Thursday, November 13, 2008

Would national testing really improve our schools?

Now that Bill Gates has offered to write new “national standards” and a national exam to go along with them, and many of our most influential opinion makers, including usual opponents like Joel Klein and Diane Ravitch, appear to support this notion as a solution to the country’s educational problems – I have a question:

Why wouldn’t a national test be open to the same sort of grade inflation than the state tests are now? With rampant test prep, cheating, narrowing of the curriculum, and all the rest – as predicted by Campbell’s Law?

And wouldn’t this lead to the loss of one of our few semi-objective measures of how schools are doing – the NAEPs?

The NAEPs, after all, are considered more reliable than state tests for a number of reasons which derive directly from the fact that they are not used for accountability purposes and no stakes are attached to theo9r results.

Only a random sample of students are assigned to take these assessments and because of their low-stakes nature, many of the questions can remain the same over the years, which makes the scaling (or difficulty level) more consistent; finally, there is little or no concerted attempt on the part of districts or schools to “game” the system.

Again, why do is there this national obsession with testing, and assuming that with more or better testing, more learning will follow? Where’s the evidence for this anywhere, in the research or the experience of any state or district? In North Carolina, the state that led the nation in terms of emphasizing standards and testing, the chair of a recent blue-ribbon commission concluded,

We’re testing more but we’re not seeing the results....We’re not seeing graduation rates increasing. We’re not seeing remediation rates decreasing. Somewhere along the way testing isn’t aligning with excellence.”

Finland, whose students score the highest in the world on the international assessments called the PISA, has no standardized tests in their schools, which, as the Washington Post points out, is “a stark contrast to the current test obsession in this country.”

Since when did more or better testing become identified as more learning? Or am I missing something here? Please post your responses on the blog.


Anonymous said...

I thought Gates and his ilk only liked "experimenting" on urban schools. Wonder how the suburbs would feel about him sticking his nose into their educational system and assessments? So, Scarsdale and Ardsley and Rockland, what do you think? Are you ready?

Anonymous said...

I'm not sure I see an "obsession" with national testing, but I would favor it for one reason alone: transparency. Right now, states have no incentive to set and maintain high standards. It's far easier to establish the illusion of progress by making the test easier each year (something that seemed to be happening while I taught 5th grade in NYC over several years). It allows apples to apples comparisons across states so in theory, we can get a better idea of what's working and not. Is it a panacea? Of course not. And as long as tests are high-stakes, there will be cheating, period.

If I could waive a wand, I'd have a single national standard(content, not empty process standards), a single curriculum, and a single test based on that content. And there would be no stakes. Let the feds present as clear and accurate a picture as possible as to where every school stands. If states and localities don't have an interest in improving failing schools, then as Diane Ravitch has noted, we have much bigger problems.

Diana Senechal said...

I agree mostly with Robert here. I favor a national test for two reasons: its transparency and its protection of subject matter. It should come with an optional national curriculum. The test would have no stakes; it would help states and districts ascertain how much students are learning.

For this to be of value, we would need an excellent curriculum to accompany the test. Both test and curriculum would test students' knowledge of actual subjects like literature and history. Right now, under programs like Balanced Literacy, the emphasis is on reading and writing strategies, not works of literature. In social studies, the emphasis on history is rather weak; you can pass a social studies test with minimal knowledge of the past. An excellent national curriculum would help ensure that students not only learn important skills but read powerful literary works and study the past closely.

Of course such a curriculum would need ideas and contributions from teachers, scholars in the individual subjects, education scholars, and other interested members of the public. Of course it should give states and districts some flexibility. But the underlying premise is golden: that students should actually know and read certain things.

Now, that's not always what people mean when they say "national standards" or "national test." I suspect that many who support a national test differ widely in their ideas of what it would be. If the vague standards (such as the New York State ELA standards) became national, we would be in trouble. So the question is: "Who says your idea of standards are better than other people's? What if some people love the NYS standards for their emphasis on "higher order thinking"?"

Well, we should argue it out. But let's have the debate. Let's have a public forum on national tests, standards, and curriculum. Let's discuss which literary works and authors students should read, which historical topics and primary documents they should study, which mathematical theorems they should understand, which areas of science they should know, and much more. I'll come if at all possible.

Diana Senechal

Anonymous said...

The first objection that will come up is "I don't want any federal bureaucrats deciding what my child should learn." The other object is invariably, "learning is not one size fits all."

But you know what? At it's most basic level, it IS one size fits all. Shouldn't ALL children learn the three branches of government? Shouldn't ALL children understand that matter exists in three forms? Shouldn't EVERY child learn how to add, subtract, multiply and divide? Is there something inherently wrong with suggesting, as Americans, we believe every kid should learn that the Earth revolves around the Sun?

These are standards. These are the things we think it's important for children to know. I'm bewildered that people take issue with that.

The bigger issue is, as you suggest Diana, process standards vs. content standards. Process standards are useless and content-free (literally).

Leonie Haimson said...

I agree that any national test should have no stakes attached; however, most of the people pushing them feel the opposite, and believe that stakes are critical for so-called "accountability" purposes, for districts, schools, and increasingly, staff and students in those schools.

Thus, the results of these tests will be less reliable and we will lose all sense of which schools are improving and which are not.

Anonymous said...


As a parent advocate, you're in a better position to lead and shape opinion on this issue than anyone. Rather than merely assume that testing will lead to no good end, why not try to lead the debate? Parents are clearly the greatest stakeholders in education. Therefore they are uniquely positioned to say to the Feds, "You bring the data, we'll bring the accountability." Concerned about what the tests will do to schools? In a no-stakes testing environment, who better that the parents at a particular school to say, "We don't want a minute of instructional time wasted on test prep." We want the tests to show what our kids can really do.