ROANOKE TIMES
Copyright (c) 1996, Roanoke Times
DATE: Sunday, May 5, 1996 TAG: 9605040007
SECTION: EDITORIAL PAGE: 3 EDITION: METRO
SOURCE: LIL TUTTLE
AS GOVERNORS and business leaders met recently to refocus attention on academic standards, testing and accountability, it was acknowledged that Virginia is ahead of most other states in these areas.
In fact, during the same week as the Palisades summit, the Virginia Board of Education met to work on statewide testing, which is the key to any systemic improvement. Without objective, reliable student testing, true accountability is a pipe dream and academic standards are paper tigers.
Testing should be far less controversial then standards, yet there is confusion among some who are aiming strong charges at the standardized, machine-scorable form of testing. Their argument is that traditional standardized tests only test basic skills and rote learning, but not the higher-order thinking skills, They advocate the use of performance assessments, asserting that these assess higher-order skills better than standardized tests.
Performance assessments are not new. Teachers routinely assess a variety of student performances: homework, projects, quizzes, etc. Classroom-performance assessments tell teachers whether teaching practices are effective and where learning breakdowns are occurring. Parents see the results of these assessments when they receive their children's report card.
This report is incomplete, however, in that it fails to advise how one student's composite performance compares to others of his age group. As one researcher noted, "...at some point, a child and his parents have a right to know whether the child's progress is reasonable for his or her age and experience."
Current attempts to adapt the small-scale classroom tool known as performance assessment to a large-scale state testing system are problematic in several ways:
Performance assessments are hand-scored, meaning that students' scores are determined by human judgment, which contains innate biases.
Despite good intentions and aggressive efforts to overcome it, human subjectivity inevitably finds its way into scoring, meaning that performance assessments offer increased opportunity to discriminate against a person or class of persons. The subjective scoring of performance-based assessments is inherently unfair to students, since their future opportunities may depend upon the results of large-scale statewide testing.
Costs associated with creating, administering and scoring performance-assessment tasks are enormous - perhaps five times the per-pupil cost of standardized testing.
Estimated cost of a recent performance assessment of fourth-grade math by the National Assessment of Educational Progress was $150 per pupil. The NAEP board has wisely decided to use more standardized, machine-scored test items in future tests.
Performance assessments also cost dearly in instructional time.
Researchers evaluating Great Britain's pilot of large-scale performance assessments found that it took two to five weeks of the school year for performance assessments, compared to three or four days for standardized testing. The North Central Regional Education Laboratory estimated that "in the time it would take a student to complete one or two performance assessments, that same student could have completed 200 items on a multiple-choice test."
Not incidentally, 200 multiple-choice items can cover substantially more subject matter, giving parents and teachers a far more accurate indication of a student's overall achievement.
Standardized tests are machine-scored, ensuring that the exact same score is awarded for the exact same student performance. Researchers remind us that standardize tests were created almost 50 years ago to surmount the problem of inconsistent scoring of students' college-entrance essays - the performance assessments of that time.
Since those early days, standardized testing of large numbers of students has become a legitimate science, offering high levels of reliability and fairness that have withstood scholarly and legal challenges.
As to the charge that machine-scorable testing only assesses rote, lower-order thinking skills, it should be noted that it was through standardized testing that the United States has been able to document the decline of students' higher-order skills.
State-of-the-machine scorable tests measure students' ability to reason, analyze, synthesize and logically evaluate knowledge as well as their ability to read, calculate and communicate. The Scholastic Aptitude Test, still a strong predictor of college performance, is a machine-scored test almost totally devoted to assessing higher-order skills.
Clearly, the indictments against standardized testing are bogus, so why the objections? According to University of Tennessee researchers, "... educators have come to view standardized tests as the root of the public's disenchantment with the schools and the prime reason for the denigration of the teaching profession. It's no wonder there has been such an outcry against their use." If this is true, it appears the message has been confused with the messenger.
The Virginia Department of Education should facilitate and encourage the creation and distribution of classroom-based performance-assessment instruments for instructional use. Created through the collaborative efforts of classroom teachers, such assessments can enhance teaching and cultivate skills in students within and across the various fields of scholarly and scientific investigation.
But the primary purpose of any statewide student-assessment program is accountability. Statewide testing must provide a system of checks and balances that reveals to the public whether local school experiments and innovations, including performance assessments, are succeeding or failing.
Are the public schools teaching what society expects them to teach? Are children learning what society expects them to learn? Standardized, machine-scorable tests answer these questions - whether or not the news is welcome.
How foolish it would be to abandon this most objective and reliable means of measuring student and school performance at this crucial juncture of education reform. To do so would virtually guarantee that the accountability that the governors and business leaders seek, and that society needs, remains just a pipe dream.
Lil Tuttle is vice president of the Virginia Board of Education.
LENGTH: Long : 115 linesby CNB