STANDARDIZED TESTS

Whatever their inadequacies, standardized tests are one measure of school achievement. They do not measure everything that schools do. Often they do not even reflect accurately those things they intend to measure. But even though their limitations are often unrecognized, test results do provide necessary evidence of some kinds of learning. Unfortunately, a concerned public searching for simple ways to judge school effectiveness may exaggerate their importance and, as a consequence, misuse their results.

Test results are frequently misunderstood by the public and educators alike. Many do not comprehend the meaning of grade equivalents or understand that because "grade level" is a term reflecting the median, at least half the children in the country simply must perform below grade level. Test results indicating that some children are below "grade level" are often mistakenly taken to mean that schools and teachers are not doing their job. In some cases school funding may be tied to performance on tests. In others, school districts may wrongly attempt to evaluate teachers or grant tenure on the basis of test results. Tests may even cause a narrowing of the curriculum into score-producing areas. In such instances public ignorance as to the true meaning of scores becomes all the more crucial.

Despite these problems test results may provide us with a basis for some of what is needed to interpret school success. Educational evaluation must always look at test scores in relation to other factors such as the socio-economic background of the children tested, school organization and resources, staff experience and morale and others. But, when viewed in their proper context, test scores are important. Recent reports of a decline in test scores may, for example, indicate less content emphasis in the curriculum. Or, they may be a result of changes in the pool of students taking tests--for example, that more dropouts are staying in school; that more women are aspiring to higher education; or that absenteeism is declining. The point is that such results must be thoroughly analyzed because they provide some of the information needed for educational decision-making. Test information is part of what is needed to diagnose educational problems and to develop corrective solutions.

Obviously test results are valuable for some things and not for others. The problem of test abuse will not be solved by simply calling for the abolition of standardized tests. Such tests do provide us with a general measure of academic achievement in the basic skills. It is the responsibility of educators to insist on a continuous effort to improve tests. A concerted effort must also be made to educate the public as to their strengths and weaknesses and to develop a more thorough awareness of their proper uses and definite limitations. The American Federation of Teachers recommends the following with respect to the uses of standardized tests:

• The AFT encourages the passage of state and federal legislation insisting that test publishers accompany all tests with a clear explanation of what any given test purports to do and how its results may be interpreted and used. Publishers should also be required to issue warnings that would prevent misuse of the results. Such information might include: a) estimates of measurement error; b) interpretation of validity and reliability; c) information on the revision of scales or, norms; d) scoring procedures and methods of interpretation; e) explanation of norming systems. These same issues must also be addressed when test scores are released to the public.

• Test results should not be used as the basis for school funding. Estimates of monetary need must also be based on such factors as family income, the incidence of special problems among the student population, school violence, etc. Making funding dependent on test scores may encourage schools to retain low scores to avoid loss of funds. The results should not be used for evaluating teachers or staff performance. To judge teaching quality on the basis of test scores when many other factors impinge upon results is simply unfair.

• The AFT warns against any narrowing of the curriculum or the focus of schooling to include only those areas which produce measurable results. At the same time, the AFT urges that test development continue and that existing tests be further refined. New test forms such as criterion-referenced tests should be further developed and their uses explored.

• The public must be made aware that schooling is a comprehensive process involving many kinds of learning. The current movement to establish minimum skill levels as the basis for high school graduation totally ignores this premise since in some states (Florida, California) test success is taken to mean that a prospective graduate can leave school early. While such tests may be a useful indicator of student progress on measurable skills they should not be used as the basis for lowering the minimum school-leaving age. Students thrown into the labor market without the full benefits of a high school education are only being cheated and will probably end up on the unemployment rolls.

• The AFT recognizes that tests are often misused for purposes of labeling and tracking individuals who may, for a variety of reasons, have high intelligence and aptitude but be poor test performers. We encourage the kind of refinement of the testing process that allows the strengths of such individuals to be recognized. Such refinements must not lean in the direction of providing different tests for different cultural or ethnic groups, but must recognize socio-economic, ethnic and cultural differences. Such a practice is patronizing in its implication that some groups' performance should be measured by lesser standards than others. The AFT also rejects the notion that differing levels of group performance on tests is by itself evidence of test bias. While in some instances this may be a reflection of bias, it may just as well be an indicator of unequal educational opportunity.

• The AFT urges the exploration of new ways of administering tests that would help alleviate the anxiety test-taking causes to many children.

• Individual diagnostic or achievement test results must be balanced by teacher observation and judgment before any decisions affecting a particular child are made.

• School systems must make a greater effort to train teachers and other educational personnel on testing procedures and analysis. They must also involve teachers more fully in planning and executing testing programs.