Putting Assessments to the Test

At Stuart Elementary School in Richmond, Tonika Shackleford takes a practice test in preparation for standardized testing. (By Jay Paul For The Washington Post)

? A test designed to screen students for learning disabilities is not used to measure student progress in reading acquisition.

? A test that says it predicts college performance actually does. The old SAT said it did, but experts said the test had limited ability to predict a student's performance in the first year and none beyond that. The test has been changed and, experts say, does not intend to predict achievement.

? A test is not used to guide curriculum.

"There has been an explosion of mandates for more and more standardized tests with very little evidence to support their use," said Walter Haney of Boston College's Center for the Study of Testing, Evaluation and Educational Policy.

The No Child Left Behind program has ushered in an unprecedented era of high-stakes standardized testing, which has dramatically changed what goes on in classrooms across the United States and caused fierce debate over the approach.

The issue of what the tests actually measure has become more important than ever because the results do, indeed, have high stakes, with jobs of teachers and administrators sometimes riding on the single administration of a test. Many experts say that, in this environment, there should be much more effort to ensure that tests are valid.

"If indeed in the long run No Child Left Behind and the accountability movement is going to really have traction in improving education for kids in the United States, I think it's going to have to subject itself to a serious level of scrutiny," said Robert Pianta, director of the University of Virginia's Center for Advanced Study of Teaching and Learning.

What does validity actually mean in the context of student testing?

Testing experts generally refer to three major areas of validity: