
                         Roanoke Times
                 Copyright (c) 1995, Landmark Communications, Inc.

DATE: SUNDAY, May 13, 1990                   TAG: 9005130092
SECTION: NATIONAL/INTERNATIONAL                    PAGE: D-4   EDITION: METRO 
SOURCE: The New York Times
DATELINE: WASHINGTON                                LENGTH: Medium

SCHOOL TEST ADMINISTRATORS AGREE TO NATIONAL STANDARDS

A federal education panel that administers a 20-year-old test in the three R's voted unanimously on Friday to set the first uniform national standards to indicate how well the students are doing and how many of them are coming up short.

The unanimity of the vote by the governing board of the Department of Education's National Assessment of Educational Progress masked weeks of debate about what students should know, who should set the standards and whether the whole process of setting goals for fourth-, eighth- and 12th-graders will provide any incentive to improve achievement.

Compromises were needed to achieve unanimity, said Lawrence W. Feinberg, a board staff member.

The most important, he said, was replacing an original proposal to set a single minimum standard for reading, writing and mathematics with a stepped approach.

Instead of being judged adequate or inadequate, reading skills will be classified as deficient, basic, proficient or advanced.

"The NAEP results have never been reported against any anchored sense of how good is good enough," said Chester E. Finn Jr., a former assistant secretary of education who is chairman of the agency's board.

"That's the crux of this."

The board's intent, he said in a telephone interview, was to set achievement standards in reading, writing and mathematics that would remain constant for years, so the results of each biennial test could be compared with the results measured either in 1990 (for the pilot program in eighth-grade math) or 1992 (for all three subjects at all three grade levels).

In effect, Finn said, the system would set off four groups of students: those who score above the advanced standard, those between proficient and advanced, those between basic and proficient and those who do not meet the basic standard.

The pilot program is scheduled to go into effect late this year after panels of scholars, teachers and parents decide what kinds of problems and concepts on which eighth-grade math students will be tested, Feinberg said.

Must a student be able to compute an arithmetic mean and know how to compute the area of a circle? Or must he merely be able to master only enough of such concepts and skills to achieve a certain score, while no individual skills are singled out as essential?

"One of the major judgment calls the panels are going to have to make," Finn said, "is, `Are we talking about specific individual competencies or an average score or a bit of both?' "

The test is administered to a random sample of students every two years.

States and territories can decide for themselves whether to take part in a broadened testing effort that would allow state-by-state comparisons.

by CNB