In the mid-1990s the former Conservative government introduced standards tests for students in Grades 3, 6, 9, and 12. These tests were designed to reflect the widespread concern that learning outcomes needed to be assessed at various stages during the educational lives of students.
Even though most parents and many teachers support standards testing, these tests have been criticized by a number of parties including the Manitoba Teachers’ Society (MTS). Such critics usually argue, as one writer did in a recent letter to the editor of the WFP: "What does standards 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."
As the NDP promised during the election campaign, Drew Caldwell, the Minister of Education, has already announced that the standards tests in math and language arts for Grade 3 students have been eliminated. He has also promised to re-evaluate the standards tests at the other grades and to announce his decision sometime this month. If the tests are abolished, Manitoba will be one of only a few educational jurisdictions in North America without tests of this type. If the tests are abolished, Mr. Caldwell will be doing the students, parents, and teachers of Manitoba a terrible disservice.
Technically speaking, the tests used in Manitoba are "standardized criterion-referenced tests." To call a test "standardized" means that it is a relatively objective test that yields the same score for all students across the province who achieve the same performance outcome. To call a test "criterion-referenced" means that it measures student achievement against a predetermined standard, or criterion, such as specific Grade 6 reading skills. In other words, standards tests evaluate the material that students are expected to cover in the curriculum at the level that a committee of distinguished teachers and specialists who designed the test thought was adequate for the particular subject at the specific grade level.
In addition, the items are pre-tested 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, following specific protocols, to ensure that the same level of achievement from students in various schools and divisions receive the same score.
Good classroom assessment, of course, begins with a teacher’s own observations and measurement of what students are learning. Overall, teachers spend between 20 and 30 percent of their time assessing the work of students. But many teachers never take courses in the specialized field of test construction and evaluation, and many never study ways of improving the reliability and validity of their assessment instruments and procedures. This means that many teacher-constructed tests have low reliability and validity. In comparison, standardized tests have high reliability and validity. More importantly, and contrary to what many people seem to believe, standards tests are generally fairer 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 but have been thoroughly demonstrated in the research literature.
Some critics charge that the emphasis on standards tests has lead to an epidemic of "teaching to the test," while others charge that these tests "kill the creativity" in teaching and learning. These criticisms are misplaced. First, the tests are derived from the objectives of the curriculum, which means that teaching to the test is, in fact, teaching to the objectives. Second, the tests are basic competency tests, which means that they cover the core objectives. As a result, there is considerable opportunity for teachers and students to creatively move beyond those objectives.
What do standards tests offer that parents, students, and teachers cannot obtain without them? In one word, they offer "comparability"–comparability in the context of all the students in a particular grade throughout the province. It isn’t very useful for teachers to compare their students’ achievement with the achievement of students one room down the hall and then use that information to make decisions about their teaching. It isn’t very useful for parents to compare their children’s performance with the performance of a few other children and then to make decisions about the effectiveness of teachers and schools. It isn’t very useful for students to compare their performances with the performances of their friends and classmates. These comparisons lack a meaningful context.
Assessing Teachers and Schools
Some school administrators and teachers are justifiably concerned that the average performance of students on one test, in one classroom, during a single year, will be used to evaluate the effectiveness of a teacher, a principal, or a school. As we know, the achievement of students is affected by many variables other than the actual teaching that takes place in a classroom. A number of students, for example, may be ill with the flu or there may be a very high turnover rate of students during the school year. When these things happen, the assessment of the students does not represent the effectiveness of a specific teacher or a specific principal. Consequently, it is often inappropriate to generalize from an assessment of individual students to an assessment of classrooms and schools.
All of this simply means that the results of standards tests must be interpreted carefully; without careful analysis and without specific qualifications, the results of tests should not be used to attack teachers or principals. Standards tests can provide a valuable measure of student achievement and a valuable measure of instructional effectiveness, but they cannot be used as the only measure of the strengths and weaknesses of either.
By themselves, tests are incapable of harming students, teachers, or principals. It is the way in which the results of tests are used that can be potentially harmful. Critics of standards tests, such as the MTS, often overlook this important distinction, preferring to blame the tests rather than the people who interpret, or misinterpret, the results of the tests. It is unfair for people to blame test results, particularly those they do not like, on tests while absolving the people who interpret the tests—-newspaper reporters, parents, teachers, principals, superintendents, and trustees–from any responsibility. In other words, people who interpret the results of tests need to understand what tests can and cannot tell them.
All scientific achievements, both physical and social, have been dependent on the standardization of measurements, weight, distance, mass, time, academic achievement, etc. Few people question the usefulness of standardized tests and procedures in most scientific and practical endeavours. Not many people, for example, argue that standardized accounting procedures should be discontinued or that the police should be forbidden from using breathalyzers and laser-guns. Not many people, including Mr. Caldwell, would argue against using elections, following standard procedures at standard intervals, to hold governments accountable. Nevertheless, a number of these same people seem to discount the use of standardized assessment in education.
The research literature reveals that well-designed achievement tests have much higher reliability and validity than tests that have been developed to measure other social and psychological characteristics of students, such as socio-economic status and self-esteem. In fact, of all the assessment instruments that have been developed in the social sciences, the best instruments are standardized criterion-referenced achievement tests. Properly used, well-designed standards tests can give teachers and parents feedback to determine whether students have attained the desired learning objectives. Properly interpreted, the results of good tests can inform teachers, students, parents, and other citizens about the effectiveness of instructional programs.
Finally, standards tests are not designed to predict the future. When a person passes a driver’s test, a standard instrument that measures both knowledge and skill, no one can say with certainty that the person will never speed, run a red light, or have a serious accident. Similarly, when students achieve provincial standards on an English Language test, no one can say that they will be good at reading and writing throughout their lives. In other words, no one can claim that standards tests are flawless. They can claim, however, that standards tests are useful because 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. So, Mr. Caldwell, please serve the interests of students, parents, and teachers: keep the standards tests.