The Risks of Comparing NAEP Results
Backers of the National Assessment of Educational Progress's plan to conduct a field test allowing state-by-state comparisons of student achievement early next year rationalize in several ways the decision of only about 20 states to participate in the test ("Fewer Than Half of States Join naep 'Pre-Test', " Sept. 7, 1988).
Indeed, the plan's advocates contend that support for such comparisons has shown "extraordinary" growth.
But the choice of 30 states not to participate may in fact reflect widespread concern in the states about the implications of such plans--and indicate the weakness of the arguments for expanding assessment.
In the magic world of making and giving tests, standardized minds, standardized thinking, and standardized curricula often drive a pretense of accountablity. State-by-state comparison is the logical extension of educational micromanagement, and multiple-choice questions become the ultimate determinant of educational quality.
This centralist mentality frequently confuses educational assessment with academic improvement. And without a balance wheel, testing will become the raison d'etre for education itself.
Where are all of the parents, local educators, and communities who--naep's deluded backers imply--are deluging the U.S. Education Department, state legislatures, and school boards with their demands for more testing?
In fact, many parents are already confused by the numerous measures of "outcome" to which their children are currently exposed. Many school districts use pre- and post-tests as part of the curriculum; administer standardized tests such as the California Achievement Test and the Iowa Test of Basic Skills; oversee individual district criterion-referenced tests; and participate in their own state's minimal-competency tests.
And we need more information about educational outcomes? Perhaps districts and states should undertake to explain the meaning of all the tests already administered rather than further confuse this mix by expanding existing measures or adding new ones.
So what if one state scores higher or lower than another state? With data presently available, we can already guess the general outline of such rankings.
And what changes would a low-ranking state need to make in order to raise its test scores? Would a state-by-state comparison help determine the relative importance of, say, higher pay for teachers? Restructuring of schools' organization? More competent principals? Smaller class size? Different curricula? A change in the basal reading program?
Might not a higher-scoring state be lulled into self-righteous complacency?
And so the race is on. "My test scores are better than your test scores" becomes the refrain of the day, and curricula are blatantly distorted to conform to the limited content a standardized, multiple-choice achievement test can measure.
These tests ignore character traits--determination, idealism, wisdom--as well as talents and skills--judgment, writing ability, creativity--essential to the educational process.
Has anybody's wall chart ever suggested that school systems be evaluated on the basis of their libraries? By the kind of citizenship training they provide their students?
What are we doing ignoring 99 percent of the important factors in evaluating educational performance? If school districts are told to raise test scores, indeed they will--by teaching to the test. The test, in essence, then becomes the curriculum.
The National pta agrees with the American Academy for Education that "when test results become the arbiter of future choices, a subtle shift occurs in which fallible and partial indicators of academic achievement are transformed into the major goals of schooling." When the test becomes the curriculum, major decisions about programs and goals will be made by the testmakers--not by involved parents and communities.
Indeed, the National pta believes that educational accountability is imperative in improving our schools for all children. But measurements must include other factors in educational quality, such as teacher competence, class size, instructional leadership, and parental involvement.
The psychometrics of assessments, however, are not advanced to the point that such comprehensive measurements are possible.
In fact, the U.S. Education Department explained away as as an "anomaly" the huge drop in naep reading scores between 1984 and 1986. Do we want to pin a profile of an entire educational system on "anomalies"? Educators and parents must not accept the imputation of the plan's supporters that opponents of state-by-state comparisons do not value accountability.
Could it be that an unofficial plebiscite on testing has just been held? That 30 states, for various reasons, are questioning the wisdom of state-by-state comparisons? That the stigma testing places on some children is being questioned by some states?
If indeed--as Gordon M. Ambach, executive director of the Council of Chief State School Officers predicts--all states eventually "find money and the way" to fit naep into their testing programs, it will be because of the politics of assessment rather than the wisdom of state-by-state comparisons.
In that case, groups like Friends for Education Inc.--the West Virginia group that last year charged standardized-test scores in most states were artificially inflated--will have more work than they can handle, and public confidence in education will decline. And children will be the ultimate losers.
Vol. 08, Issue 06, Page 24Published in Print: October 12, 1988, as The Risks of Comparing NAEP Results