- The Washington Times - Monday, July 3, 2000

Any farmer knows that you don't fatten cattle simply by weighing them every day. Why then, have legislators and governors fallen into the trap of thinking that more and more tests, with higher and higher consequences for students and their schools, will improve education?

Standardized tests are a valuable tool for looking at how a child, a class, or a school is progressing in school. Properly used, they provide information to teachers, parents and the community about how students are doing in the subject areas tested, and can indicate where additional resources and focus might be needed.

A good system of accountability clearly demonstrates to a community the level of progress attained by schools or a school district. A good accountability system includes not only standardized tests, but report cards, teacher observations, attendance records and student performance on real life tasks, as well. When schools teach only what is taught on a narrow, multiple-choice test, the curriculum is narrowed. Public school parents surveyed in October 1999 indicated they know schools are "teaching to the test," and they don't like it. Standardized tests can measure factoids; they cannot as reliably measure problem-solving, honesty and perseverance. What is important to remember is this no single test can substitute for a system of accountability.

Standardized tests were not designed to be used for high-stakes decisions such as promotion or graduation for individual students. They are not valid for that purpose. They are imprecise; the same student's score will vary from one day to the next. And yet, about half our states use standardized tests for high-stakes decisions affecting students. Such use of tests has altered the role of teaching in many of our schools, and replaced investigation and exploration with memorization and repetition.

National polls show that, despite the increasing use of standardized tests for the high-stakes decisions of promotion or graduation, the public profoundly disagrees that a student's academic year can be summarized by a single standardized test.

A national poll of registered voters, conducted in May, showed that 63 percent of American voters do not agree that a student's progress for one school year can be accurately summarized by a single test. This is an overwhelming vote of "no confidence" that every politician and policy-maker should know about. Fewer than half of those polled agreed that standardized tests accurately reflect what children know about the subject tested.

This level of ambivalence is significant. Standardized testing is not a powerful education reform tool in the eyes of the American public. When student results on one state test become so important that teaching and learning are focused exclusively on the test, education is distorted. The atmosphere in the school becomes tainted, and teachers and principals feel less like professionals and more like drill sergeants.

Testing is a big "good news, bad news" joke for America's children. The good news is that, when used properly, tests can help teachers and administrators make smart decisions about how best to teach children and improve instruction. The bad news is that when used as a single determination of child's future, they are unreliable, unrealistic and unfair.

Paul Houston is executive director of the American Association of School Administrators.



Click to Read More

Click to Hide