Facebook Twitter

Some school reforms pay off, Rand reports

SHARE Some school reforms pay off, Rand reports

WASHINGTON — Some education reforms work better than others, according to a massive Rand Corporation report released here Tuesday.

Researchers from the California-based research center concluded that educational systems in states such as North Carolina, Texas and Michigan did the best job of improving student achievement between 1990 and 1996.

But the researchers did not endorse other reforms that had been introduced during the previous decade in poor-performing states such as California, Louisiana and Mississippi.

The researchers found that the highest-ranking states had:

— Higher per-pupil spending.

— Lower teacher-pupil ratios in the four lowest grades.

— More teachers with adequate instructional materials and other resources.

— Higher attendance at public pre-kindergarten classes.

— Lower teacher turnover.

But they found that some factors accounted for little or none of the performance gaps between states.

"Other things being equal, higher teacher salaries, higher teacher educational levels and increased experience (in teaching) do not show significant effects on achievement," said the report. It also said that teacher aides don't make much difference in how well a state's students perform.

However, pay levels, education, longevity and help from teacher aides may be more important in the competition for the best teachers between school districts within a state, the researchers said.

The Rand researchers measured achievement by comparing the average scores from students in 44 states on reading and math tests administered by the federal government's National Assessment of Educational Progress. States with big gains or with higher scores after adjustments for socio-economic differences were rated as performing well.

"The main message is, public schools are reformable," said veteran Rand researcher David Grissmer, who led the three-year study.

"It takes more spending to raise academic achievement, but the additional resources must be well-targeted," Grissmer said at the National Press Club. "Several states have shown it can be done with moderate amounts of money if it is done in the right way."

He said the reforms with the biggest bang for the buck were:

— In all states, providing adequate classroom materials for all teachers in kindergarten through eighth grade.

— In states with more disadvantaged students, offering pre-kindergarten and reducing pupil-teacher ratios in the early grades below the national average of 21 to 1.

— In states with an average number of disadvantaged students, reducing pupil-teacher ratios in the early grades to the national average.

— Not spending money to hire teacher aides.

The report contains ammunition both for supporters of both major-party presidential hopefuls, Grissmer said. For example, the report's conclusions support programs such as the early reading initiative of Gov. George W. Bush in Texas and the national class-size reduction program touted by Vice President Al Gore.

"It's up to them how they spin it," Grissmer said. But he said the report showed that, like other current office-holders, Bush and Gore can't claim yet that their pet reforms are working, because "reforms take 10 or 15 years to play out."

For example, he said, decisions made many years ago helped Texas students earn much higher scores than California students on NAEP tests administered from 1990 to 1996.

High Texas scores could be traced back to a state commission headed by Ross Perot, which obtained the expansion of public pre-kindergarten programs in the mid-1980s.

Low California scores could be traced back as far as 197nd increased class size as immigration rates soared.

Published by Rand, "Improving Student Achievement: What State NAEP Test Scores Tell Us," is based on three years of research underwritten by the ExxonMobil Foundation, Danforth Foundation and U.S. Education Department.


On the Web: Rand Corporation — www.rand.org

Andrew Mollison's e-mail address is AndyM@Coxnews.com.