information.
Aiming beyond the K–12 curriculum, the ACT Assessment and the College Board's SAT-I and SAT-II attempt to predict students' readiness to undertake college-level work. Most colleges consider these tests' results as part of evaluating applicants for admission. The ACT measures general educational development in English, reading, mathematics, and science (ACT, 2001), and the SAT-II assesses achievement in such subjects areas as English, biology, and foreign languages (College Board, 2001). The SAT-I, on the other hand, measures verbal and mathematical reasoning abilities that students develop both in and out of school; it does not test any particular state, school, or district curriculum or high school course (College Board, 2001).
Asking Too Much of Tests
Used for their intended purposes, these assessments can assist education reform by tracking the progress and levels of achievement of individuals or groups of students and by indicating who is ready to tackle college-level work. Unfortunately, the pressure on educators and policymakers to demonstrate accountability in schools has driven some to use the test results inappropriately. The NAEP Guide cautions that casual inferences related to subgroup membership, the effectiveness of public and nonpublic schools, and state or district-level educational systems cannot be drawn using NAEP results. (Brown, Dabbs, Kostad, & Horkay, 1999, p. 30) Nonetheless, Donald Gratz (2000) points out that "educational accountability is still in its infancy" and "testing is often handled poorly" (p. 681).
To address the concerns raised by Gratz, David Pearson and his colleagues (2001) contend that we must not ask the tests to perform tasks that they cannot do. Karen Mitchell (1997) agrees that many reform efforts have been derailed by misaligned assessment, and she cautions principals and teachers to track performance on a broad range of outcomes over time.
Ensuring Credibility and Validity
- Provide safeguards against the exclusion of students from assessments.
- Use multiple indicators instead of a single test.
- Emphasize the comparison of performance from year to year rather than from school to school.
- Consider value-added systems, which provide schools a reasonable chance to show improvement.
- Recognize, evaluate, and report the degree of uncertainty in the results.
- Evaluate both the intended positive effects and the more likely unintended negative effects of the system.
- Monitor the consequences of tests and identify and minimize the potential for negative consequences.
- Accompany reports of group differences in test scores with relevant contextual information. Caution users against misinterpretation.
- Explain supplemental information to minimize possible misinterpretations of the data.
- Ensure that the individuals who make decisions within the school or program are proficient in the appropriate methods for interpreting test results.
If educators and stakeholders have a thorough understanding of how to use and interpret assessment results, they will have powerful opportunities to bring about sustained, systemic improvement in our schools.
