Tests to Perform, Not to Take

`THERE are 36 sheep and 10 goats on a ship. What's the age of the captain?'' When this question is inserted in the middle of a standardized multiple-choice test, a majority of schoolchildren ignore the correct answer (``none of the above''). Instead they add the two numbers and answer 46. They have been taught that when you see the word ``and'' in a word problem, you add.

``They're being perfectly reasonable within the framework that their schooling gives them,'' says Monty Neill, associate director of FairTest, an advocacy group in Cambridge, Mass. ``You're really not supposed to think about the problems because they don't mean anything. You're just supposed to quickly figure out the rule and apply it.''

In the midst of a nationwide interest in school reform and restructuring, states and educators are being forced to rethink testing and assessment methods. Multiple-choice tests, such as the Scholastic Aptitude Test (SAT), have long been under attack and the current shift in educational goals suggests that more sophisticated measurements are necessary.

The controversy was underscored earlier this month when Education Secretary Lauro F. Cavazos released the annual state-by-state performance report known as the ``wall chart.'' The report card was released despite objections from the White House, officials acknowledged.

President Bush and the state governors met last September and outlined a 10-year program of reform for primary and secondary schools. Many educators suggest that a new yardstick is required to assess progress toward these goals.

``You can't go very much into school reform and school restructuring ... without tackling the problem of assessment,'' says Grant Wiggins, director of research for Consultants on Learning, Assessment, and School Structure (CLASS) in Rochester, N.Y.

In the process of revamping curricula, a number of states across the United States have begun to experiment with innovative testing methods. These alternative measurements - often called ``performance-based testing'' - take a variety of forms:

California now uses open-ended problems as part of its mathematics test for 12th-graders and is piloting new testing in English, science, and history.

Connecticut is introducing hands-on math and science testing for high schoolers.

New York requires all 4th-graders to conduct a science test and report the results.

Vermont is developing a program to include student work portfolios with standardized tests.

Some other states have plans to move away from multiple-choice testing to more authentic forms of testing.

For example, Kentucky, which is in the process of restructuring its schools, is aiming for performance-based testing across the grades by 1995.

Writing assessment has been the most readily adopted addition to multiple-choice testing. As many as 30 states now include essay writing of some sort in their testing programs.

``I think it's no accident that much of the work has been fueled by the success of local, and regional, and state writing projects, which easily led to developing writing assessments,'' says Mr. Wiggins.

While the quality and character of the various writing assessment programs vary greatly, the willingness to adopt such a practice suggests that administrators are willing to break away from strictly standardized, impersonal testing.

``Over the last 15 to 20 years we've simply become addicted to test scores,'' says Rexford Brown, director of communications at the Education Commission of the States in Denver. He calls for ``reeducation'' of the public and school officials on the issue.

``There's been a myth in this country that important evaluation should not involve human judgment,'' says CLASS's Wiggins. The United States is the only major Western country that holds that view, he says.

``The multiple-choice format is the American method,'' according to Mr. Neill. The appeal of this form of testing is its perceived objectivity, say experts. The tests are scored by machine, which provides an efficient, cost-effective manner in which to deal with a mass of material.

However, the limits and drawbacks of these traditional measurements are being scrutinized more and more.

`THE issue of trust is at the heart of a lot of our discussions of testing,'' says Mr. Brown. Parents, administrators, and policymakers don't trust teachers to judge and grade their students fairly. Therefore, external measures are imposed to get around human judgment.

``The supposedly objective measure is going to be supplanted by somewhat sloppier and messier but more interesting and useful kinds of indicators that involve human judgment,'' Brown says.

Opponents claim that performance-based testing is too expensive and cumbersome to be workable. The labor-intensive grading process concerns many critics.

``Grading essays is clearly more expensive than a machine-scored multiple-choice test,'' says Neill. He suggests, however, that testing does not need to be done every year.

The average student takes at least five tests per year in the United States, according to Neill. Some students in the lower-level classes do nothing but prepare for and take tests, he says.

``The cost differential [between standardized and performance-based tests] is not as great as is usually feared,'' according to Wiggins. ``You don't have to pay teachers very much ... to do this work because it is such wonderful professional development.''

Joan Boykoff Baron, the Connecticut state assessment coordinator, supports this view: ``You're gaining a tremendous amount if you use the scoring opportunity to develop the skills of your teachers.''

Connecticut, unlike most other states, has its own teachers score all essay exams. ``The professional development that emerges from the scoring is so worthwhile you don't want to pay somebody else to do it,'' says Ms. Baron.

THE restructuring of testing methods is expected to enhance classroom teaching in other ways as well. ``Performance-based assessment assumes you're doing projects, and thinking, and interacting - a whole different kind of curriculum,'' says Neill of FairTest.

``What you want is the testing or assessment and the instruction to look very much alike,'' says Jack Foster, Kentucky's secretary of the education and humanities cabinet.

``[That way] it wouldn't matter what day you came in and watched what a child was doing. ... Learning becomes an ongoing activity demonstrating the very kind of things that you're looking for in the end.''

Almost all testing experts agree that there is a time and a place for standardized multiple-choice testing.

``One of the virtues of standardized testing,'' says Brown, ``is that it has enabled us to see ... comparative results of different populations.''

Mr. Foster has suggested that states should form a consortium of parties interested in innovative testing.

``If 15 or 20 states can get together and map out all the things we want to ultimately be able to test,'' he says ``then ... we can each invest in a piece of it and then exchange it among ourselves.''

``It's going to take a different kind of politics of school improvement in order to make these things fly,'' says Brown. ``But I don't think there are any serious intellectual barriers to our being able to do this.''