ROANOKE TIMES 
                      Copyright (c) 1996, Roanoke Times

DATE: Tuesday, May 28, 1996                  TAG: 9605290012
SECTION: EDITORIAL                PAGE: A-7  EDITION: METRO 
SOURCE: DAN FLEMING


NEW TESTS WON'T REFORM PUBLIC SCHOOLS

VIRGINIA'S State Board of Education, following the lead of Gov. George Allen, has produced a revised set of Standards of Learning designed to reform our K-12 public schools. They will be accompanied by a new assessment program to hold schools accountable for student performance.

The General Assembly this spring appropriated more than $12 million to establish a new assessment system, and Allen asserts that for the first time, Virginia will have accountability for school performance.

Several vital questions still have not been answered by the State Board of Education. What will be done as a result of the new assessment program if it is determined that many or most of our schools are really in bad shape? What determines if a student, school or school division fails? Will there be some sort of punishment given students, schools or a school system, such as cutting state funding? Will the state send in remediation teams from Richmond to take over a school system, as has been done in other parts of the country? How will such remediation be paid for?

Another option mentioned by state officials is to convert the failing public schools into some form of charter school. And yet another idea floated suggests that schools doing well will be rewarded financially by the state.

Let us now turn to another feature of this subject, that of the assessment instruments to be used to evaluate the SOLS. Contrary to Allen's claim, Virginia has more than 20 years of statewide standardized, machine-scored testing, using nationally normed results. We know how students have performed in reading at three grade levels for every school in the state.

For example, in 1974, fourth-grade students in Falls Church on the Science Research Associates achievement tests, scored at the 77th percentile in reading, putting them well above the national average, while fourth-grade students in Sussex County scored at the 27th percentile, far below the national average. Turning to the 1994-95 Iowa Tests of Basic Skills, we find Falls Church fourth-graders scoring at a lofty 74th percentile, and the Sussex fourth-graders scoring at a dismal 35th percentile.

The same pattern of achievement exists over a period of 20 years. Does anyone doubt that even with all the new SOLS and testing, five years from now, Falls Church will still score well above the national and state average and Sussex County well behind? Obviously, we have voluminous data telling us where schools are doing well and where others have difficulty.

Another topic ignored by state leaders to date is where all the new materials, equipment and teacher training for the new SOL programs are to come from. Will this be yet another unfunded mandate from Richmond, placing even greater pressure on the limited tax base of local cities and counties?

Our leaders in Richmond would be wise to listen to the advice of the old farmer who said, "You can't fatten a pig by weighing it." Having a new set of tests won't fatten what is already a very lean educational system.

One of the few voices that pointed this problem out in the recent General Assembly was Sen. Madison Marye, who suggested tax dollars would be better spent on more funds for the public schools than for new tests. Virginia state government ranks near the bottom of the nation in the amount of money provided local governments for public education, placing a severe burden on the property-tax base at the local level.

Assuming that the state is going to move ahead with a new assessment program, one has to consider what form these new assessment instruments will take. In a recent article in The Roanoke Times, Lil Tuttle, vice president of the State Board of Education, implied that all the new tests will be machine-scored, probably relying on multiple-choice questions. Tuttle is quite correct that for many SOL standards ,such tests are quite suitable and can measure higher-order thinking skills. They also are much less expensive than performance tests.

Unfortunately, some of the new standards can't be measured adequately by machine-scored tests in areas such as writing and computer skills. It may cost much more to evaluate competencies in skills such as these, but not to do so would defeat much of the purpose of the reform effort.

Other issues must be addressed concerning assessment. A prominent national consultant for the State Board of Education on the SOLS, Dr. Diane Ravitch from Columbia University, co-authored a book on the national assessment of history. As a result of her work with this major national project, she declared, "Assessments in history should not rely exclusively on multiple-choice questions" and should include essay questions. She concluded, "The short-run cost of scoring essay answers seems to be less than the long-term, invisible cost imposed by a method that encourages superficiality and rewards guessing," as she maintains multiple-choice questions tend to do.

A serious problem that lies ahead is that of test security. If schools can fail a test and be punished for doing so, there is the likelihood that some teachers and even school divisions may engage in unethical practices to jack up test scores. One highly publicized example of this took place in Virginia Beach in 1980, where school officials were caught providing students with study materials containing verbatim test questions.

Finally, Ravitch expressed concern over another aspect of the easily scored pick-a-winner type of question, that of "smart kids psyching out the test-makers by using good test-taking strategies." She concluded "the difficulty is that the test then rewards students not on the basis of what is ostensibly tested, but on their mastery of test-taking skills."

All of these legitimate concerns make it imperative that the State Board of Education develop a long-term comprehensive assessment plan involving teachers and school administrators in the process.

Let us now turn to the great challenge of how to improve our public schools. We know from past test data in Virginia that money alone won't solve the problem. We have examples of students in small, rural school divisions with limited resources doing very well on standardized tests. So other significant factors are obviously in play.

Demographer Harold Hodgkinson tells us that nearly one-fourth of our nation's preschool children are living in poverty, "often in a single-parent home without adult supervision after school." He goes on to add that "almost one-third of our preschool children are destined for school failure because of poverty, neglect, sickness, handicapped conditions and lack of adult protection and nurturance."

It is really not difficult to conclude that if we truly want to reform education, we must address the societal failures that are mirrored in our public schools. Creating a new set of SOLs in Richmond, accompanied by tough talk about failing schools, will not solve the problems of those schools where serious problems exist often beyond their control.

There is no quick fix to reforming public schools, particularly from the top down, without the participation of classroom teachers. Although having high-quality standards and a good system of assessment is essential, we are really avoiding the larger issue of a breakdown in some of our basic institutions, particularly in urban areas and other communities suffering from economic and social distress.

I strongly question that many of our schools in Virginia are failing. Most are doing a good job, although certainly there is always room for improvement. What is now needed is for Virginians to start asking their state officials and lawmakers to come up with some answers to the questions that I have raised above before we head much further along an uncharted course to another dead-end effort at public school reform.

Dan Fleming is a retired professor from Virginia Tech's College of Education, and served as a consultant for the Educational Testing Service in developing standardized tests.


LENGTH: Long  :  129 lines
ILLUSTRATION: GRAPHIC:  Nancy Ohanian/LATimes. 



























































by CNB