Is it fair to students when major decisions -- affecting not only their education but in some cases, their future -- are based on the results of a single test? According to the National Center for Research and Evaluation, a student who takes a standardized test a second time may have only a 30-50 percent chance of scoring within 5 points of his initial score. In fact, a score change of as many as 10 points may be completely attributable to the test. Other factors that may influence a test score include whether the child receives clear directions, follows those directions carefully, takes the test seriously, and is comfortable taking tests.
In the world outside of K-12 education, major decisions are routinely based on more than one type of assessment, says John Merrow, author of Choosing Excellence: Good Enough Schools Are Not Good Enough. A medical doctor, for example, wouldn't operate on a patient based on the results of a single test. Instead, the doctor would take a second measurement and look for other indicators. Similarly, college admissions offices never base their decisions solely on test scores. "You don't get into Harvard because you got 1600 on your SAT," explains Merrow. "They use multiple measures, yet we're willing to take a single measure and say this determines whether a student goes on to the eighth grade or graduates from high school."
Peter Sacks, author of Standardized Minds: The High Price of America's Testing Culture and What We Can Do About It, recommends "performance assessments" intended to reflect real-life situations, which could include more open-ended testing questions, portfolios, essays, presentations, exhibitions, and large projects carried out over a period of time. Compared to standardized tests, Sacks says these types of assessments would provide a more accurate gauge of student achievement.
Gary Orfield, an education professor at the Harvard Graduate School of Education and the Kennedy School of Government, suggests a set of alternative assessments that can be used along with standardized tests. These assessments would reflect different learning styles of students, provide timely feedback, address curriculum actually taught in the classroom, and be developed in collaboration with teachers.