A Harvard professor responded to an New York Times article critical of his study on school vouchers yesterday, writing a letter to the Times defending his work and suggesting that the article exaggerated the criticism.

Last Friday, the Times reported that Mathematica Policy Research, a think tank that provides academic data, took the unusual step of questioning the conclusions of a school vouchers study by Shattuck Professor of Government Paul E. Peterson.

Yesterday, Peterson and his colleagues wrote a joint letter to the Times suggesting that the article generated controversy where there was none to be found.

"Your story suggests a level of disagreement among researchers that simply does not exist," the authors write.

The two-year study found that, in three different cities, vouchers had a statistically significant effect in raising the test scores of black students although no corollary effect was seen for other racial groups.

Peterson's findings were quickly picked up by voucher supporters and lauded as some of the first genuine evidence in favor of the use of vouchers.

But last week's press release by Mathematica called the validity of the study into doubt, questioning Peterson's use of data. The release prompted Kate Zernicke's article, "New Doubt is Cast on Study that Backs Voucher Effects," in Friday's Times.

The release claimed the New York segment of the study shows "no overall difference in test scores between those who were offered vouchers and those who were not."

Peterson said some questions do remain over exactly how to interpret the data from his study, but added that there is no controversy over the importance of the findings to the voucher debate.

No Debate Among Academics?

The research study gathered data in Dayton, Ohio, Washington D.C. and New York. In each city, the study reported that test scores of the black students using vouchers rose considerably, averaging a six percent increase over the two years of the study.

In New York, black students' test scores over the grades analyzed rose four percent. This number, Peterson said, is not and should not be under debate.

"Those findings are not in any way disputed by Mathematica. They agree that the effects in New York were statistically significant," Peterson said.

However, when the figures were broken down by grade, Mathematica found no increase in test scores among third through fifth graders, with the only significant increase in the sixth grade classes.

It was this analysis that prompted the press release.

"Until we understand this better, we cannot place much policy weight on it," the release quoted David Myers, an investigator in the study, as saying.

To Peterson, the dispute remains a question only of technicalities.

"There is no real debate among academics," Peterson said, noting that Myers was a joint author of one of his studies.

According to Peterson, the number of study participants was not sufficient to make a grade-by-grade analysis worthwhile, although he said he hopes it will be done in the future.

Research into voucher education currently remains thin--a necessary consequence of the relative novelty of vouchers, Peterson said.

The Voucher Story

Education vouchers represent an attempt to jumpstart school reform without increasing funding.

For many years in the U.S., more federal money in schools meant better performance. In recent years, however, this relationship has not held up.

Now, many academics and politicians say they hope vouchers can inject new life into the school system. By simultaneously giving lower-income students the chance to attend private schools and increasing the competition for students among public schools, vouchers can give better schooling to all students, supporters say.

But to opponents, this vision is a pipe dream that will end up funneling federal funds into private schools.

"Public money ought to be used for public schools," said Stephen E. Gorrie, the president of the Massachusetts Teachers Association. "You're just setting the schools up for failure by taking money away from them for a set few."