The Past, Present And Future Of High-Stakes Testing

Jan 22, 2015
Originally published on February 6, 2015 2:14 pm

After a long stretch as the law of the land, annual standardized tests are being put to, well, the test.

This week, the Senate education committee held a hearing on the reauthorization of the No Child Left Behind (NCLB) law and, specifically, on testing. The committee's chairman, Lamar Alexander, R-Tenn., has released a draft bill offering a lot more leeway to states in designing their own assessment systems.

But Education Secretary Arne Duncan and Sen. Patty Murray, the ranking Democrat on the committee, have dug in their heels to say that annual tests should remain mandatory.

All this comes as parents, students and educators around the country are asking serious questions about the number of tests children are taking and the reasons they're taking them.

I've just written a book on this topic, The Test: Why Our Schools Are Obsessed with Standardized Testing — But You Don't Have to Be, and Steve Inskeep sat down with me to ask me a few questions about it.

How much more widespread are standardized tests than they were 10, 15 or 20 years ago?

Since No Child Left Behind became federal law, every state has been required to test every child every year in third through eighth grade in math and reading, plus once in high school. And districts have added many tests to follow on to state-mandated tests.

Why would they do that?

Well, the state tests are tied to consequences for districts, schools and teachers as well as students. Districts are adding on benchmark, practice and interim tests, and that's how they get these multiplying and ballooning requirements. That's why the Council of the Great City Schools found that students are taking 113 standardized tests in grades K through 12.

I heard from an elementary school teacher recently that she felt like she wasn't doing anything other than preparing for, or administering, tests.

I hear that from educators all around the country. Even the ones who want the data and the information, they lament the fact that the whole school experience is increasingly becoming defined by testing and test prep. And I've walked into lower-income schools around the country where test scores are posted right in the front entryway and the message is very clear: That we care about you as a person and everything, but what really matters is the score that you post in April.

In talking with teachers you even hear stories of the entire school experience being distorted by these tests because the school is worried only about kids who are just on the edge of passing or failing the tests. They don't seem to care about the kids who are doing really well or really badly.

What you're describing is an example of perverse response to incentives. And everybody does that to some extent. And researchers have documented what your friend noticed, which is called educational triage or rationing. More and more attention goes to students just on the verge of passing, and schools don't have capacity to focus on students who are doing really well or really badly.

This is in some ways an old debate. We've heard about it for decades. We are in this data-driven world where people want information and accountability for schools. How do you get beyond that?

Well, this is an interesting inflection point in the debate. On the one hand you've got parents and teachers saying, "This is too much." But the other question is, What can we do instead? I talk about a couple of different approaches in the book, like statistical sampling and "big-data" approaches [see this story for more details].

So how open is the political system to some kind of reform?

Well, this is a really interesting moment, Steve, because the political alignments around education are always confusing, but right now you have a situation where Sen. Lamar Alexander is proposing a draft bill to eliminate annual testing requirements and really leave it up to the states to have much more flexibility.

But Education Secretary Arne Duncan and Democratic leaders in Congress are saying we must protect annual testing at all costs — this being the policy initially introduced by George W. Bush.

OK, what is the case in favor of annual tests?

Well the core argument in favor is that NCLB forced schools to report the performance of historically disadvantaged groups — minorities, students with disabilities, English language learners and low-income students. So instead of hiding those students behind the average, schools were made responsible for the welfare and success of every single child. And so some civil rights groups are saying that equity must be ensured by testing every child every year.

Then you have groups on the other side saying equity isn't just about measuring students, it's about ensuring equal outcomes and equal access and equal opportunities, and we don't really believe that more testing is really going to do that.

Copyright 2015 NPR. To see more, visit http://www.npr.org/.

Transcript

RENEE MONTAGNE, HOST:

And these days, something that's been vexing many American students, parents and educators is standardized testing in schools.

(SOUNDBITE OF SPEECH)

SENATOR LAMAR ALEXANDER: Are there too many tests? Are they the right tests? Are the stakes for failing them too high? What should Washington, D.C., have to do with all of this?

MONTAGNE: That's Senator Lamar Alexander, chair of the Senate Education Committee, speaking at hearings underway about standardized testing. Once seen as the way to improve America's failing schools, standardized tests are now widely seen as part of the problem.

