Test’s Accuracy Raises Concern From Teachers

During the next couple of months, every student in California in grades 2 through 11 will be taking the Stanford 9 test as part of a statewide testing program known as STAR (Standardized Testing and Reporting). Much is at stake here since every school’s scores will appear in the newspaper and on the Internet. Because the state curriculum guidelines and standards are either in place or in progress, it would be logical to expect that the test will be tied closely to those established standards and that our students will be tested on what it is we have agreed they should know.

The Stanford 9, however, was developed by a private company, Harcourt Brace, for national use and not by California educators for California’s students. Hence, some teachers are understandably concerned about how accurate this assessment will be.

To better understand the role testing plays in education, let’s start with a single teacher’s classroom. Along with that teacher’s personal beliefs, available materials, and agreements made with colleagues about how to teach the course come school guidelines, district requirements, and state level demands. So what is actually taught in the classroom evolves from a number of sources. From all this, the teacher develops a course of study and a set of strategies to implement a curriculum. At some point that teacher needs to know if the students are in fact learning what is being taught.

This is where testing comes in. Ideally, the teacher examines the students on what has been deemed worth knowing, but it clearly doesn’t stop there.

Teachers look at the test results carefully because there is much to learn from them. If the scores are high, it may mean students paid attention and the teacher was amazing, or it could mean the expectations were too low. In the case of weak scores, it could mean the students did not try or pay attention. Then again it could also indicate the questions were ambiguous or the material was not taught well. And regardless of the actual numerical scores, are the students nevertheless improving? Or how do these students compare with a similar class taught by a different instructor?

Assuming the test is a valid measure of student achievement, then teachers can focus on the way they are delivering the material. If it’s a good test, then it should encourage good teaching. If students don’t do well, the teacher tries a different method of instruction. If the test grades go up, the new strategy has paid off. In this way the testing has given the teacher some valuable information that can be transformed into better teaching.

Effective testing should be part of a larger structure that also contains the curriculum and the methods used to teach that curriculum. These factors are interdependent and one cannot exist without the other.

If a district test on U.S. history, for example, includes questions about the civil rights movement, it should be based on an agreement that all teachers will cover certain aspects of the subject. Thus it follows that they will teach this material the best they can and that students will be held responsible for knowing it. In other words, there shouldn’t be any big surprises on the test. Everyone, including the students, is aware of what is expected.

If the statewide test seems to be following this sort of structure or at least moving toward it, then we may end up with a worthwhile tool for improvement, and not simply the illusion that we’re addressing problems. If, however, there seems to be a gap between what is being taught, how it’s being taught, and what is being tested, then we as parents, administrators, and teachers need to consider this carefully when the results come in before passing judgment on our schools.

It is reasonable to expect schools to be accountable and to demand that they do their job. Assessing that job, however, is a complex endeavor and cannot be based on what’s merely popular or expedient.

Nor can any true measure of learning be based solely on the results of one standardized test scored by a machine. Bubbling in “True” or the letter ‘C” cannot possibly indicate what a student truly knows and can express about a subject as broad as government or as deep as literature. The STAR test can, at best, offer us one view, but let’s not confuse that view with the total picture.