Previous Section | Chapter 2 Table of Contents | Next Section

Say "test" to nearly anyone-student, teacher, administrator-and the face clouds over. Beyond the simple fact that testing by its very nature tends to intimidate, there is good reason for this reaction. Indeed, in recent years the entire subject of testing and assessment has come into intense scrutiny at all levels of education from the lower schools on up. In adult literacy the issue has assumed particular relevance.

In April 1988 Congress enacted legislation which for the first time calls for using standardized tests to evaluate ABE and ESL programs funded under the Adult Education Act. The Adult Education Amendments of 1988 (Public Law 100-97) and the implementing regulations of the U.S. Department of Education (August 1989) require that the results of standardized tests be used as one indicator of program effectiveness.

For the adult education and literacy community this new mandate brings special urgency to what was already a matter of growing concern: the use and misuse of standardized tests.

From the sheer volume of standardized test-giving, it would appear that we are a nation obsessed. For example, a study by the National Center for Fair and Open Testing estimates that U.S. public schools administered 105 million standardized tests during the 1986-87 school year alone. This included more than 55 million tests of achievement, competency, and basic skills which were administered to fulfill local and state mandates, some 30-40 million tests in compensatory and special education programs, two million tests to screen kindergarten and pre-kindergarten students, and 6-7 million additional tests for the GED program, the National Assessment of Educational Progress, and the admissions requirements of various colleges and secondary schools.

A major reason that standardized tests have come into such pervasive use is that they are relatively easy to administer on a wide scale, no small matter when dealing with a large population. Moreover, they are viewed by their advocates as scientific measuring instruments that yield reliable and objective quantitative data on the achievement, abilities, and skills of students, data that are free from the vagaries of judgment by individual teachers. Because the test and the conditions under which they are administered are (theoretically) constant, except for the skill being tested, they are thought to be useful for comparing a person's ability form one time to another, as in pre- and post-testing. By the same token they are viewed as useful for evaluation program effectiveness and by extension as a tool for improving educational quality.

However, as standardized tests have come into sweeping use throughout education and employment, so have complaints about them and challenges to their validity. They have been the subject of criticism in congressional hearings and state legislatures, and are increasingly the subject of lawsuits in state and federal courts.

Not surprisingly, when the new federal requirements for standardized testing in ABE and ESL were set forth this past August, it was over the objections and protest of many members of the adult basic education community. [Note: See the Federal Register, August 18, 1989.]

The reasons are compelling. Assessment in adult literacy is a central issue with high stakes. The authority vested in these tests can determine the way programs are developed, what is taught, and the climate of teaching and learning. It shapes legislation and the funding policies of public and private agencies. It is tied to welfare eligibility for young parents. It drives government job training programs. It can deny entry into the military, or crucial access to a diploma or a job.

The growing concern of literacy service practitioners, theorists, and test designers, among others in the field, is sparking much debate and a hard look at just what standardized tests actually test and for what purposes, and whether the results tell us anything of real value, indeed whether they are not harmful. It is also beginning to result in a search for alternative assessment approaches.

The complexities of the testing controversy are vast and beyond the scope of this general article, but opponents of standardized basic skills tests fault them for a host of reasons, some of which are discussed below. Objections tend to fall into two broad categories: their intrinsic defects, and their misuse.

MAKING GRADE LEVEL COMPARISONS

The most commonly used general literacy tests are off-the-shelf commercially-produced tests of reading achievement. Virtually all are "normed" on children. That is, their scores are based on the average performance of children at various grade levels. Because adults bring years of prior knowledge and experience to the acquisition of literacy skills, comparisons with the performance of children are considered by most experts to be inappropriate.

Test scores are usually in the form of grade-level equivalents. A person may score at a 4.2 grade level, say, meaning that he or she reads on the level of a child in the second month of the fourth grade. Not only is this humiliation to people already the victims of past school failure, charge the critics, but it is meaningless to tell adults of any age that they read like a nine-year-old. More importantly, it is not a useful measure of what adults can do in terms that are contextually meaningful and it does not point to an appropriate instructional program.

In fairness, it must be noted that the Test of Adult Basic Education (TABE) which appears to be the most widely used of all general literacy tests and which has been mandated for use throughout New York Statehas recently been improved. Analysts indicate that while TABE is still strongly tied to childhood norms, the newer version does make it possible to interpret test scores in relation to other adult in certain ABE programs, rather than to children. It also produces scaled scores rather than grade-level equivalents (though many administrators are apparently falling back on the grade-level scoring system they know because they find the scaling system hard to interpret).

TESTING TRIVIAL SUB-SKILLS

"TABE and other standardized general literacy tests are not a true representation of how people read." says Clifford Hill, Professor of Applied Linguistics at Columbia University Teachers College. "They force the reader to recycle very low level trivial details and don't really represent the reading process with all its complexity." The questions they pose deal with isolated, decontextualized bits and pieces of reading sub-skills such as word recognition spelling, or paragraph comprehension. Questions are framed in a multiple choice format, and they dictate one right answer. There is no applied use of reading or math, no writing component, no higher order thinking or problem solving. "The way the tests are set up, the research shows tha