Wednesday, February 27, 2002
The Chronicle Herald

Standardized exams: the test of a good school

By Brian Lee Crowley

ON MONDAY, New Brunswick released school-by-school test results for core curriculum areas. But parents and students in Nova Scotia get no such objective test-based information on the quality of their schools. It’s time to ask why not.

Even though most parents and many teachers support standardized testing, such tests have been much criticized. Critics usually argue, as one writer did recently: “What does standardized testing show? Absolutely nothing more than how completely a teacher uses class time to teach to the exam and prepare the kids for the exam. These are not reliable assessments and evaluation methods.”

But according to Prof. Rod Clifton, an expert on the sociology of education at the University of Manitoba, this criticism doesn’t hold water when exams are properly designed. When they are, they yield the same score for all students across the province whose performance is the same. Moreover, such tests evaluate the material that students are expected to cover in the curriculum at the level that a committee of distinguished teachers and specialists thought was adequate for the particular subject at that grade level.

The exams themselves are tested in advance on samples of students and rewritten to eliminate ambiguities so that the tests reflect the content of the curriculum. After they are written, the tests are then graded by committees of specially trained teachers to ensure that the same level of achievement from students receives the same score, no matter where the student writes the exam.

Contrary to what many people seem to believe, standardized tests are very fair to students – particularly to disadvantaged students – because they are created by committees of teachers and subject-area specialists, more fully cover the curriculum, and more accurately measure the varying performances of all students. These claims are not a matter of conjecture, according to Prof. Clifton, but have been thoroughly demonstrated in the research literature.

Do standardized tests lead to an epidemic of “teaching to the test,” while “killing creativity” in teaching and learning? Come on. Proper tests are derived from the objectives of the curriculum. Teaching to the test is, in fact, teaching to the objectives that the Department of Education has established for our children. And as basic competency tests, they leave lots of scope for teachers and students to go beyond those objectives.

Of course, test results must be interpreted carefully. They can provide a valuable measure of both student achievement and instructional effectiveness, but they cannot be used as the sole measure of either. On the other hand, without such tests, the school system cannot be held properly accountable for the results that it is producing. And higher standards of accountability are something that the public schools need.

Most great scientific achievements have depended on standard measurements – weight, distance, mass, time, academic achievement, etc. – and such standards allow other scientists to test the validity of their colleagues’ work. In other words, they enhance accountability.

Few people question the usefulness of standardized tests and procedures in most scientific and practical endeavours. Especially post-Enron, most people, for example, think we should improve standardized accounting procedures, not get rid of them. Ditto for the breathalysers and laser-guns police use, or the regular elections, following standard procedures, we use to hold governments accountable. Nevertheless, some people oppose the use of standardized exams in education. But without such performance measures, how can we know where our schools are letting down our students and the community, or indeed where they are doing us proud?

The research literature reveals that well-designed achievement tests have much higher reliability and validity than other tests that have been developed to see how students are doing, such as socio-economic status and self-esteem. In fact, the social science literature is quite clear: Of all the student assessment instruments, the best are standardized tests. Properly used, well-designed exams give teachers and parents feedback to determine whether students have learned what they were supposed to learn. Properly interpreted, the results of good tests can inform teachers, students, parents and other citizens about the effectiveness of the schools.

Standardized tests, however, do not predict the future. Passing a driver’s test, a standard instrument that measures both knowledge and skill, doesn’t guarantee that you will never speed, run a red light, or have a serious accident. Similarly, when students achieve the provincial standard on an English language test, no one can say that they will be good at reading and writing throughout their lives. They can say, however, that when the exam is well-designed, the responses of students are assessed in an objective manner on items designed to measure the core objectives of a course in a way that is consistent and fair for all students in the province. That’s good for parents, teachers, employers, post-secondary institutions and, above all, students. Isn’t that what the schools are supposed to be about?

Brian Lee Crowley is president of the Atlantic Institute for Market Studies, a public policy think tank in Halifax. E-mail: bcrowley@herald.ns.ca