||Volume 17, Number 7, May/June 2005 |
High-stakes testing leads to bad education outcomes
by Sandra Mathison and Wayne Ross
Fearing that the newly instituted Grade 10 examinations will result in an increase in drop-out rates and intensification of teachers’ work, the Vancouver school trustees have called for its elimination.
Recent research on the impact of these kinds of tests on teaching and learning validate the trustees’ concerns and more.
The primary tool of educational reform in North America is now high-stakes testing (the use of tests to make critical decisions about students, teachers, and schools.) The rationale most often cited for increased use of tests is the need for information to help educational decision-makers shape both policies and practices that will enhance the academic success of all students.
The accumulated evidence of two decades of research indicates, however, that high-stakes testing does not lead to better educational policies and practices.
Moreover, there is evidence that such testing leads to educationally unjust consequences, even though it occasionally upgrades teaching and learning conditions in some classrooms and schools.
Here is what we know.
First, when tests are used for important educational decisions any early benefits are quickly outweighed by substantial negative effects on learning and teaching.
For example, focussing on test scores undermines student motivation to learn, and promotes superficial understanding of the subjects studied. Test-driven education assumes students ought to be thinking constantly about improving their performance. The assumption that test scores are all that counts overlooks a substantial body of psychological research suggesting that a focus on how well one is doing is different from a focus on what one is doing.
When raising test scores becomes the most important indicator of school improvement, pressure mounts for teachers to teach only what is covered on the test. The tests become the curriculum as teachers adjust instruction to enhance test scores rather than teaching in ways that are responsive to the interests and needs of their students.
Teaching to the test narrows the curriculum, forcing teachers and students to concentrate on memorization rather than critical thinking. Instructional time increasingly means rote drill, preparing for (and predicting) specific test items and taking practice tests. In effect, high-stakes tests transfer control over the curriculum to the people who write the exams.
Second, test-driven educational practices are technically unsound and amount to educational malpractice. No test is good enough to serve as the sole or primary basis for important educational decisions. This is a widely-held principle in the testing profession.
When test scores are central to decision-making, people tend to treat test results as the major goal of schooling, rather than as a potentially useful but fallible indicator of achievement. The fact is that high-stakes tests are not accurate representations of students’ performance.
Despite efforts to create tests that are reliable, scores on standardized tests can be surprisingly inconsistent; that’s why every major standardized test publisher tells schools not to use them to make decisions about grade retention or graduation. In addition, recent studies suggest that standard test scores more accurately indicate family income than students’ educational achievement.
Third, high-stakes tests threaten the most vulnerable students.
High-stakes tests undermine, rather than enhance, equity for aboriginal students, immigrant students, special needs students, and students from low-income families. For these students, access to innovative and successful programs are restricted in the rush to measure success solely by test scores.
The use of high-stakes tests is an effort to treat teaching and learning in a simple and fair manner, but in a world where education is complex and with inequitable distribution of opportunity.
The proven way to improve schools is to build strong curriculum, provide ample professional development for teachers and, most importantly, foster local community involvement.
Sandra Mathison is head of the Department of Educational and Counselling Psychology and Special Education, University of B.C. Wayne Ross is acting head of the Department of Curriculum Studies, UBC. They are co-editors of the book Defending Public Schools: The Nature and Limits of Standards-based Reforms.
Source: The Vancouver Sun, April 4, 2005. Sandra Mathison and Wayne Ross.