Less Testing, More Learning

Nice Op-Ed piece by Guisbond and Neill.

-Patricia

My guests are Lisa Guisbond and Monty Neill of the National Center for Fair and Open Testing, known as FairTest, a non-profit organization that works to end the misuse and flaws of standardized testing.

By Lisa Guisbond and Monty Neill
During his campaign, Barack Obama was a quick study, learning he could get wild applause and cheers by bashing No Child Left Behind, especially from audiences of public school teachers and parents.

Many felt hope, even exhilaration, to hear Obama say that “we can meet high standards without forcing teachers and students to spend most of the year preparing for a single, high-stakes test.”

They hoped that Obama understood, as they did, that while test makers were getting fat and happy with the NCLB-driven testing explosion, the people giving and taking the tests were getting little return for a huge investment of time and resources.

Now more than three years after No Child Left Behind came due for reauthorization, President Obama’s administration has unveiled its blueprint for reforming the law.

It’s good to see that Obama has finally ditched the preposterous demand that all students make “adequate yearly progress” toward 100% “proficiency” on state tests by 2014. That could be a real victory.

Overall, however, we at FairTest conclude that the blueprint’s details show a deeply disappointing failure to learn from NCLB’s many other big mistakes.

For instance, it still makes standardized testing the centerpiece of the law. (For a more detailed analysis, see this.)

George W. Bush’s education secretary, Margaret Spellings, often said we need to give state tests every year, because otherwise we would have no way of knowing if students were falling behind.

But that’s hogwash, as any teacher, student or parent could have told her.

In the classroom of any reasonably competent teacher, student progress is being evaluated constantly, each time he or she looks at classroom work, not to mention frequent quizzes, papers, projects, and discussions.

All of these allow for valuable, rapid feedback, producing information, or “data” if you wish, that can be used by teachers to address student weaknesses and build on student strengths. Want to make sure students don’t fall behind? Then pay close attention to this kind of data. Want to ensure teachers do this well? Then ensure they have time to work together and learn from one another.

The data from annual state testing, on the other hand, is just not that useful for improving instruction. For all the talk of “data-driven education”--and there is a lot of talk--by the time teachers receive the results of spring testing, usually the next fall, it’s too late to respond efficiently and effectively to their students’ needs.

Teachers lament, for example, that state test data may show one year’s fourth graders are weak in one area of the math curriculum, so there’s a big effort to rejigger things to address that weakness.

But the next year’s fourth graders’ weaknesses might lie in a completely different area. (Not to mention different strengths and weaknesses from student to student.) Oops, too bad. And with just a couple questions in any area, there is too little information to use for teaching. To say nothing of the fact that the tests are mostly multiple-choice.

What about the need for the public to hold schools accountable? To see whether district
A is falling behind and decide whether to intervene? Testing in a few grades, say 4th, 8th and 10th, along with high-quality ongoing assessment in between, would be more than adequate for both assessment and accountability purposes.

The Joint Organizational Statement on NCLB recommends the testing burden on states be decreased by allowing states to assess students annually in selected grades in elementary, middle schools, and high schools.

The statement has been endorsed by 151 national education, civil rights, religious, children’s, disability and civic organizations.

The Forum on Educational Accountability has a more detailed report on how to develop a high-quality assessment system from the classroom up. These provide an assessment blueprint for Congress that could turn NCLB into a tool for improvement, not overtesting and punishment.

If we did adopt this “less is more” approach to testing, we’d look more like high-achieving countries such as Finland, which has no high-stakes testing program, or Singapore, Hong Kong and typical European nations or Canadian provinces, which test the equivalent of once each in elementary, middle and high school.

In a nutshell, the feds should help states develop systems that build on the assessments teachers already do, ensure the quality is good and reasonably comparable across the state, and then use statewide tests as an occasional supplement. That will enable teachers to go back to teaching, not running test prep programs.

In the coming weeks, we will talk about other parts of the blueprint.