For a closer look at the issue, our colleague Steve Inskeep spoke with NPR education reporter Anya Kamenetz. She's also the author of the new book "The Test: Why Our Schools Are Obsessed With Standardized Testing - But You Don't Have To Be."

STEVE INSKEEP, BYLINE: How much more widespread are standardized tests than they were 10 or 15 or 20 years ago?

ANYA KAMENETZ, BYLINE: Well, since No Child Left Behind became federal law in 2002, every state has been required to test every child, every year in third through eighth grades in math and reading, plus once in high school. And districts have added on many tests to follow on to those tests that are state mandated.

INSKEEP: Why would districts be adding tests?

KAMENETZ: Well, the deal with the No Child Left Behind tests is that they are tied to consequences for districts, for schools, increasingly for teachers. And so when you make a test high-stakes, districts really care about the outcome. And so what they're doing is adding in practice tests, interim tests and that's how we get these multiplying and ballooning requirements.

INSKEEP: This is helping to explain something that I heard from an elementary school teacher that I know. She was describing a situation where she felt like she wasn't doing anything other than preparing for tests and administering tests.

KAMENETZ: You know, I hear that from teachers all over the country. And, you know, I've walked into lower-income schools where the students' test scores are posted right in the front entryway. And the message is very, very clear that, you know, we care about you as a person and everything, but what really matters is the score that you post in April.

INSKEEP: But this is, in some ways, an old debate, Anya. We've heard about it for more than a decade since No Child Left Behind. People complained about the amount of standardized tests even before. But we're in this data-driven world where people want information, and they want accountability for schools. And how do you get the argument beyond that, that question of what would you do other than test?

KAMENETZ: You know, I talk about a couple of different approaches in the book. One of them is very simple, and the idea is statistical sampling. So instead of testing every student every year, as we're doing now, we could follow the PISA exam - the famous international test - which tests a statistical sample of students. And that would give us a reliable indicator of how students are doing without having tests be so dominating,

INSKEEP: And is that valid? Can it work for this situation?

KAMENETZ: If what you're looking for is accountability on a district level or a state level, absolutely, it can work. You know, sampling obviously doesn't provide information about every single student or every single school. So, you know, that's why people who object to it say, well, we need these tests for different reasons. We need them to get information about students.

INSKEEP: OK, so that's one idea. What's another?

KAMENETZ: Well, you mentioned big data, and the fact is that technology provides us the ability to get much more broad-based information about how students are performing. All states now pretty much have statewide longitudinal data systems. They're tracking students from kindergarten all the way through the workforce in some cases. And so they have really good long-term indicators of what works well where it really matters, which is ensuring that students are able to succeed in college and beyond.

INSKEEP: So now, Anya, you're talking about something that will remind people of what Google does with your search data or Facebook might do with information that it gathers on you. It's just quietly gathering information about you and acting accordingly, and you never even know the information is being gathered about you.

KAMENETZ: Well, exactly. So on an individual basis every single day, more and more students in schools are using software to learn. And while the software is, you know, giving them the math problems or the English problems to answer, it's getting incredibly fine-grained information about the students' performances, how they think, how they approach difficult problems. Do they try really hard? Do they give up easily? And many people who, you know, run testing companies and software companies believe that that kind of information could replace the activity of stopping and testing.

INSKEEP: Wait, even the people who are making money creating the tests think there might be a better way to measure school performance than their own tests?

KAMENETZ: Oh, absolutely. I mean, I talked to chief research scientist at Pearson and ETS who said you know, we believe that this invisible integrated assessment is really the future because stopping and testing is this clunky kind of a way of going about it.

Businesses in the 1970s used to have to shut down at the end of the year and do inventory for three weeks. And that's kind of how we do it in schools today. We spend eight days taking tests. And so how could schools possibly use data in that same sort of agile, just-in-time way?

INSKEEP: Anya Kamenetz, thanks very much.

KAMENETZ: Thank you.

MONTAGNE: Anya is NPR's education reporter speaking there with Steve Inskeep. On MORNING EDITION from NPR News, I'm Renee Montagne

DAVID GREENE, HOST:

And I'm David Greene. Transcript provided by NPR, Copyright NPR.