This document has been formatted for printing from your browser from the Web site of the Illinois Association of School Boards.

COPYRIGHT NOTICE -- This document is © copyrighted by the Illinois Association of School Boards. IASB hereby grants to school districts and other Internet users the right to download, print and reproduce this document provided that (a) the Illinois Association of School Boards is noted as publisher and copyright holder of the document and (b) any reproductions of this document are disseminated without charge and not used for any commercial purpose.

Illinois School Board Journal
July August 1999

Too much testing, ETS study charges

Too much testing is going on in K-12 schools, and too much of it is the wrong kind, charges a report published by the Educational Testing Service.

"The testing enterprise in K-12 education has mushroomed in the last quarter-century," charges author Paul E. Barton, director of the ETS Policy Information Center.

"Americans want numbers when they look at students, schools, state education systems, and how America’s students compare to those of other countries," he says. "Among political leaders, testing is turning into a means of reform, rather than just a way of finding out whether reforms have been effective."

According to "Too Much Testing," most standardized testing used today "rarely results in better information to aid instruction and achievement." Among the shortcomings of standardized testing:

• Commercial standardized tests are used that measure a blend of what is being taught across the nation — not what is taught in a school or district (and not what is supposed to be taught).
• The test content changes from time to time to reflect changing views of what should be taught. Yet the scores from year to year are used to judge whether progress is being made.
• In many cases, norm-referenced tests designed to show how one school’s students compare with those in the entire nation are used to track change in the school’s performance over time, a task they are not designed to do.
• While the tests are presumed to judge the quality of what the school does, a large part of an individual’s score is attributable to family background and opportunities outside the classroom. Tests that measure the quality of current in-school instruction and out-of-school development are used unfairly to reward or punish schools.
• While tests presumably are used to determine how well the school instructs from the beginning of one grade to the beginning of the next grade, the tests actually determine the cumulative level of knowledge of eighth graders (for example, not what knowledge was added during the eighth grade). Tests rarely measure the "value added" — the change in the levels of knowledge between two points in time.

Barton recommends more use of sample-based testing — like that used in the National Assessment of Educational Progress. This approach would provide better information about schools with less intrusion into the classroom.

According to the report, "The greatest promise continues to be in intensifying efforts to establish strong standards for the content of instruction, developing curricula reflecting this content, and aligning assessments to the curricula actually being taught."

By 1998, the Council of Chief State School Officers (CCSSO) reported that almost all the states had some kind of content standards in place. But 29 states had not yet aligned their assessments with the standards. The result, says Barton, is new content standards and old tests that do not reflect the new curriculum.

Once states develop assessments to reflect the new content standards, they still must establish performance standards — determining what score is necessary for performance to be judged acceptable or advanced.

CCSSO reported in 1998 that 21 states had established performance standards, but we do not know how directly they are linked to the content standards, Barton notes.

"While we need to complete the content-assessment-performance triad," he concludes, "we do not need this ever-larger volume of standardized testing of individual students to render individual scores."

Aligned standardized assessments can be used on a sampling basis, or without assigned individual scores, for school accountability purposes and tracking achievement changes, he recommends. And well-trained teachers should be responsible for judging individual performance — by giving out grades and passing or failing students.

This article was provided by the National School Boards Association News Service.