For years the U.S. Department of Education has championed standard-based reforms in the nation's public schools. Central to these reforms is our belief that schools, school districts and states should be held accountable for helping all students -- including students with limited English proficiency and minority students in poor schools -- achieve high academic standards.

The Post has understood these reforms and their potential to deliver educational excellence and equity. That is why we find puzzling the Aug. 23 editorial "If Not Tests?" suggesting that our draft resource guide on nondiscrimination in high-stakes testing could undercut the drive for high standards.

Our vision of reform -- challenging learning standards for all children, coupled with assessment systems that monitor progress and hold schools accountable -- has been the centerpiece of our educational policy agenda throughout this administration.

To be sure, tests aligned to challenging academic standards may reveal performance gaps among racial groups. When such performance gaps are based on educationally valid and reliable tests, we should direct our concern not toward the fairness of the test but toward the quality of the educational opportunities afforded to underperforming students.

Tests can help indicate inequalities in the kinds of educational opportunities students are receiving -- and the unfortunate fact is that minority students are more likely than their white counterparts to be in schools with watered-down curricula, unqualified teachers, large classes and fewer instructional resources.

The proper response to these circumstances is not to abandon testing but to measure all students against the same high standards while improving curriculum and instruction so that all students have equal opportunity to achieve those standards. In cases involving racial disparities in performance on high school graduation tests, our office for civil rights has preserved the use of the test in question while taking steps to ensure that students are being provided better opportunities to learn the material for which they are held accountable.

Educationally sound testing helps advance both learning to high standards and the provision of equal educational opportunities to all students, but if a school makes promotion or graduation decisions based on a test that does not validly measure what students have been taught, then those decisions will not be fair to any student. Moreover, when an invalid test produces lower promotion or graduation rates for minority students, it may run afoul of federal civil rights law. A test that is not overtly "racist" can still be unfair in this way. As one federal appeals court has put it, "An invalid test cannot measure merit."

MARSHALL S. SMITH

Deputy Secretary

U.S. Department of Education

Washington