Results of the 2007 Trends in International Mathematics and Science Study (TIMSS) were released Tuesday, and the data proves to be a bit of an educational Rorschach Test. The New York Times sees “solid achievement gains” in math by U.S. 4th and 8th graders, and “spectacular progress” by students in Minnesota and Massachsetts, while science performance remained flat nationwide. “The results showed that several Asian countries continued to outperform the United States greatly in science and math,” notes the Times, “subjects that are crucial to economic competitiveness and research.”
USA Today’s Greg Toppo sees American students “consistently better than average,” but notes that “if there were a math-and-science Olympics for elementary and middle schoolers…the USA never quite makes it to the medal podium.”
At Flypaper, the Fordham Foundation finds reasons to be cheerful. “American students have made steady gains in mathematics performance over the past decade. This progress was especially noteworthy at the eighth grade level, where the U.S. made gains since 1995 that were at least as strong as all of our major economic competitors.” Diane Ravitch disagrees however that 8th grade gains are “noteworthy.”
The gains posted by 8th graders are certainly not a vindication of No Child Left Behind’s testing regime. Eighth-graders registered a 12-point gain in math from 1995-2003, before the imposition of NCLB testing. They posted a 4-point gain from 2003-2007. The students who were tested by TIMSS in 2007 had been subject to NCLB annual tests in every year from third grade onward, yet their scores did not show a dramatic improvement. If anything, the gains were no greater (and possibly smaller) than those registered pre-NCLB.
Democratic Congressman George Miller sees “significant gains” in 4th grade math, but tells the Washington Post it’s “troubling that our students are still behind their international peers in both math and science.” Tom Loveless of the Brookings Institution, tells USA Today that the new TIMSS results belie complaints that U.S. students are lagging behind the rest of the world in math. “It’s just not true,” he says. “It hasn’t been true for a long time.” Meanwhile the National Science Teachers Association pronounces itself “discouraged” by the results, noting science scores for minority students are “dismal.” Many districts simply do not value science education, says a statement released by the NSTA Tuesday. ”Science is being eliminated from many K-6 classrooms.”
Ravitch misleads in dividing testing as pre and post 2003.
Many states began standardized testing in the 1990s. Indeed, NCLB only codified for all states what many states had undertaken.
Massachusetts, for example, launched MCAS in 1998.
If Ravitch believes the 12-point 1995-2003 gain is legit, then she can either they happened in spite of millions of American kids taking state standardized tests for the first time, or that those gains happened in part *because* of the testing regime.
Mike G.
No! No! No!.
It’s never that simple. Sure early adopters of numerous reforms have a pattern of doing better. For instance, Chicago had greater increases in the 90s before stakes were attached to testing.
But look at the big picture. When the economy grows, test scores go up. When recession hits, learning goes down. The economy is the biggest factor, but it’s not the whole story.
Also, rich states like Massachushetts have the resources to make holistic reforms.
Standardized testing, I suspect, was more destructive since 2003 because of the panic created by NCLB. Under pressure, people make worse decisions.
Mike,
NCLB fundamentally changed the way we do testing. Yes some states did testing before then, but a) the eighth-graders taking TIMSS in 1995 or 2003 very likely didn’t take annual end-of-course tests every year preceding, like this year’s cohort did (who are much more children of standards-based testing), and b) few states even in ‘03 had particularly robust standards/testing regimes. So if you want to look at testing, there has I think been more changes from 2003 to now, than from 1995 to 2003.
I think TIMSS is interesting more as a measure of where we sit in relation to everyone else, not how we have increased or decreased (we have other measures like NAEP focusing more specifically on that).
A comparison with PISA may also be instructive. In math, U.S. 4th & 8th graders outperform students in a number of countries–like Germany, Australia, Sweden, etc.–that outperform the U.S. on PISA. Do U.S. students really lose so much ground between ages 13 and 15? Or do the tests assess different things?
After we interviewed PISA guru Andreas Schleicher two months ago, Stanford mathematician Jim Milgram sent us a long critique of PISA–http://www.publicschoolinsights.org/node/2125#comments. He implies that Finland’s stellar showing on PISA says at least as much about PISA’s shortcomings as about Finland’s success. He believes TIMSS is the stronger test.
I don’t have a horse in this particular race, but the dissonance between the two assessments serves as another reminder that we should be careful in interpreting international assessments.
There was an interesting article in the WashPo by Neil Howe the other day about how the late Baby Boomers & early Gen Xers have done significantly worse from an academic achievement standpoint than those who came before & after them. If there really is something to his argument about the causes, then the rise in test scores during the 1990’s may not have been due to the implementation of pre-NCLB state accountability measures.
I have reached an age at which I am apparently at risk for more health conditions than at any other point in my life. Every time to go for a physical check-up it seems that I am due for some new screening exam. If we add up all the annuals, biannuals, every five years, once after age X, etc, it keeps me pretty busy.
Someone is paying attention to these things–that is someone has figured out that before a certain age, the cost/risk of the test outweighs the cost/risk of the condition. In the end, there is a combination of tests that generally works out to an expedient rate of identification and treatment. The tests, of course, don’t cure anything. On the other hand, without the tests, there might be no treatment until there were symptoms so obvious and annoying that I knew something wasn’t right and sought help.
By the same token–if I was screened and a health problem identified, then lost my insurance and couldn’t access the cure or treatment (or recoiled from the possibility and never followed up), then the test would be futile.
The point is, folks, the tests alone don’t matter. The tests plus the will and ability to respond appropriately to the results can and do. They are necessary but not sufficient. NECESSARY. But not sufficient.