Arizona's student testing model is flawed, and the state's top education official is exaggerating student success on standardized tests, a conservative researcher charged Thursday.

"It's a bit like watching Barry Bonds and Sammy Sosa beating these baseball records," said Matthew Ladner, vice president of research at the Goldwater Institute. "It could be that they're just better baseball players. Or it could be that the ball is juiced or the players are taking steroids."

Ladner debated Tom Horne, Arizona's superintendent of public instruction, at an annual meeting of education researchers held at Arizona State University's Downtown Phoenix Campus.

Horne called Ladner a "demagogue" and said the Goldwater Institute is selective with facts and spreads false information as a scare tactic.

"They can't stand the idea that there could be anything good in public education," Horne said.

The Goldwater Institute is a Phoenix-based think tank advocating limited government and individual responsibility.

Ladner said Horne uses questionable statistics to claim that Arizona students score well above the national average in college entrance exams and standardized tests such as the TerraNova, which replaced the Stanford 9 in 2004 as a measure of student achievement that can be compared nationally. It is taken by all second- and ninth-graders in Arizona.

He pointed to a letter in the Arizona Daily Star in Tucson in which Horne said that Arizona students are scoring 8 percent above the national average on the TerraNova exam.

"Parents are out there making very important decisions based on these data," Ladner said. "I can no longer be certain my decision is based on data I can rely on."

Horne later said Arizonans score about 1.9 percent above the national average on TerraNova.

Ladner didn't dispute the number but said the TerraNova exam is an imprecise method for comparing students nationally.

He said Arizona's TerraNova results may be artificially high because some scores from third- through eighth-graders are merely estimates based on results from another test, Arizona's Instrument to Measure Standards (AIMS).

"If there's teaching to the test going on with AIMS, that would inflate TerraNova scores," Ladner said.

Horne responded with a graph showing relatively consistent results between students who took the TerraNova test and those whose scores were estimated from AIMS.

Ladner advocated another exam, the National Assessment of Educational Progress. Arizona students score below the national average on that test, known as The Nation's Report Card.

Congress created the Report Card in 1969 to measure student performance over time. It tests a representative sample of students in all 50 states and is often used to compare results between states.

Horne has been a sharp critic of the test, saying Arizona's scores are low because the test doesn't align with the state's curriculum.

He said it's not a measure of student achievement, but rather "a measure of how well as state has aligned its standards" with the test.

Arizona's schools are performing well but the results have been diluted by attacks from the media and "ideologues," he said.

"We'd like to see the Diamondbacks win the World Series," Horne said. "But that's not going to happen if people falsely think they're in ninth place."