« Is Testing Designed to Promote Inequality? | Main | Should Communities Determine Their Own Accountability Measures? »

Why Are We Using Standardized Tests?

| No comments

This week Jack and Julian are discussing the uses and abuses of standardized tests. They continue their conversation today.

Schneider: In our last conversation we discussed some of the philosophical underpinnings of standardized tests. But I want to talk about moving beyond test scores in piecing together a picture of school quality.

What should be included? And if not by standardized test scores, how should we measure student achievement?

Heilig: We need a paradigm shift from our current conception of assessment. For example, a large state like Texas spends half a billion dollars for students to sit for a set of Pearson exams that ultimately provide information at 10,000 feet. If you ask teachers how valuable these particular assessments are to curriculum and pedagogy, they will usually tell you not much. These expensive high-stakes assessments may get one or two glances at the start or end of a school year from educators because the information is not-specific enough and already dated when it arrives.

Instead, we should focus on summative and formative assessments that occur on the local level and provide the information for data-inspired instruction. For example, performance-based assessments are a more appropriate form of summative assessments being implemented by a consortium of New York Schools. 

As Monty Neill wrote in the Washington Post, performance-based assessments can include both written and oral components. The typical components of performance-based assessments are analytic essays, a social studies research paper, a science experiment, and an applied mathematics problem. 

Performance-based assessments essentially model how the real world works, as individuals are typically judged on the outcomes of integrated multi-faceted work products, rather than choosing from a set of multiple choice answers.

Schneider: I agree with you here. And I think it would be hard to argue that teachers don't already know much of what standardized tests tell them.

Of course, the tests aren't really designed to inform teachers. They're designed to provide universally-interpretable information that can be consumed by those outside of schoolsparents, community members, district officials, policymakers, etc. And whatever the problems of current evaluation methods, it is important to recognize that there's value in helping those outside the school understand what's going on inside of it. That is, at least if we believe that it's the job of outside stakeholders to help build school capacity.

I agree that multiple choice tests are insufficient for this task (in fact, I wrote about it for Ed Week), and that performance-based assessments would be far preferable. But that presents two new problems: cost and uniformity. It takes a lot longer to read a student essay, and to train raters to read them, than it does to run a Scantron sheet through a computer.

In short, you're right. But we also have to recognize that there are no simple solutions here.

That said, I certainly don't favor the kind of testing currently taking place. And I am particularly disturbed by the narrow-minded view that standardized tests capture even a fraction of what matters in schooling.

Heilig: It's also worth noting that we have leaned too heavily on high-stakes assessment as the lynchpin of accountability, teacher evaluation, and our general societal evaluation of school performance. We need to think more broadly in terms of multiple measures of data to evaluate our schools rather than the Pearson paradigm—a very narrow view from two miles above the earth.

Schneider: You raise an important point here. Rather than approaching test scores with the understanding that they are narrow and imperfect, policy leaders often assume a level of precision that is simply impossible to achieve. It's almost farcical.

You mentioned that we lean heavily on high-stakes tests in our evaluations of school performance. What else should be a part of that equation? What other information should the public have access to? I know I have my wish list; but what's on yours?  Maybe we can address that in an extra post tomorrow.

You must be logged in to leave a comment. Login |  Register
Ground Rules for Posting
We encourage lively debate, but please be respectful of others. Profanity and personal attacks are prohibited. By commenting, you are agreeing to abide by our user agreement.
All comments are public.

Advertisement

Most Viewed on Education Week

Archives

Recent Comments