What You Missed When You Were Learning About Dropouts

While nearly everyone else on the education beat was falling all over themselves to cover the latest dropout report -- they're smart, they're bored, it's a silent epidemic -- AP's Ben Feller and a select few others pursuaded themselves (and their editors) that it might be more worwhile to focus instead on the issue of whether kids seem to be improving on NAEP and how incredibly large the gaps are between what NAEP says and what state tests say. Less sexy, to be sure, but probably more substantive.


It's unusual that the EdTrust comes in behind the dropout folks. Usually, the Trust has done astoundingly well at getting the word out and dominating the day's coverage. But this time, at least, the Gates-funded organization that did the dropout study had their number. Daria Hall, senior policy analyst -- a new name to me who is she? -- must have been wondering why her phone didn't ring more. Blame it on the dropout study, Daria.

Of course, the fact that kids in many states appear to be much smarter on state tests than on NAEP is not news to most of us. Lots of folks have talked about it, and some use the gap to push for a national testing system or some NAEP-NCLB linkage. I don't think that's happening anytime soon. Why not? As Feller puts it: "The state tests have consequences for schools. The federal test does not."

But it's good to be reminded of how big the gap is: 30, 40 or 50 percentage points lower on the federal exam.Feller calls out some states you'd expect -- Mississippi, Colorado, West Virginia -- and some you wouldn't like North Carolina for large gaps between state and NAEP results. I never knew the NC state tests were set that low. Nor did I know that some states have higher requirements than NAEP, and kids do better on NAEP than on the state tests.

Interestingly, the report isn't really about the NAEP-state gap. Its focus is on whether anyone's making progress on NAEP from 2003 to 2005, by state. "Most states have made progress in raising achievement in the elementary grades, but secondary schools still struggle to close gaps between poor and minority students and their White and more affluent peers."

Primary Progress, Secondary Challenge (PDF).


Anonymous Bob said...


As one who has followed the NAEP achievement levels fairly closely since their inception, I find the continual NAEP-state gap discussion fairly infuriating. I have two main problems with it. First, why is the assumption always that the NAEP levels are right and the state levels are wrong? Yes, there is the suspicion that states are low-balling their standards so they look better. But there is no evidence that the NAEP levels reflect "grade-level" performance, as Susan Aspey says in Ben Feller's piece. In fact, there's plenty of evidence that the NAEP levels are unreasonably high. As an NRC committee found, NAGB chose to lower the levels in science because TIMSS had just come out showing that US students performed pretty well compared with other students, and nobody would have believed NAEP results that showed that virtually no US student was advanced in science.

Second, the levels might be different because NAEP and state tests are *different tests*! Just because they both measure "math" doesn't mean they measure the same things. Although there are doubts that state tests are truly aligned with state standards, at least they are designed to measure those standards. States don't pretend to align their tests with NAEP. And NAEP is generally broader than state tests because the framework on which it is based represents a consensus among all states. When people tried to discredit Texas by pointing out that gains on the state test were not matched by gains on NAEP, they failed to point out that Texas did show gains on the parts of NAEP that matched their state test.

In addition to these issues, tere are also serious technical problems with the NAEP achievement levels, which many testing experts have found. NCES includes a disclaimer about the levels, yet reports them anyway.

I hope you're right that a NAEP-NCLB linkage won't happen anytime soon.

9:52 AM  

Post a Comment

<< Home