Rethinking Testing

Under the twin banners of “Back to the basics!” and “More accountability!” enthusiasm has grown across the country for standardized testing as a way to improve the schools. In the Milwaukee Public Schools testing is increasingly determining how students experience schools.

Before we continue our headlong charge into test driven curriculum, we need to pause and carefully look at what tests can and cannot accomplish. There are disturbing signs that over-exuberant testing is beginning to interfere with learning. Consider a few examples:

  • Research has found that the administration of group tests to young children often diverges widely from what are considered ideal testing conditions. As Ken Wodtke points, out in this issue, group testing in some kindergarten classrooms departs so dramatically from appropriate procedures that it invalidates test results completely.
  • Every year, elementary school teachers are encouraged to use stacks of materials to prepare students for spring tests. While we want our students to be “test-wise”, the extensive use of these materials take time away from the classroom curricula that teachers are responsible to cover.
  • In order to receive extra money through the state sponsored P-5 program, 14 elementary schools with low achievement scores are giving new tests in five subjects. As Bob Peterson notes in his article on the P-5 program, many teachers involved think that the increased testing is narrowing the curriculum in a way which hurts the quality of instruction.
  • In all elementary schools, students’ reading levels are determined by the basal end of section and end of book tests. These tests, which do a poor job measuring comprehension and higher order thinking skills, are frequently criticized by teachers for inaccurately measuring children’s reading levels.
  • At the high school level, English teachers are compelled to devote hundreds of hours to teaching students how to pass the English competency test. Yet there is little evidence that constantly practicing how to write a five paragraph essay or drilling to pass a grammar test will help students become substantially better writers. Unfortunately the reality of competency instruction is all too close to the tongue-in cheek fantasy pictured by Cynthia Ellwood and David Levine in their “Grim Fairy Tale.”
  • According to Deputy Superintendent Hawthorne Faison, MPS is developing “an outcome based curriculum, which will specify the essential, common core of learner expectations.” The new curriculum will encompass most or all subjects in all grade levels and be designed to consume 75% to 80% of instructional time. Achievement of many of the “learner expectations” will be measured by objective referenced tests. Administrators plan to begin implementing this giant step toward a more test driven system in September, although the testing component will not be in place until sometime later. There is a great danger that the test-centeredness of this new program will diminish the time and effort that teachers will be able to put into creative teaching.

Teachers, parents and administrators all look to the growing array of tests as the source of ultimate judgement on what a child has learned and is capable of. But as Peter Murrell’s article argues in this issue, standardized tests only give us a sketchy measure of a narrow range of cognitive abilities. They do not measure creativity or problem solving ability and they do not correspond with performance in real life to any significant degree. Thus tests are misused when they are accepted as the definitive statement of a student’s capacity to learn, or serve to justify guiding children into academic channels which will not challenge them to develop the full measure of their academic abilities.

Even when we consider those skills which can be measured by standardized tests, the growing time and money we are putting into testing may be distorting the picture we get of what students actually know. Teachers rarely “teach the test” in the crude sense of reviewing specific test items beforehand. But we are under increasing pressure to drill students in the isolated skills they will be tested on. The result may be a gradual rise in test scores which reflects more class’ time devoted to test preparation rather than a real gain in knowledge. Thus, students who perform better in basal skill tests may not be becoming better readers. As the Commission on Reading notes, “many children manage to pass the mastery tests (of skills management systems) without learning to read very well.”

Indeed, as test scores rise, the quality of teaching may actually be deteriorating. Extensive test preparation has already begun to warp the curriculum. If teachers feel they are constantly under pressure to prepare students for multiple choice questions which mostly survey lower-order cognitive skills and rarely require original thinking, they will not have the energy or incentive to engage their students in activities or projects which inspire love of learning, imaginative problem solving, or critical thinking. As elementary teachers scramble to drill students for the end of section tests in the basal readers, they have little time for classroom activities that encourage students to read for pleasure. Secondary English teachers, worried about what percentage of their students will pass the competency test, will be reluctant to challenge students to write creatively, read critically, and develop real insight. Thus, good teachers, who exercise their creative and professional judgment to elicit the highest possible achievement from their students, find their morale sapped by the rigidity and distorted standards that overzealous testing imposes, and may even be driven from our school system.

Perhaps most disturbingly, the brunt of this testing mania is borne by low income and minority children. It is in the low income schools where pressure is greatest to raise test scores. To the extent that low test scores do indicate a need to raise the quality of education, the danger is that we will respond by devising a curriculum that trains for the test instead of seriously examining ways to improve the process of teaching and learning.

Despite serious limitations, standardized tests do serve two important functions. Reading and math tests can provide students and their parents with one indication of individual academic progress in narrowly specified skills. Just as importantly, such tests can provide a community with one measure of how well its schools are doing, and how well different groups of students are doing. In a city such as Milwaukee, which has a long history of the undereducation of black and Hispanic students, system wide standardized testing, if done with appropriate tests, can play an important role in measuring our progress toward equal education.

While recognizing the limited, yet valid, role of, testing we need to develop an approach to evaluation which is more balanced and creative. School evaluations should consist of much more than an annual meeting during which administrators and teachers anxiously review a computer print-out to see how well their school stacks up on the system-wide testing scoreboard. Classroom observation, questionnaires, interviews with students, teachers and administrators, and portfolios of student work can all provide valuable information not available if we only look at test scores and grade point averages. Just as importantly, teachers need the time and appropriate settings to reflect together on how well their school is doing and design improvements.

When they are well designed, standardized tests are a tribute to our capacity to collect, classify, and report a huge amount of data. But if we are not careful, these same tests can become a tribute to our capacity to become the victims of the ingenious instruments we devise. The school administration has developed a dangerous obsession with standardized testing. This obsession has already caused considerable damage within our classrooms, and more damage may well be on the way. The arrival in the next few years of mandatory objective referenced tests in most or all subject areas of all grade levels is a prospect truly horrifying to consider. Before we implement this new outcome based curriculum, which sounds like such a sensible idea, we would do well to consider the warning of Stanford University Professor Larry Cuban:

“When the model curriculum standards are wired to the tests and texts, and then all are wired together to a larger accountability program (which means how you do on the tests is then publicly disseminated), what you get is measurement driven instruction, a rational, bureaucratic systems management approach to teaching. When all these things come together, as they are coming together in California, Texas, and a few other states, the intent is to determine as precisely as possible what teachers will teach…The creative and imaginative part of teaching shrinks.”

The inevitable result of such a development here would be a school system in which tests have gone beyond measuring progress to effectively dictating the content and methods of instruction. The classroom experience of students would become increasingly routineized, constricted and truncated. We have a responsibility, working with parents and school officials, to make sure that students and teachers are not forced to live out the administration’s testing obsession.