Ruth Mitchell, the associate director of the Council for Basic Education, writes in Testing for Learning that "a new model of schooling is a national imperative." One aspect of the model she envisions will be standards whose achievement can be verified by assessment practices going beyond multiple-choice.
In arguing for the rapid implementation of performance-based testing, Ms. Mitchell underscores its benefits and reliability, but admits that it is an unknown quantity in some areas. In the excerpt below, she discusses its cost:
No one knows accurately how much performance assessments cost, partly because there is no accurate information about how much any form of educational assessment costs. When school-district or state personnel put a dollar figure on evaluation, they may be counting only the cost of buying the test from the publisher, complete with reported results. The district or state education office typically will not include the costs of the testing-and-evaluation office maintained (in big school districts and in state departments of education) to oversee the administration of the tests or the time administrators spend with test salespeople and, later, their own colleagues to decide on a test. Nor will they include the teachers' time spent preparing students for the test--time that, on the admission of many teachers, is not spent "teaching" in the sense that they recognize.
But the costs of a testing-and evaluation office, of consultants, and of teachers' time scoring assessments will be included in the cost of performance assessments because the work will be done in-house. When the cost comes out to be several times that of buying a test from a publisher, that figure is reported, but the basis for the comparison is forgotten or not recognized. The Educational Testing Service's 1990 annual report cites a state "with a strong commitment to educational assessment" that found that performance assessment would cost 10 times more than the existing state program. The basis for the alarming figure is not given.
Certainly, performance assessments cost more to score and report because it costs more to pay people to read papers than to load up a machine with bubble sheets. Even if the costs of development were the same, the extra cost for scoring still makes performance assessment a more expensive proposition than multiple-choice examinations--no matter whether school systems buy them from a test publisher or develop them in-house.
But is the cost of scoring an assessment cost alone? When you pay teachers to score, do you just get reports back? As [was shown in] the California writing assessment,... the return is more than graded papers. The teachers [who do scoring] understand more about what students need to know and be able to do. As they themselves say, scoring assessments is the best professional development they ever experienced.
The costs of scoring could just as easily be charged to professional development as to assessment. The same is true of developing assessment tasks in states, such as California and Maryland, where teachers are involved, and districts, such as South Brunswick, N.J., where the teachers modify the K-2 portfolio annually. Teachers who face the difficult question of finding suitable tasks, making them clear and fair to all students, ensuring that they do test what is valued, are probing into the heart of their professional concerns.
Thus, the question of comparative costs depends on what you want to count and where you want to put it in the budget. For performance assessment, but not for multiple-choice tests, some of the costs can be attributed to other parts of the budget, since working with performance assessment can replace other forms of professional development.
In any case, the basis of comparison must be clear before any figures are accepted, especially the frightening tenfold, 30-times, 100-times claims that appear in the newspapers. Further, we really do not know what performance assessment will cost because there has not been enough experience with it yet.
One more point needs to be made about the cost of assessment: Americans want to get education on the cheap, and assessment is no exception. The public and legislators on all governmental levels want information but are unwilling to pay what it really costs. Dale Carlson, the California Assessment Program director, has tried to draw attention to the disparity between needs and the money to meet them by calling for the "1 percent solution." He means that assessment might reasonably cost 1 percent of the state budget for California education, which would be 20 times more than it costs now. The situation is, if anything, worse in other states and at the federal level.
Testing for Learning: How New Approaches to Evaluation Can Improve American Schools, by Ruth Mitchell. The Free Press, 866 Third Ave., New York, N.Y. 10022; 217 pp., $19.95 cloth. Copyright (C) 1992. All rights reserved.
Vol. 11, Issue 23, Page 25