Daily Howler logo
STATE TESTS THAT WORK! Statewide test show massive gains. But the NAEP records no such progress: // link // print // previous // next //
THURSDAY, OCTOBER 20, 2005

THE SEARCH FOR A SEARCH: Did Rove and Libby (and others) commit crimes? Soon, we’ll get Fitzgerald’s (and/or the grand jury’s) judgment. Meanwhile, will 22 people really get indicted? If so, someone could shoot The Longest Yard 3 without having any prison guards play! Or could this mean that prison teams have now gone to the two-platoon system? No wonder it costs nine bucks to see a top film when Hollywood goes so crazy with costs.

Meanwhile, with all the local comedy excitement, we’ll postpone till tomorrow our troubling treatment of a basic question about the Plame matter: Since Wilson’s op-ed didn’t really contradict what Bush had said about Niger, why did the press corps act like it did? (To this day, they shave basic facts to keep this perception alive.)

In the meantime, Matt Yglesias quotes the Duelfer Report about a central question: Was Iraq ever seeking uranium? Key quote from the report: “[The Iraq Survey Group] has not found evidence to show that Iraq sought uranium from abroad after 1991.” In his State of the Union, Bush said that British intelligence had learned that Iraq sought uranium from Africa. The Iraq Survey Group has “not found evidence” of any such pursuit.

That’s significant information, but Matt is almost surely wrong when he says “we can now rely upon the Duelfer Report for everything we need to know about Iraq’s nuclear program.” That would treat the Duelfer as scripture—not the wisest procedure. But then, we’re all inclined to go a bit scriptural at super-charged partisan moments like this. For example, try to believe that someone as smart as Matt could make a statement like this:

YGLESIAS: We also know that the CIA made a criminal referral in this case long ago, so there whole notion that Plame somehow wasn't "really" covert is a nonstarter.
Incredible! If the CIA says it, it has to be true! That someone so smart could say something so odd helps define the strange state we’re now in.

STATE TESTS THAT WORK: Is the No Child Left Behind program “working?” There’s no way to tell from the newly-released, 2005 scores on the National Assessment of Educational Progress (NAEP), the “comprehensive reading and math examination given to hundreds of thousands of students periodically since 1990" (Sam Dillon, New York Times). The program passed in December 2001; if it’s going to have an effect, it wouldn’t necessarily be visible yet. And if it were causing marginal upticks in the short time it has been in effect, those upticks might be disguised by other factors in the student population—factors which might be hard to measure in the short term.

That said, the NAEP is worth tracking. We’re not experts on the NAEP (we’ll get ourselves up to speed), but we’d guess that its results are more reliable than those from many other measures. It tests large numbers of students nationwide, and because it isn’t a locally-devised, “high stakes” program, local teachers have little incentive to monkey around with its administration. Teacher salaries and principal rankings don’t turn on results of the NAEP. We’d guess that it’s a more reliable measure than many local “high stakes” tests—tests which may be more widely discussed.

That said, NAEP reading scores are quite flat over its fifteen-year history. Fourth-grade reading? On a 500-point scale, the average score was 217 in 1992 (first year available)—and 219 this year. And 219 was also the score in 2002—the year before No Child Left Behind had any imaginable effect. According to the NAEP, the fourth-grade reading score hasn’t changed in the past three years—and has barely changed in the past thirteen.

But that just isn’t what we see when we look at some state-made test programs. In the past few weeks, we’ve been reviewing test scores from three states involved in the recent PBS program, Making Schools Work with Hedrick Smith. (Does Hedrick Smith have to be there before these schools can work?) Each of these states (Washington, North Carolina, Illinois) administers an annual, state-produced, high-stakes test battery. And in two of those states, elementary reading scores have shot up over the past dozen years. These results are pleasing—but they suggest student gains which haven’t been found by the NAEP on the national level.

Consider the state of Washington. Each year, the state administers its own “WASL” tests (never mind what that stands for). And holy cow! In 1997 (first year available), 47.9 percent of fourth-graders passed the WASL reading test. By 2005, that had jumped to 79.5 percent. Citizens of Washington probably think fourth-graders are doing much better in reading. Indeed, the WASL results suggest massive gains—a type of progress the NAEP hasn’t found on the national level.

Ditto for North Carolina. The state administers annual “End-of-grade” tests—and according to these high-profile tests, Tarheel fourth graders have also been making big progress in reading. In 1994 (first year available), 44.0 percent of fourth graders tested “proficient” on the state reading test. Last year: 83.5 percent.

Only Illinois, of the three states we’ve looked at, has recorded fairly stable scores (albeit over a shorter time span). The state’s “ISAT” program tests third- and fifth-graders in reading each year. In 1999 (first year available), 61.3 percent of third-graders seem to have “met or exceeded standards.” (Warning: The state’s reports are quite confusing.) In 2004, the number seems to have been 65.0 percent. Fifth-graders showed a more modest gain. In 1999, 59.3 percent seem to have “met or exceeded” in reading; in 2004, 60.9 percent seem to have done so. (Those values can be found here. Elsewhere in the state’s confounding reports, slightly different numbers obtain.)

Everybody loves it when passing rates jump. But the NAEP records no such progress on the national level. In 1992 (first year available), 62 percent of fourth graders tested “at or above basic” on the NAEP reading test. In both 2002 and 2005, the number was 64 percent. In 2003, it was 63 percent.

Statewide audiences rock and roll when they see those statewide passing rates rise—but the NAEP suggests this may be an illusion. Are fourth graders in North Carolina and Washington doing much better? Or are some state-made, high-profile tests getting easier over the years?

A (POSSIBLE) GOOD POINT: What a surprise! In the “nothing much ever changes” category, Bush and Ed Sec Margaret Spelling were out there pimping the new non-results, looking for signs of pleasing progress. But Spellings made a (possible) point about those long-term score trends:

DILLON (10/19/05): Mr. Bush, meeting with Education Secretary Margaret Spellings at the White House, said he was pleased with the test. ''It shows there's an achievement gap in America that is closing,'' Mr. Bush said.

In an interview, Ms. Spellings called attention to the improvement in math by fourth graders. She said the less robust increases and outright declines in some reading scores were understandable in part, because the nation’s schools are assimilating huge numbers of immigrants.

''We have more non-native speakers, there are lots of so-called at-risk, hard-to-educate students, and in spite of that, steady progress is being made,'' she said. ''We're on the right track with No Child Left Behind.”

In short: If schools have many more non-English speakers, then stable scores may suggest improved overall performance. But are “native speakers” doing better on the NAEP? Dillon quoted Spellings’ pitch, then declined to say.