A Testy Subject : Exams Tell Public How Schools Are Doing and What Teachers and Parents Should Be Focusing On, Expert Says

Two years ago, the state of California unveiled the first results from an ambitious new kind of academic exam. Educators vowed that it would produce better information on what students know than traditional, multiple-choice standardized tests used in schools.

Students’ performance in reading, writing and math on the California Learning Assessment System, or CLAS, was generally awful. Even worse, critics said the test virtually ignored basic skills, included controversial reading passages, asked students probing personal questions and, because not all tests were graded, produced inaccurate results. The ensuing controversy was so intense that the entire system was scrapped.

Now the state is in the early stages of trying to find a new way to assess students. The Los Angeles Unified School District is trying to create its own means of measuring students in reading, math, science and history. And testing experts are sorting through a variety of difficult issues to make sure these efforts to produce a tool for improving schools.

Eva Baker, co-director of the UCLA-based National Center for Research on Evaluation, Standards and Student Testing, discussed the issue of testing with Times Education Writer Richard Lee Colvin.

*

Question: Academic testing has become one of the hottest issues in education. Why is testing so volatile, and so important?

Answer: People are concerned about the quality of education, and testing is one way people trust to let them know how schools are doing. Also, I think we’re a society that really likes numbers. We keep track of all sorts of statistics . . . and there’s always little tests to take, in Cosmopolitan or other magazines, where you answer some questions and get a score.

Recently, academics and the policy community have been talking about whether we are getting the best information from the tests that we have normally used. And whether tests can be used as a way of targeting attention on things we want to accomplish. So the purpose has changed from just a measurement--finding out how we did after the fact--to a way to communicating to people--parents, teachers, children--what they should be focusing on.

Q: Real estate agents--at least those in neighborhoods where scores were high--loved standardized tests because houses there sell for more. Why do educators say the scores on such tests don’t mean much?

A: Standardized tests are fine for . . . measuring in a very general way how well a child or a school are doing. The tests have to be general because it has to be useful for kids all across the country. But because it is very general, it isn’t much good for communicating to people what they should be teaching. It would tell you, for example, you should be teaching math. But not what kinds of problems you need to work on.

Q: How can the tests be made more valuable for educators?

A: I’m encouraged by the standards approach--trying to link tests to what it is . . . important to teach. If the tests are connected to important goals, the nature of the test will change and we’ll ask kids to do pretty complicated things related to those goals. Like we might ask someone to analyze a historical document or actually solve a complicated math problem that requires measuring and weighing. I think those tests will be more like learning sessions.

If we do that, it will be harder for people in schools to say the results don’t count . . . and schools will be put under pressure to fix things if they’re not working.

Q: How about for parents?

A: Tests have always been a mystery for parents. What I like about the new kinds of tests is that people will be given real examples of what kids ought to be doing at particular grade levels. And it seems to me, parents can look at what’s expected and get a rough sense of whether their kid is in the ballpark, rather than getting just a number. At least the parent can go to the school and say, “It looks like my daughter isn’t doing well, and what are you going to do about it?”

Q: What should the public role be in developing the new tests?

A: The public should have a very strong role in deciding two things: the general standards kids should meet and making sure that the content meets broad-based community standards, that it’s not biased in favor of one group or another and that it is not intrusive. We just cannot sustain another loss of confidence like we had with CLAS because the content is seen as being inappropriate.

Q: Can tests regain the credibility lost with the failure of CLAS?

A: There are two levels of credibility. One is that the tests are seen as being fair to all kids. Second, they have to be seen as really challenging, that they’re not something the system is using to cover up. People also have to think that the tests are worth the kids’ time to do well on them. And that is another problem with standardized tests, that people didn’t think they were worth the time.

The test has to be credible for the teacher as well, that it is not just busy work, that it is something . . . they can use to help them figure out how to teach their kids better.

Q: Could you explain these terms we’ve heard so much about--performance-based assessments, portfolios, authentic assessment? Are they reliable? Aren’t the results subjective?

A: Performance-based assessment means kids have to do something with what they know. Authentic assessment means the task has real meaning for kids. A lot of these new kinds of assessments are pretty interesting. But I don’t think there’s one kind of assessment that fits all.

It isn’t fair to say the grading is subjective because people who do the scoring are trained . . . to agree to explicit standards, not just their point of view.

The portfolio assessment is where you accumulate student work over a period of a year or longer. Studies of that show they’re best used by teachers and not as part of a very formal accountability system.

Q: Many school districts are concerned about a state law requiring them to test all kids, including those whose native language is not English and those in special education. Should all students be tested?

A: The idea is that if you don’t test these children, they may be forgotten. The pressure now is for inclusion and the problem is how to do that in a way that maintains high standards but is also fair to all children. So, it’s not easy.

Q: What role will technology play in the future of testing?

A: Being really conservative, in 10 years all testing will be computer-based. It will occur in simulated environments where the data is being collected as the child is working through problems and it won’t even feel like a test.

In that world, testing and instruction really merge. And that is what we’re looking for.