NOXVILLE, Tennessee - Contrary to claims by George W. Bush that education in Texas has vastly improved under his watch, a new report finds a growing racial gap in the state as well as signs that reading and math advances have been greatly overblown.
Although students did much better on a state-administered test in 1998 than in 1994 - giving rise to the idea of a ''Texas miracle'' in education, a centerpiece of the Bush campaign - those improvements were barely reflected when the same students took a national test, according to findings released yesterday by the Rand Corporation, a nonpartisan think tank.
And while the gap between minority and white students shrank dramatically in the state test, giving Bush reason to boast of great advances during his first term, the racial gap grew in the national test. That led researchers to voice ''serious questions about the validity of the gains'' in Texas - and cast doubt on Bush's claim that he has revolutionized the classroom as Texas governor.
Stephen Klein, a senior R and researcher who helped author the report, said the ''Texas miracle'' was a ''myth.''
That assessment, two weeks before the election, created an instant outcry from Bush aides, who denounced the report as partisan and lightweight. Calling the findings ''highly suspect,'' Bush communications director Karen Hughes pointed to a three-year Rand study released last summer that showed Texas test scores dramatically improved between 1990 and 1996.
The Gore campaign seized on the report, and quickly fit it into a broader strategy of discrediting Bush's six-year record of public service.
''This Rand report totally undercuts confidence in what Governor Bush has been telling us about his great success in Texas. It's one more part of the Texas record for the American people to worry about,'' said Senator Joseph I. Lieberman, Gore's running mate.
Bush, governor since 1995, has made education a key focus of both his state administration and his campaign, promising to end what he has dubbed the ''education recession'' if elected president. On the campaign trail, he frequently holds events at public schools, introduced by his wife, Laura, a former school teacher.
He has two competing visions: On the one hand, Bush often chastises the Clinton administration for having failed to raise national test scores, promising that a Bush administration would demand ''greater accountability'' from schools that receive federal funds. But Bush is also firm in his belief that most education priorities should be left up to states.
And almost always, the soaring test scores in Texas are a centerpiece of his federalist argument - proof that governors are the best shepherds of public schools.
Last summer, a R and study confirmed part of his claim, although it only covered two years Bush was in office. Between 1990 and 1996, Texas students made much greater improvements on the National Assessment of Educational Progress than their counterparts in other states, in a study that weighted the scores to eliminate socioeconomic factors and race.
The report released yesterday analyzed a different set of data. Focused almost entirely on the time Bush has been governor, 1994 to 1998, the study looked only at Texas, to determine whether state tests and national tests showed similar trends.
In fact, students who did dramatically better on the Texas Assessment of Academic Skills test showed only modest gains on the national test. That raised concerns that Texas teachers were teaching students how to do well on th e specific test rather than overall reading and math skills.
Another possibility, Klein said, is that the Texas test is not difficult enough.
Acknowledging they did not know the reasons for the disparity, the researchers concluded: ''The large discrepancies between TAAS and NAEP results raise serious questions about the validity of the TAAS scores.''
The Bush campaign dismissed the report, saying it was the work of an isolated group of researchers with suspicious motives. ''The timing of this opinion, this 14-page opinion paper, conducted by four researchers, is highly suspect and the conclusions are dead wrong,'' Hughes said.
But R and president James A. Thomson fully backed the report, saying it had undergone review by field experts. He also said it had no political motivation; had it been released after the election, R and officials said, the researchers would have been accused of protecting Bush.
''The timing of our report was not based on political consideration,'' Klein said.
The report was a distraction for Bush, who had hoped to spend the day focused solely on tax cuts, an issue he highlighted at the Thomas Middle School in Arlington Heights, Ill.
Bush appeared confident, although his oratory took several unexpected turns. Responding to a question about abortion, Bush said he would promote a ''culture of life,'' adding that ''it's not only life of babies, but it's life of children living in the dark dungeons of the Internet.''
